This action might not be possible to undo. Are you sure you want to continue?

BooksAudiobooksComicsSheet Music### Categories

### Categories

Scribd Selects Books

Hand-picked favorites from

our editors

our editors

Scribd Selects Audiobooks

Hand-picked favorites from

our editors

our editors

Scribd Selects Comics

Hand-picked favorites from

our editors

our editors

Scribd Selects Sheet Music

Hand-picked favorites from

our editors

our editors

Top Books

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Audiobooks

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Comics

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Sheet Music

What's trending, bestsellers,

award-winners & more

award-winners & more

P. 1

Introduction to Economic Growth - Daron Acemoglu - 2006|Views: 160|Likes: 2

Published by dinh son my

Introduction to Economic Growth - Daron Acemoglu - 2006

Introduction to Economic Growth - Daron Acemoglu - 2006

See more

See less

https://www.scribd.com/doc/153922411/Introduction-to-Economic-Growth-Daron-Acemoglu-2006

06/29/2015

text

original

- Introduction
- 1.1 A Quick Look at the Facts
- 1.2 Interpretation
- 1.3 The Agenda
- 2.1 The Basic Model in Discrete Time
- 2.1.1 The Production Structure
- 2.1.2 Endowments
- 2.1.3 Fundamental Law of Motion of the Solow Model
- 2.1.4 Deﬁnition of Equilibrium
- 2.1.6 Transitional Dynamics in the Solow Model
- 2.2 The Solow Model in Continuous Time
- 2.2.1 From Diﬀerence to Diﬀerential Equations
- 2.2.3 A First Look at Sustained Growth
- 2.3 Solow Model with Technological Progress
- 2.3.1 Balanced Growth
- 2.3.2 Neutral Technological Progress
- 2.3.3 The Steady-State Technological Progress Theorem
- The Solow Model and the Data
- 3.1 Growth Accounting
- 3.2 Solow Model and Cross-Country Income Diﬀer-
- 3.2.1 Solow Model with Human Capital
- 3.2.2 Problems with the Mankiw, Romer and Weil Approach
- 3.2.3 The Macro Mincer Approach (Bils-Klenow-Rodriguez-Hall-
- Fundamental Determinants of Diﬀerences in Income
- 4.1 From Proximate to Fundamental Causes
- 4.2 Hypotheses
- 4.3 Europe’s Expansion and Colonial Origins of Insti-
- Neoclassical Growth
- Towards Neoclassical Growth
- 5.1 Representative Consumer
- 5.2 Problem Formulation
- 5.3 Welfare Theorems
- 5.4 Optimal Growth in Discrete Time
- 5.5 Optimal Growth in Continuous Time
- Dynamic Programming and Optimal Growth
- 6.1 Brief Review of Dynamic Programming
- 6.2 Digression: Technical Details
- 6.2.1 Contraction Mappings
- 6.2.2 Application of Contraction Mappings to Dynamic Program-
- 6.3 Back to the Fundamentals of Dynamic Program-
- 6.3.1 Basic Equations
- 6.3.2 Dynamic Programming Versus the Sequence Problem
- 6.4 Optimal Growth in Discrete Time
- 6.5 Competitive Equilibrium Growth
- Brief Review of Optimal Control
- 7.1 Finite-Horizon Optimal Control
- 7.1.1 The Fundamental Problem
- 7.1.2 Variational Arguments
- 7.1.3 Simpliﬁed Maximum Principle
- 7.1.4 Generalizations
- 7.1.5 Limitations
- 7.2 Inﬁnite-Horizon Optimal Control
- 7.2.1 The Basic Problem: Necessary and Suﬃcient Conditions
- 7.2.2 Lack of Transversality Conditions
- 7.2.3 Discounted Inﬁnite-Horizon Optimal Control
- The Neoclassical Growth Model
- 8.1 Preferences, Technology and Demographics
- 8.2 Characterization of Equilibrium
- 8.2.1 Deﬁnition of Equilibrium
- 8.2.2 The Consumer Problem
- 8.2.3 Equilibrium Prices
- 8.3 Optimal Growth
- 8.4 Steady-State Equilibrium
- 8.5 Transitional Dynamics
- 8.6 Technological Change and the Canonical Neoclas-
- 8.7 The Role of Policy
- 8.8 Quantitative Evaluations
- 8.8.1 Policy Diﬀerences
- 8.8.2 Extensions
- 8.9 Variants of the Neoclassical Model
- Growth with Overlapping Generations
- 9.1 Problems of Inﬁnity
- 9.2 Overlapping Generations and Overaccumulation
- 9.2.1 Demographics, Preferences and Technology
- 9.2.2 Consumption Decisions
- 9.2.3 Equilibrium
- 9.2.4 More Speciﬁc Utility Functions
- 9.2.5 Pareto Optimality
- 9.3 Role of Social Security in Capital Accumulation
- 9.3.1 Fully Funded Social Security
- 9.3.2 Unfunded Social Security
- Recitation Material: Stochastic Growth
- 10.1 The Brock-Mirman Model
- 10.2 Application: Risk, Diversiﬁcation and Growth
- 10.2.1 The Environment
- 10.2.2 Equilibrium
- 10.2.3 Dynamics
- 10.2.4 Eﬃciency
- 10.2.5 Implications
- 10.2.6 Ineﬃciency with Alternative Market Structures
- Endogenous Growth
- First-Generation Models of Endogenous Growth
- 11.1 AK Model Revisited
- 11.1.1 Demographics, Preferences and Technology
- 11.1.2 Equilibrium
- 11.1.3 Transitional Dynamics
- 11.1.4 The Role of Policy
- 11.2 The Extended AK Model
- 11.3 Growth with Externalities
- 11.3.1 Preferences and Technology
- 11.3.2 Equilibrium
- 11.3.3 Pareto Optimal Allocations
- Multiple Equilibria and the Process of Development
- 12.1 Multiple Equilibria From Aggregate Demand Ex-
- 12.1.1 Preferences and Technology
- 12.1.2 Equilibrium
- 12.2.1 A Simple Case With No Borrowing
- 12.2.2 The Galor and Zeira Model
- 12.3.1 Demographics, Preferences and Technology
- 12.3.2 Equilibrium
- Interdependence and Growth in the Open Economy
- 13.1 Human Capital and Technology (Nelson-Phelps)
- 13.2 Trade and Technology Diﬀusion
- 13.2.1 The Basic Krugman Model
- 13.2.2 Understanding the Eﬀects of Trade
- 13.3 Trade, Specialization and the World Income Dis-
- 13.3.1 The Model
- 13.3.2 Equilibrium
- 13.3.3 Implications
- 13.4 Growth with Factor Price Equalization
- Endogenous Technological Change
- Expanding Variety Models
- 14.1.1 Demographics, Preferences and Technology
- 14.1.2 Digression on Continuous Time Value Functions
- 14.1.3 Characterization of Equilibrium
- 14.1.4 Deﬁnition of Equilibrium
- 14.1.5 Steady State
- 14.1.6 Transitional Dynamics
- 14.1.7 Pareto Optimal Allocations
- 14.1.8 Policy in the Endogenous Technology Model
- 14.2 Growth with Knowledge Spillovers
- 14.2.1 The Role of Competition Policy
- 14.3 Growth without Scale Eﬀects
- Models of Quality Competition
- 15.1 Baseline Model
- 15.2 Pareto Optimality
- Directed Technical Change
- 16.1 Basics and Deﬁnitions
- 16.1.1 Deﬁnitions
- 16.1.2 Basic Model
- 16.1.3 Implications
- 16.2 Equilibrium Technology Bias: Some More Gen-
- 16.3 EndogenousLabor-AugmentingTechnologicalChange
- 16.3.1 Demographics, Preferences and Technology
- 16.3.2 Consumer and Firm Decisions
- 16.3.3 Asymptotic and Balanced Growth Paths
- 16.3.4 The Balanced Growth Path
- 16.3.5 Transitional Dynamics
- 16.3.6 Policy Implications
- Recitation Material: Appropriate Technology
- 17.1 DiﬀerencesinCapital-LaborRatios(Atkinson-Stiglitz)
- 17.2 The Role of Human Capital (Acemoglu-Zilibotti)
- 17.2.1 A Model
- 17.2.2 Implications
- 17.2.3 Calibration
- Epilogue: Political Economy of Growth
- 18.1 Thinking of Institutions and Growth
- 18.1.1 The Impact of Institutions
- 18.1.2 Modeling Institutional Diﬀerences
- 18.1.3 Institutions in Action
- 18.2 A Simple Model of Non-Growth Enhancing Insti-
- 18.2.1 Baseline Model
- 18.2.2 Economic Equilibrium
- 18.2.3 Ineﬃcient Policies
- 18.2.4 Revenue Extraction
- 18.2.5 Factor Price Manipulation
- 18.2.6 Revenue Extraction and Factor Price Manipulation Com-
- 18.2.7 Political Consolidation
- 18.2.8 Subgame Perfect Versus Markov Perfect Equilibria
- 18.2.9 Lack of Commitment–Holdup

**451 Introduction to Economic Growth
**

Daron Acemoglu MIT Department of Economics January 2006

14.451: Introduction to Economic Growth

ii

Contents

I Introduction 1

5 5 16 18 19 19 19 23 27 28 29 35 41 41 42 49

1 Stylized Facts of Economic Growth and Development 1.1 A Quick Look at the Facts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 The Agenda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 The Solow Growth Model 2.1 The Basic Model in Discrete Time . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 2.1.2 2.1.3 2.1.4 2.1.5 2.1.6 The Production Structure . . . . . . . . . . . . . . . . . . . . . . . . Endowments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fundamental Law of Motion of the Solow Model . . . . . . . . . . . . Deﬁnition of Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . Equilibrium Without Population Growth and Technological Progress Transitional Dynamics in the Solow Model . . . . . . . . . . . . . . .

2.2 The Solow Model in Continuous Time . . . . . . . . . . . . . . . . . . . . . . 2.2.1 2.2.2 2.2.3 From Diﬀerence to Diﬀerential Equations . . . . . . . . . . . . . . . . The Fundamental Equation of the Solow Model in Continuous Time . A First Look at Sustained Growth . . . . . . . . . . . . . . . . . . . iii

14.451: Introduction to Economic Growth 2.3 Solow Model with Technological Progress . . . . . . . . . . . . . . . . . . . . 2.3.1 2.3.2 2.3.3 2.3.4 Balanced Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral Technological Progress . . . . . . . . . . . . . . . . . . . . . The Steady-State Technological Progress Theorem . . . . . . . . . . . The Solow Growth Model with Technological Progress: Continuous Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 The Solow Model and the Data 3.1 Growth Accounting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Solow Model and Cross-Country Income Diﬀerences . . . . . . . . . . . . . . 3.2.1 3.2.2 3.2.3 Solow Model with Human Capital . . . . . . . . . . . . . . . . . . . . Problems with the Mankiw, Romer and Weil Approach . . . . . . . . The Macro Mincer Approach (Bils-Klenow-Rodriguez-Hall-Jones) . . 59 63 63 66 66 71 75 79 83 83 84 89 51 51 53 55

3.3 An Alternative Approach to Estimating Productivity Diﬀerences (Treﬂer) . . 4 Fundamental Determinants of Diﬀerences in Income 4.1 From Proximate to Fundamental Causes . . . . . . . . . . . . . . . . . . . . 4.2 Hypotheses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Europe’s Expansion and Colonial Origins of Institutions . . . . . . . . . . .

II

Neoclassical Growth

95

99

5 Towards Neoclassical Growth

5.1 Representative Consumer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 5.2 Problem Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 5.3 Welfare Theorems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 iv

14.451: Introduction to Economic Growth 5.4 Optimal Growth in Discrete Time . . . . . . . . . . . . . . . . . . . . . . . . 109 5.5 Optimal Growth in Continuous Time . . . . . . . . . . . . . . . . . . . . . . 111 6 Dynamic Programming and Optimal Growth 113

6.1 Brief Review of Dynamic Programming . . . . . . . . . . . . . . . . . . . . . 114 6.2 Digression: Technical Details . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 6.2.1 6.2.2 Contraction Mappings . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Application of Contraction Mappings to Dynamic Programming . . . 123

6.3 Back to the Fundamentals of Dynamic Programming . . . . . . . . . . . . . 135 6.3.1 6.3.2 Basic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Dynamic Programming Versus the Sequence Problem . . . . . . . . . 138

6.4 Optimal Growth in Discrete Time . . . . . . . . . . . . . . . . . . . . . . . . 141 6.5 Competitive Equilibrium Growth . . . . . . . . . . . . . . . . . . . . . . . . 146 7 Brief Review of Optimal Control 149

7.1 Finite-Horizon Optimal Control . . . . . . . . . . . . . . . . . . . . . . . . . 150 7.1.1 7.1.2 7.1.3 7.1.4 7.1.5 The Fundamental Problem . . . . . . . . . . . . . . . . . . . . . . . . 150 Variational Arguments . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Simpliﬁed Maximum Principle . . . . . . . . . . . . . . . . . . . . . . 154 Generalizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

7.2 Inﬁnite-Horizon Optimal Control . . . . . . . . . . . . . . . . . . . . . . . . 160 7.2.1 7.2.2 7.2.3 The Basic Problem: Necessary and Suﬃcient Conditions . . . . . . . 160 Lack of Transversality Conditions . . . . . . . . . . . . . . . . . . . . 163 Discounted Inﬁnite-Horizon Optimal Control . . . . . . . . . . . . . . 164 v

14.451: Introduction to Economic Growth 8 The Neoclassical Growth Model 167

8.1 Preferences, Technology and Demographics . . . . . . . . . . . . . . . . . . . 167 8.2 Characterization of Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . 173 8.2.1 8.2.2 8.2.3 Deﬁnition of Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . 173 The Consumer Problem . . . . . . . . . . . . . . . . . . . . . . . . . 174 Equilibrium Prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

8.3 Optimal Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 8.4 Steady-State Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 8.5 Transitional Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 8.6 Technological Change and the Canonical Neoclassical Model . . . . . . . . . 185 8.7 The Role of Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192 8.8 Quantitative Evaluations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 8.8.1 8.8.2 Policy Diﬀerences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 . . . . . . . . . . . . . . . . . . . . . . . 199 203

8.9 Variants of the Neoclassical Model

9 Growth with Overlapping Generations

9.1 Problems of Inﬁnity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 9.2 Overlapping Generations and Overaccumulation . . . . . . . . . . . . . . . . 206 9.2.1 9.2.2 9.2.3 9.2.4 9.2.5 Demographics, Preferences and Technology . . . . . . . . . . . . . . . 206 Consumption Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 More Speciﬁc Utility Functions . . . . . . . . . . . . . . . . . . . . . 210 Pareto Optimality . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

9.3 Role of Social Security in Capital Accumulation . . . . . . . . . . . . . . . . 217 vi

14.451: Introduction to Economic Growth 9.3.1 9.3.2 Fully Funded Social Security . . . . . . . . . . . . . . . . . . . . . . . 217 Unfunded Social Security . . . . . . . . . . . . . . . . . . . . . . . . . 219 221

10 Recitation Material: Stochastic Growth

10.1 The Brock-Mirman Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 10.2 Application: Risk, Diversiﬁcation and Growth . . . . . . . . . . . . . . . . . 223 10.2.1 The Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 10.2.2 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 10.2.3 Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 10.2.4 Eﬃciency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 10.2.5 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 10.2.6 Ineﬃciency with Alternative Market Structures . . . . . . . . . . . . 234

III

Endogenous Growth

239

243

11 First-Generation Models of Endogenous Growth

11.1 AK Model Revisited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 11.1.1 Demographics, Preferences and Technology . . . . . . . . . . . . . . . 244 11.1.2 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246 11.1.3 Transitional Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 247 11.1.4 The Role of Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 11.2 The Extended AK Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 11.3 Growth with Externalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 11.3.1 Preferences and Technology . . . . . . . . . . . . . . . . . . . . . . . 256 11.3.2 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 11.3.3 Pareto Optimal Allocations . . . . . . . . . . . . . . . . . . . . . . . 261 vii

14.451: Introduction to Economic Growth 12 Multiple Equilibria and the Process of Development 263

12.1 Multiple Equilibria From Aggregate Demand Externalities . . . . . . . . . . 264 12.1.1 Preferences and Technology . . . . . . . . . . . . . . . . . . . . . . . 264 12.1.2 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 12.2 Human Capital Accumulation with Imperfect Capital Markets . . . . . . . . 275 12.2.1 A Simple Case With No Borrowing . . . . . . . . . . . . . . . . . . . 276 12.2.2 The Galor and Zeira Model . . . . . . . . . . . . . . . . . . . . . . . 279 12.3 Learning-by-Doing, Structural Change and Non-Balanced Growth . . . . . . 283 12.3.1 Demographics, Preferences and Technology . . . . . . . . . . . . . . . 283 12.3.2 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 13 Interdependence and Growth in the Open Economy 289

13.1 Human Capital and Technology (Nelson-Phelps) . . . . . . . . . . . . . . . . 289 13.2 Trade and Technology Diﬀusion . . . . . . . . . . . . . . . . . . . . . . . . . 291 13.2.1 The Basic Krugman Model . . . . . . . . . . . . . . . . . . . . . . . . 291 13.2.2 Understanding the Eﬀects of Trade . . . . . . . . . . . . . . . . . . . 295 13.3 Trade, Specialization and the World Income Distribution . . . . . . . . . . . 296 13.3.1 The Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 13.3.2 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 13.3.3 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303 13.4 Growth with Factor Price Equalization . . . . . . . . . . . . . . . . . . . . . 304

IV

Endogenous Technological Change

307

311

14 Expanding Variety Models

14.1 The Lab-Equipment Model of Growth with Product Varieties . . . . . . . . 312 viii

14.451: Introduction to Economic Growth 14.1.1 Demographics, Preferences and Technology . . . . . . . . . . . . . . . 312 14.1.2 Digression on Continuous Time Value Functions . . . . . . . . . . . . 314 14.1.3 Characterization of Equilibrium . . . . . . . . . . . . . . . . . . . . . 315 14.1.4 Deﬁnition of Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . 317 14.1.5 Steady State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317 14.1.6 Transitional Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 319 14.1.7 Pareto Optimal Allocations . . . . . . . . . . . . . . . . . . . . . . . 320 14.1.8 Policy in the Endogenous Technology Model . . . . . . . . . . . . . . 322 14.2 Growth with Knowledge Spillovers . . . . . . . . . . . . . . . . . . . . . . . 324 14.2.1 The Role of Competition Policy . . . . . . . . . . . . . . . . . . . . . 326 14.3 Growth without Scale Eﬀects . . . . . . . . . . . . . . . . . . . . . . . . . . 328 15 Models of Quality Competition 333

15.1 Baseline Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 15.2 Pareto Optimality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 16 Directed Technical Change 341

16.1 Basics and Deﬁnitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343 16.1.1 Deﬁnitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343 16.1.2 Basic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343 16.1.3 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350 16.2 Equilibrium Technology Bias: Some More General Results . . . . . . . . . . 354 16.3 Endogenous Labor-Augmenting Technological Change . . . . . . . . . . . . . 356 16.3.1 Demographics, Preferences and Technology . . . . . . . . . . . . . . . 356 16.3.2 Consumer and Firm Decisions . . . . . . . . . . . . . . . . . . . . . . 361 16.3.3 Asymptotic and Balanced Growth Paths . . . . . . . . . . . . . . . . 364 ix

.1. . . .14. . . .1 Thinking of Institutions and Growth . . . . . . . . . . . . . . . . . . . 416 18. . . . . . . . 405 18. . . . . . .3. . . . . . 390 18. . . . . . . . . 370 17 Recitation Material: Appropriate Technology 373 17.1 Diﬀerences in Capital-Labor Ratios (Atkinson-Stiglitz) . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 The Role of Human Capital (Acemoglu-Zilibotti) . 403 18.2. . . . .5 Factor Price Manipulation . . .2. . . . 374 17. . 392 18. . . . . .2. . . . . . . . . . . . . . . . . .3 Ineﬃcient Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 16. . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . 380 17.3.2 Implications . . . . . 380 18 Epilogue: Political Economy of Growth 383 18. . . . .2. . . . . . 408 18. . . 406 18.1 A Model . .451: Introduction to Economic Growth 16.2. . . . . . . . .8 Subgame Perfect Versus Markov Perfect Equilibria . . . .4 Revenue Extraction . . . . . . . . . .2.9 Lack of Commitment–Holdup . 413 18. . . . . .3. . . . . . . . . . . . . . . . . . . . . . . .2 Economic Equilibrium . . . .2. . . . .2.2. . . . . . . . . . . . . . . . . . . . . .1 Baseline Model . . . . . . . . . . . . .1. . . . . . . . .2 A Simple Model of Non-Growth Enhancing Institutions . . . 375 17. . . . . . . .1. . . . . . . . . 396 18. . . . . . . . .1 The Impact of Institutions . . . . . .4 The Balanced Growth Path . . . . . . . . . . . . . . . . . . . . . . . 375 17. . 400 18. . . . . . . . . . . .5 Transitional Dynamics . . . . . . . . . . 366 16. . . . . . . . .3 Institutions in Action . . . . . 384 18. . . . . . . . .6 Revenue Extraction and Factor Price Manipulation Combined . . .2 Modeling Institutional Diﬀerences . . . . . . .6 Policy Implications . . . . . . . . . 409 18. . . . . . .2. . . . .3 Calibration . . . . .7 Political Consolidation . . . . . . . . . . . . . 417 x . . . . . . . . 385 18. . . . . . .

.2. . .14. 427 18. .3. . . . 428 18. .11 Ineﬃcient Economic Institutions . . . . . . . .3. . . . . . . . .4 Institutional Change and Persistence . . . .2. . . 419 18. . 429 18. . . . . . . . .451: Introduction to Economic Growth 18. . . . .10 Technology Adoption and Holdup . . . . . . . . . . . . . . . . . . . .3. . 433 xi . . . . . .1 Dictatorship of the Middle Class . . . . . . . . . .3 Ineﬃciency of Political Institutions and Inappropriate Institutions . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 Democracy . . . . 422 18. . . .3 Modeling Political Institutions . . . . . . . . 431 18. . . . . . . . .

14.451: Introduction to Economic Growth xii .

Part I Introduction 1 .

.

The purpose is to both prepare us for the analysis of more modern models of economic growth with forward-looking behavior and explicit capital accumulation and technological progress.451: Introduction to Economic Growth We start with a quick look at the stylized facts of economic growth and the most basic model of growth. the Solow growth model. I will also discuss diﬀerences between proximate and fundamental causes of economic growth and development. and also give us a way of mapping the simplest model to the data.14. 3 .

451: Introduction to Economic Growth 4 .14.

income per capita is much lower in many other countries: $9000 in Mexico. $2500 in India. GDP per capita in the United States was $32500 (valued at 1995 $ prices). $4000 in China. in 2000. In contrast. $1000 in Nigeria. Mali (all ﬁgures adjusted for purchasing power parity). For example. and much much lower in some other sub-Saharan African countries such as Chad. 5 . The gap is larger when there is no PPP adjustment. Ethiopia.1 A Quick Look at the Facts There are very large diﬀerences in income per capita or output per worker across countries today. Countries at the top of the world income distribution are thirty times as rich as countries at the bottom in PPP adjusted dollars. The next ﬁgure shows a cross-sectional look at these income-level diﬀerences in the year 2000.Chapter 1 Stylized Facts of Economic Growth and Development 1.

there are striking diﬀerences in the quality of life. Understanding how some countries can be so rich while some others are so poor is one of the most important. perhaps the most important. challenges facing social science. 6 . living standards and health level of richer countries are appreciably higher than those with lower income per capita. standards of living and health. rich country with a less-developed one. it is even diﬃcult for us to imagine the burden of poverty at the levels experienced by countries in sub-Saharan Africa. High income levels reﬂect high standards of living. In fact. pollution increases and individual aspirations may also increase so that the same bundle of consumption may no longer make an individual as happy. It is true that together with economic growth. when one compares an advanced. These gaps represent big welfare diﬀerences. But at the end of the day. There is little doubt that the consumption level.451: Introduction to Economic Growth Should we care about cross-country income diﬀerences? The answer is a big yes.14.

Therefore.451: Introduction to Economic Growth How could a country be 30-times or so richer than another? The answer lies in diﬀerences in growth rates. Take two countries. A and B. we see tremendous diﬀerences in growth rates across countries. while country B grows at 2% per capita. In 200 years’ time country B will be more than 52 times richer than country A. so its income per capita remains constant. with the same level of income to start with. This is shown in the next picture for the postwar era: This picture shows how East Asian tigers have grown at much higher rates than the rest of the world over the past 40 years. the United States is considerably richer than Nigeria because it has grown steadily over an extended period of time. 7 . In fact. while a number of countries in sub-Saharan Africa and Central America have experienced negative growth. while Nigeria has not. even in the historically-brief postwar era.14. Imagine that country A has 0% growth per capita.

For another.48) + 1.06) ln y 1960 (1. it may be precisely the poor countries that are growing faster. despite some big growth successes and disasters. South Korea. these growth diﬀerences may be small relative to those necessary to cause large per capita income level diﬀerences. The next question is therefore when this growth gap opened up. with a slight tendency towards becoming more unequal.14. For one thing. The next ﬁgure shows this relationship diagrammatically. but the “world income distribution” has been more or less stable. For example. For instance. The answer is that much of the divergence happened during the 19th century and early 20th century.1) with R2 = 0. A regression of log income per worker in 1990 on log income per worker in 1960 gives the following relationship: ln y 1990 = 0. There are striking growth diﬀerences during the postwar era. Singapore and Taiwan were substantially poorer than the United States and Western Europe in 1960. the substantial growth diﬀerences in the postwar era do not mean that these growth diﬀerences are responsible for the current diﬀerences in income levels. countries that were rich in 1960 are very very likely to be rich today. 8 .78.56 (0. Hong Kong.451: Introduction to Economic Growth However.00 (0.

there is a small but notable increase in the dispersion of incomes.451: Introduction to Economic Growth If we look at output or income per worker.14. the overall shape of the world income distribution has been relatively stable in the postwar period. This is shown in the next ﬁgure which depicts the standard deviation of log income per capita in the world and the ratio of the income of the ﬁve richest to the ﬁve poorest countries in the world. 9 . Instead. There is certainly no narrowing of income gaps.

This is shown in the next ﬁgure: 10 .14. whereby some of the middle-income countries of the 1960s appear to have joined either the low-income or the high-income club. there is also a pattern of stratiﬁcation.451: Introduction to Economic Growth Moreover.

How do we capture conditional convergence? Consider a typical “Barro growth” regression: gt. they refer to whether the income gap between two countries increases or decreases irrespective of these countries’ characteristics.g. the determinants of steady state income and/or growth).2) where gt.. this 11 . we can look at the “conditional” distribution (e. the income gap between countries that share the same characteristics typically closes over time (though it does so quite slowly). Alternatively.14. yt−1 is per capita income at date t − 1 and X is a set of variables that the regression is conditioning on (in theory. When no covariates are included. 1992). Barro and Sala-i-Martin. Here the picture is one of conditional convergence: in the postwar period.t−1 is the annual growth rate between dates t − 1 and t.t−1 = β · ln yt−1 + Xt0−1 · α + εt (1.451: Introduction to Economic Growth The above statements refer to the “unconditional” distribution–that is.

was at most eight times as rich as the poorest country in the world. which is identical to (1. this is really identical to the regression equation (1.1). In fact. from 1870 to today.451: Introduction to Economic Growth regression leads to a positive or zero estimate of β . The income gap between countries was much smaller during the 19th century than today.S. there is divergence.t−1 ' ln yt − ln yt−1 . for example. This suggests that in 1870.1) equal to 1 implies that β ' 0. the U. If we look at a longer period. roughly at the rate of 2 percent a year. The estimate of (1 + β ) in (1.2) can be written as ln yt ' (1 + β ) ln yt−1 + εt . But when Xt−1 includes some human capital-related variables such as years of schooling or life expectancy. since gt. Pritchett illustrates this point using data from Angus Maddison and deriving an absolute lower bound on country incomes due to “subsistence”.14. thus no unconditional convergence. Here. This is illustrated in the next ﬁgure: 12 . so equation (1. He argues that $250 in terms of 1985 purchasing power parity is a practical lower bound below which the death rate would be extremely high. while it is over 30 times as rich today. without covariates. reiterating the absence of unconditional convergence as shown in the estimation of equation (1. the pattern is quite diﬀerent. β is estimated to be approximately -0.1) above. indicating that the income gap between countries that have the same human capital endowment has been typically narrowing over the postwar period.1) above. however.02. Therefore there has been signiﬁcant divergence over the past 130 years.

Johnson and Robinson (2002) show that in 1500. among the societies that were later to be colonized by European powers.S. among this set of countries there was a pattern of reversal. Therefore. Aztecs and Incas were much more urbanized and densely settled than the civilizations in North America. A variety of evidence shows that in 1500 the Mughals. the pattern may be one of reversal : Acemoglu. Today the U. Canada. and is more wide13 . those that were relatively prosperous are today relatively poor.451: Introduction to Economic Growth If we go even further back.. Ecuador or Peru. New Zealand and Australia.14. whereby those that were relatively prosperous in 1500 have become relatively poor today. How do we measure/proxy economic prosperity in 1500? It turns out that urbanization rates and population density are good proxies for prosperity during preindustrial periods (and urbanization rates are also good proxies even today). such as India. New Zealand and Australia are orders of magnitude richer than the countries now occupying the territories of the Mughal. Aztec and Inca Empires. The reversal is not conﬁned to this set of countries.

451: Introduction to Economic Growth spread among the former European colonies. PPP. PPP. This is shown in the next two ﬁgures. the ﬁrst using urbanization. the second population density as proxies for prosperity in 1500: USA 10 Log GDP per capita. 1995 CAN AUS SGP HKG NZL CHL 9 ARG VEN URY BRA DOM PRY JAM PHL IDN SLV BOL LKA HND NIC PAK VNM IND HTI LAO BGD MYS COL PAN CRI ECU BLZ PER GTM MEX TUN DZA MAR 8 GUY EGY 7 0 5 10 Urbanization in 1500 15 20 10 CAN AUS USA SGP HKG NZL CHL BRB BHS ARG Log GDP per capita. 1995 BWA BRA NAM VEN ZAF GAB MYS KNA PAN COL TTO CRI MEX 8 SUR GUY PRY SWZ CPV BOL TUN DZA JAM PHL IDN MAR SLV AGO LKA ZWE HND NIC CMR GIN CIV COG MRT SEN COM GHA IND SDN PAK LSO VNM GMB TGO CAF HTI LAO KEN BEN UGA NPL BGD ZAR BFA TCD MDG ZMB NGA NER ERI MLI BDI RWA MWI MOZ TZA SLE ETH LCA ECU GRD PER BLZ DOM DMA VCT GTM EGY 6 -5 0 Log Population Density in 1500 5 14 .14.

451: Introduction to Economic Growth When did this reversal take place? Consistent with the discussion from Pritchett’s paper above. industrialized and increased their GDP per capita. previously prosperous places continued to be somewhat more prosperous. Up to the late 18th century.14. The next two pictures give a sense of these processes: Timing of the Reversal Urbanization in excolonies with low and high urbanization in 1500 (averages weighted within each group by population in 1500) 25 20 15 10 5 0 800 1000 1200 1300 1400 1500 1600 1700 1750 1800 1850 1900 1920 low urbanization in 1500 excolonies high urbanization in 1500 excolonies 15 . It was the age of industrialization. when previously less-prosperous former colonies became rapidly urbanized. the 19th century. the evidence suggests that the reversal among the former European colonies took place during the 19th century as well.

14.2 Interpretation This discussion points to the following set of facts and questions that are central to an investigation of the determinants of long-run diﬀerences in income levels and growth: 1. The relative stability of the postwar income distribution has suggested to many economists that we should look for diﬀerences across countries leading to very large “permanent” diﬀerences in income.451: Introduction to Economic Growth Reversal. 2. UK in 1900 = 100 (from Bairoch) 400 350 300 250 200 150 100 50 0 1750 1800 US 1830 Australia 1860 Canada 1880 New Zealand 1900 Brazil 1913 Mexico India 1928 1953 1. but not necessarily large “permanent” diﬀerences in 16 . This immediately takes us to questions of why some countries grow (or have grown) while other countries have failed to grow and stagnated. Industrialization and Divergence Industrial Production Per Capita. The major pattern to be explained is why there are such large diﬀerences in income per capita and worker productivity across countries.

reﬂect technological or institutional changes that took place during the 19th and early 20th centuries. So this reasoning might have some merit. where countries that invest at diﬀerent rates grow at diﬀerent rates). the cross-country distribution of income across the world is relatively stable. countries have not settled into a stationary world income distribution. 3. the divergence from the 19th century to today suggests that we might want to look for a set of theories where the large diﬀerences in income per capita. 4.451: Introduction to Economic Growth growth rates in the recent decades. economists with this view argue that the ﬁnding of conditional convergence suggests the presence of transitional dynamics taking countries towards their “steady state” values as in the basic Solow and neoclassical models. The reversal (among the former European colonies) suggests that theories that empha17 .14. For example. 5. Equally puzzling is how the very large income diﬀerences we observe today can persist in this age of free-ﬂow of technology. some countries may have taken advantage of industrialization opportunities. we should expect significant divergence. Clearly. trade and ﬁnancial integration. This is based on the following reasoning: with substantially diﬀerent long-run growth rates (as in models of endogenous growth. We saw above that despite some widening between the top and the bottom. We therefore need theories which can shed light on why certain societies may fail to take advantage of better technologies. while other societies have failed to do so. or may have only started adopting technologies very late. Moreover. at least to some extent. we have seen that there there is still some notable (though perhaps not so large) divergence in the world income distribution. Furthermore. Nevertheless. It is important to understand why even in this age of free-ﬂow of technology some countries are growing faster than others.

18 .3 The Agenda In the rest of the class. persistence not a reversal).451: Introduction to Economic Growth size diﬀerences in (economic and perhaps political) institutions or social organization or more generally man-made factors as key determinants of economic performance may be more promising than theories emphasizing ﬁxed environmental factors such as geography or climate. and productivity growth.14. we will look at models that can help us understand the mechanics of economic growth. the approach will be two pronged: on the one hand. we should expect countries that were relatively rich 200 or 500 years ago to be also relatively rich today–i. Alternatively. This means understanding a variety of models that underpin the way economists think about the process of capital accumulation. technological progress. and why some countries are rich and some others are not.e. we want to understand what these models and others have to say about which key parameters or key economic processes are diﬀerent across countries and why. Only by understanding these mechanics can we have a framework for thinking about the causes of why some countries are growing and some others are not. More ambitiously. we may want to investigate whether and why certain characteristics that make countries richer at some point contribute to their relative poverty during other episodes. 1. on the other. Therefore. (With such environmental factors as the main determinants of income diﬀerences. we want to understand the mathematical structure of these models as well as possible. we may want to see what type of shocks could cause a “reversal” in the relative incomes of countries over long periods..

or simply as the Solow growth model after our own Bob Solow. 2.1 The Basic Model in Discrete Time The Production Structure We start with the simplest growth model. weeks.. you should know 19 . This is a closed economy. 1. with a unique ﬁnal good. We return to what this assumption of the representative consumer involves below. or years. you may want to assume that all households are identical.. who was awarded the Nobel prize for his contributions to growth theory. So far we do not need to take a position on this. they will not be optimizing. The economy is in discrete time running to inﬁnite horizon. As an aside. Time periods here can correspond to days. To ﬁx ideas. so that the economy admits a representative consumer. The economy is inhabited by a large number of households.1. sometimes referred to as the Solow-Swan model after two economists who developed versions of it. so that time is indexed by t = 0.1 2. and for now we are going to make relatively few assumptions on the households because in this baseline model.. .Chapter 2 The Solow Growth Model 2.

L (t) is total employment and A (t) is technology. L (t) . Thus the ﬁrm does not have to pay for it. A (t)] (2.451: Introduction to Economic Growth from basic general equilibrium theory that most economies do not admit a representative consumer. But much of macroeconomics (unfortunately) ignores this basic theorem. Much of the neoclassical growth theory is about understanding exactly how much individuals save and how capital accumulates. Both the capital stock and technology are taken to be single indices. Heterogeneity of preferences. in fact the celebrated Debreu-Mantel-Sonnenschein theorem states that we can say relatively little about the preferences of a consumer obtained by aggregating a number of well-behaved neoclassical consumers. and works with representative consumers. abilities and income are in fact quite important to understand the process of economic growth. The key assumption of the Solow model will be that each household saves an exogenous fraction s of their income. In the basic model this is taken as exogenous. non-rival good. and at some level. The other key agents in the economy are ﬁrms. Here I will adopt the same defense and for much of this course I will limit myself to models with representative consumers. the important assumption is that technology is free. they are treated as “black boxes”–we will later discuss how such models can be extended to think of multiple types of technologies and capital goods. The capital stock here denotes the quantity of “machines” used in production. In many situations this can be justiﬁed on the basis of parsimony.14. 20 . but many of these topics are beyond the scope of this class.1) where Y (t) is the total amount of production of the ﬁnal good. it is publicly available as a non-excludable. For now. Let us assume that the economy also admits an aggregate production function for the unique ﬁnal good Y (t) = F [K (t) . K (t) is the capital stock.

It speciﬁes that marginal products are positive (thus ruling out some production functions). L. and to be strictly concave in K and L. L.451: Introduction to Economic Growth As an aside. The other important assumption is that of constant returns to scale. L. you might want to note that some authors use xt or Kt when working with discrete time and reserve the notation x (t) or K (t) for continuous time. i. A) ≡ FL (K. Recall that F exhibits constant returns to scale in K and L if it is linearly homogeneous (homogeneous of degree 1) in these two variables. Positive Marginal Products. A) > 0.14. FKK < 0 and FLL < 0.e. for simplicity. we have: Assumption 1 (Continuity. A) ≡ ∂K 2 FK (K. I will drop time dependence when this causes no confusion. L. In particular. Since I will go back and forth between continuous time and discrete time. A) FLL (K. Concavity and Constant Returns to Scale) F is twice continuously diﬀerentiable in K and L. ∂K ∂ 2 F (K. More speciﬁcally: 21 . F exhibits constant returns to scale in K and L. but include it when there is any chance of such confusion. A) ≡ ∂F (K. A) < 0. ∂L ∂ 2 F (K. We will see below that the degree of diminishing returns to capital will play a very important role in many of the results of the basic growth model. The production function F : R3 → R is. L. L. Diﬀerentiability. All of the components of Assumption 1 are important. Throughout. except when discussing dynamic programming where the subscripts are the usual notation. FKK (K. A) ≡ < 0. ∂L2 Moreover. L. A) > 0.. but more importantly that there are diminishing returns both to capital and labor. and satisﬁes ∂F (K. I use the latter notation all throughout. L. assumed to be twice continuously diﬀerentiable and increasing in all of its arguments.

z ) x + gy (λx.2) with respect to x: λgx (λx. z ) = λm g (x. y. 22 (2. y. Diﬀerentiate both sides of equation (2. λy. λy.2) . Proof. z ) for all λ ∈ R+ and z ∈ RK . z ) = gx (λx. diﬀerentiate both sides of equation (2.2) with respect to λ. y. The function g (x. z ) = λm g (x. Linearly homogeneous (constant returns to scale) production functions are particularly useful because of the following theorem: Theorem 1 (Euler’s theorem) Suppose that g : RK +2 → R is continuously diﬀerentiable in x ∈ R and y ∈ R. z ) y for any λ. z ) are themselves homogeneous of degree m − 1 in x and y. which gives mλm−1 g (x. Setting λ = 1 yields the ﬁrst result. λy. Then mg (x. y. y. We have that g is continuously diﬀerentiable and g (λx. y. λy. y. z ) x + gy (x. z ) y for all x ∈ R. y ∈ R and z ∈ RK . z ) = gx (x. z ) . z ) = λm gx (λx. z ) and gy (x. λy. with partial derivatives denoted by gx and gy and is homogeneous of degree m in x and y .14. y. y.451: Introduction to Economic Growth Deﬁnition 1 Let z ∈ RK for some K ≥ 1. z ) . z ) is homogeneous of degree m in x ∈ R and y ∈ R if and only if g (λx. To obtain the second result. gx (x. Moreover. λy.

and we take their initial holdings of capital. For now how this initial capital stock is distributed among the 23 . If there is population growth. we will work with aggregate production functions as a representation of underlying production structure of the economy. that the assumption of an aggregate production function could be quite restrictive.451: Introduction to Economic Growth Dividing both sides by λ establishes the desired result. we will further assume that product markets are also competitive. for example. In particular. The households also own the capital stock of the economy. 2. K (0). when the economy consists of a large number of ﬁrms all having access to the same constant returns to scale production function.14. there is no diﬀerence between assuming an aggregate production function or working with a large number of ﬁrms competing for factors of production. Throughout this course we are going to assume that all factor markets are competitive. households own all of the labor. Notice. or new households being born. so ours will be a prototypical competitive general equilibrium model. In that case.2 Endowments Let us imagine that all factors of production are owned by households. In particular it rules out heterogeneity of productivity among ﬁrms.1. Until we come to models of endogenous technological change. as noted above. This would be the case. as given (as part of the description of the environment). this can be thought of as existing households becoming larger. however. For our purposes here this does not matter. and it also creates problems when there are non-constant returns to scale (can you see what would go wrong with decreasing returns to scale?). for example F above. which they supply inelastically. Moreover. and this will determine the initial condition of the dynamical system we will be analyzing.

but once it is in place. it becomes ﬁxed and it cannot be turned back into consumption goods. Recall that we always have to choose a numeraire. which they can eat or rent to ﬁrms to enable them to produce further corn. we want to think of capital as corresponding to machines. there is a natural choice of numeraire in this economy which is to normalize the price of the ﬁnal good in each period to 1. There are many diﬀerent ways of conceptualizing capital. at which capital-labor ratio it can be used. where diﬀerent 24 . But for now let us make the rather heroic assumption that capital is essentially the same as the ﬁnal good..These types of models are sometimes referred to as “putty-putty. But this is without loss of any generality. So the economy consists of “corn”. Loosely speaking. and some of them are beyond the scope of this course.” since capital is totally malleable both before and after it is designated as capital.. The more important point is that the households will rent their capital to ﬁrms. but here we are making a normalization in each period.14. Then K (0) is the amount of corn that individual households have at the beginning of period t = 0... This discussion should already alert you to a central fact: you should think of all of the models we are going to be talking about as general equilibrium economies. [. cannot be changed. Then in competitive markets a representative ﬁrm is solving the problem of proﬁt maximizing.451: Introduction to Economic Growth households is not important. and certain features of it. because the interest-rate between periods will play the role of relative prices. Another important set of issues involves how to think of “capital”. and it can use some amount of this corn as input into producing further corn. for example.] Given this structure. Alternatives include “putty-clay” models where corn can be used as capital. Let the rental price of capital be denoted by R (t) and the rental price of labor by w (t).

3) A couple of features are worth noting: 25 . A(t)] − w (t) L (t) − R (t) K (t) . This implies that the individual has given up one unit of commodity dated t − 1 for r (t) units of commodity dated t. the next important assumption is that capital depreciates. which implies L(t). as well as. combined with the normalization of the price of the ﬁnal goods to 1.14. For now it is treated as a black box. The importance of this for a household is that. We assume that this depreciation takes an exponential form.451: Introduction to Economic Growth commodities correspond to the same good at diﬀerent dates. the household will receive R (t) units of good as the rental price. L(t). in almost all of the models that we will study in this course. This raises a number of special issues in the theory of general equilibrium which we will touch on as we go along. but will get back only 1 − δ units of the capital. (2. Recall from basic general equilibrium theory that the same good at diﬀerent dates (or in diﬀerent states or in diﬀerent localities) is a diﬀerent commodity. there will be an inﬁnite number of commodities (because time runs to inﬁnity). This ﬁrm will maximize proﬁts. In the latter case.K (t) max F [K (t). the replacement of old machines by new machines. Now let us consider the problem of a representative ﬁrm. it implies that the rate of return faced by the household will be r (t) = R (t) + 1 − δ. since the rest has depreciated. This depreciation in general stands for the wear and tear of the machinery. in more realistic models. Now returning to our treatment of the basic model. Recall that every unit of capital can be eaten now or rented to ﬁrms. so that out of 1 unit of capital this period. This means that capital depreciates (exponentially) at the rate δ . Therefore. only 1 − δ is left for next period.

since F is concave (though not necessarily strictly so).4) An immediate corollary of Theorem 1 combined with competitive factor markets is: Proposition 1 In equilibrium. (2. since the price of the ﬁnal good has been normalized to 1. ﬁrms make no proﬁts. This is without loss of any generality given the representative ﬁrm (or the existence of aggregate production function). 3.5) (2. All we need to know is that ﬁrms are proﬁt-maximizing entities. L(t). This way of writing the problem already imposes competitive factor markets.e. as given. The ﬁrst-order necessary conditions of the ﬁrm’s problem (combined with diﬀerentiability of F ) imply that the competitive factor returns are equal to their marginal products: w (t) = FL [K (t). so. the ownership of ﬁrms does not need to be speciﬁed. This result is convenient. A(t)]. I set up the problem in terms of aggregate variables. since the ﬁrm is taking the prices of labor and capital. 26 . and R (t) = FK [K (t).. This follows immediately from Theorem 1 for the case of m = 1. This is a concave problem. There is nothing multiplying the F term.14.451: Introduction to Economic Growth 1. i. Proof. 2. w (t) and R (t) . since it implies that ﬁrms make no proﬁts. A(t)]. in contrast to the basic general equilibrium theory. Y (t) = w (t) L (t) + R (t) K (t) . 4. constant returns to scale. L(t). and in particular.

(2. Here the behavioral rule of the constant savings rate simpliﬁes the structure of equilibrium considerably.451: Introduction to Economic Growth In addition to these standard assumptions on the production function. in growth theory we often impose the following additional boundary conditions. From national income accounting for a closed economy.1). A) = ∞ and lim FL (K. L. 2.7). we take G (t) ≡ 0. For now. Assumption 2 (Inada conditions) F satisﬁes the Inada conditions lim FK (K.14. L. A) = 0 for all L > 0 and all A K →∞ L→∞ L→0 K →0 lim FL (K. Therefore. (2.3 Fundamental Law of Motion of the Solow Model Finally. L.1. Recall that K depreciates exponentially at the rate δ. so that national income is divided between consumption and investment. referred to as Inada conditions. feasible dynamic allocations in this economy would have to satisfy K (t + 1) ≤ F [K (t) .6) where I (t) is investment at time t. using (2. we can write the law of motion of the capital stock of the economy. we have Y (t) = C (t) + I (t) + G (t) . A) = 0 for all K > 0 and all A. so that the law of motion of the capital stock is given by K (t + 1) = (1 − δ ) K (t) + I (t) . It is important that the constant savings rate is a 27 . L (t) .7) where C (t) is consumption and G (t) is government spending. The question is to determine the equilibrium dynamic allocation among the set of feasible dynamic allocations. A (t)] + (1 − δ ) K (t) − C (t) . (2.6) and (2. L. A) = ∞ and lim FK (K.

1).8). First note that given G (t) ≡ 0 (and the closed economy assumption). i.14..451: Introduction to Economic Growth behavioral rule. their behavior is captured by a behavioral rule. it is not derived from a well-deﬁned utility function. (2. the equilibrium is essentially described by this equation together with laws of motion for L (t) and A (t).8) Thus combining (2.1. Households do not optimize when it comes to their savings/consumption decisions. 2. This means that any welfare comparisons based on the Solow model have to be taken with a grain of salt. L (t) . we have the key dynamic (diﬀerence) equation of the Solow growth model: K (t + 1) = sF [K (t) . But ﬁrms maximize and 28 .10) In the Solow growth model. aggregate investment is equal to savings. Now recall that individuals are assumed to save a constant fraction s of their income. A (t)] + (1 − δ ) K (t) .6) and (2.4 Deﬁnition of Equilibrium The Solow model is a mixture of an old-style Keynesian model and a modern dynamic macroeconomic model. S (t) = sY (t) . so that they consume the remaining 1 − s fraction of their income: C (t) = (1 − s) Y (t) (2. Instead.9) (2. We have no idea what the utility function of the individuals are. (2. S (t) = I (t) = Y (t) − C (t) .e.

an equilibrium path is a sequence of capital stocks. y (t) ≡ (2. is given by ¸ K (t) . w (t) .1).9).11) . Thus it is useful to start deﬁning equilibria in the way that is customary in modern dynamic macro models.4) and (2.14. so that A (t) = A. consumption levels. 2. let us make some further assumptions: 1. Deﬁnition 2 In the basic Solow model for a given sequence of {L (t) . so that L (t) = L. R (t)}∞ t=0 such that K (t) satisﬁes (2. wages and rental rates {K (t) .1. Y (t) . let us deﬁne the capital-labor ratio of the economy as k (t) ≡ Y (t) /L. output levels. For now.5 Equilibrium Without Population Growth and Technological Progress We can make more progress by exploiting the constant returns to scale nature of the production function. To do this. 1. 2. Let us also assume that there is no technological progress.10).451: Introduction to Economic Growth factor markets clear. L Then using the constant returns to scale assumption we have that income per capita. Y (t) is given by (2. A y (t) = F L ≡ f (k (t)) . A (t)}∞ t=0 and an initial capital stock K (0). ∙ 29 K (t) . We will relax these assumptions later. Let us assume that population is constant and individuals supply labor inelastically. and w (t) and R (t) are given by (2. C (t) . C (t) is given by (2.5).

and the other equilibrium quantities can be obtained from the capital-labor ratio k (t). and typically the economy will tend to this steady state equilibrium over time (but often never reach it in ﬁnite time). (2. Deﬁnition 3 A steady-state equilibrium without technological progress and population growth is an equilibrium path in which k (t) = k∗ for all t. zero factor prices are possible over certain ranges). which imposed that the ﬁrst derivatives of F with respect to capital and labor are always positive (with more general production functions. Most of the models we will analyze in this course will admit a steady state equilibrium. income per capita is simply a function of the capital-labor ratio.10).10) by L and obtain a simpler diﬀerence equation k (t + 1) = sf (k (t)) + (1 − δ) k (t) . we can also deﬁne a steady-state equilibrium for this model without technological progress and population growth. in the steady-state equilibrium the capital-labor ratio remains constant. At this point. in that it describes the equilibrium behavior of the key object of the model. (2. This is also the case for this simple model. Given Theorem 1. with constant returns to scale. the capital-labor ratio. we also have R (t) = f 0 (k (t)) > 0 and w (t) = f (k (t)) − k (t) f 0 (k (t)) > 0. Given this.14.12) The fact that both of these factor prices are positive follows from Assumption 1. it also can be referred to as the equilibrium diﬀerence equation of the Solow model.13) Since this diﬀerence equation is derived from (2. In other words. 30 . we can divide both sides of (2.451: Introduction to Economic Growth In other words.

14) An alternative visual representation of the steady state is to view it as the intersection between a ray through the origin with slope δ (representing the function δk ) and the function sf (k ). s k∗ (2. This establishes: Proposition 2 Consider the basic Solow growth model and suppose that Assumptions 1 and 2 hold.13).14. (2. The intersection of the right hand side with the 45◦ line gives the steady-state value of the capital-labor ratio k∗ .451: Introduction to Economic Growth This can be seen by plotting the diﬀerence equation which governs the equilibrium behavior of this economy. which is also useful in seeing the level of consumption and investment in a single ﬁgure. The next ﬁgure shows this picture. Then there exists a unique steady state where the capital-labor ratio is equal to 31 . which satisﬁes δ f (k∗ ) = .

f (k) /k is continuous from Assumption 1.13). limk→0 f (k) /k = ∞ and limk→∞ f (k) /k = 0.14). with greater values corresponding to greater productivity of factors. Equation (2. so there exists k∗ such that (2. per capita output is given by y ∗ = f (k∗ ) and per capita consumption is given by c∗ = (1 − s) f (k∗ ) . = − ∂k k2 k (2.15) The preceding argument establishes that (2. diﬀerentiate f (k) /k with respect to k. but 32 . ∞) and is given by (2.15) and (2. So far the model is very parsimonious. Since f (k) /k is everywhere decreasing.451: Introduction to Economic Growth k∗ ∈ (0.17) where the last equality uses (2. But what we are most interested in is to understand how cross-country diﬀerences in certain parameters translate into diﬀerences in growth rates or income levels. let us generalize the production function in one simple way..14.12). a zero of the diﬀerence equation (2. To establish existence. But before doing so. Proof.e. i. and assume that ˜ (k) f (k) = af so that a is a shift parameter.16) (2. This will be done in the next proposition.14) is satisﬁed. Moreover. which gives ∂ [f (k) /k] f 0 (k) k − f (k) w = < 0. and does not have many parameters.16) then follow by deﬁnition. To see uniqueness.14). there can only exist a unique value k∗ that satisﬁes (2.14) is a steady state. (2. This type of productivity is referred to as “Hicks-neutral” as we will see below. note that from Assumption 2.

∂k∗ δ (k∗ )2 = 2 ∗ >0 ∂s sw where w∗ = f (k∗ ) − k∗ f 0 (k∗ ) > 0. s. so does f ˜ (k). δ ) and the steady-state level of output by y ∗ (a. s. will tend to have lower capital-labor ratios and will be poorer. δ ) ∂k∗ (a. For example. s. s. The same comparative statics with respect to a and δ immediately apply to c∗ as well. Since f (k) satisﬁes the regularity conditions imposed above. s and δ . 33 . Denote the steadyProposition 3 Suppose Assumptions 1 and 2 hold and f (k) = af state level of the capital-labor ratio by k∗ (a. δ ) > 0. Then we have ∂k∗ (a. δ ) when the underlying parameters are given by a. Therefore. δ ) ∂y ∗ (a. ∂a ∂s ∂δ Proof. δ ) ∂y ∗ (a. s. ∗ k as which holds for an open set of values of k ∗ . > 0 and <0 ∂a ∂s ∂δ ∂y ∗ (a. > 0 and < 0. Now apply the implicit function theorem to obtain the results. and start giving us a sense of some important determinants of the capital-labor ratios and income levels across countries. δ ) ∂k∗ (a. s. s. countries with higher savings rates and better technologies will have higher capital-labor ratios and will be richer. it is straightforward to see that c∗ will not be monotonic in the savings rate (think. The proof follows immediately by writing ˜ (k ∗ ) f δ = . The other results follow similarly. s. countries. All of the results in Proposition 3 are intuitive.451: Introduction to Economic Growth for now it is just a convenient way of looking at the impact of productivity diﬀerences across ˜ (k). However. δ ) > 0.14. Those with greater (technological) depreciation.

∂s ∂s Since from Proposition 3 we have ∂k∗ /∂s > 0. ∂c∗ (s) /∂s < 0 for all s > sgold . consumption can only be maximized when f 0 (k∗ (s)) = δ. we have ∂k∗ /∂s > 0 and moreover. we have established: Proposition 4 In the basic Solow growth model. with the corresponding steady state capital level kgold such that In other words. only sgold satisﬁes f 0 (k∗ (s)) = δ and gives the unique global maximum of consumption per capita. = f (k ∗ (s)) − δk∗ (s) Now diﬀerentiating this expression with respect to s (again using the implicit function theorem). Consequently. This is shown in the next ﬁgure with 34 ¡ ∗ ¢ f 0 kgold = δ. . we have ∂c∗ (s) ∂k∗ = [f 0 (k∗ (s)) − δ ] .14. it can be veriﬁed that ∂ 2 c∗ (s) /∂s2 < 0. That f 0 (k∗ (s)) = δ is also the global maximum follows from the following observations: ∀ s ∈ [0. let us suppress the other parameters and write c∗ (s) = (1 − s) f (k ∗ (s)) . so f 0 (k∗ (s)) = δ is indeed a local maximum. f 0 (k∗ (s)) − δ > 0 by the concavity of f . and by the converse argument. when s < sgold . so ∂c∗ (s) /∂s > 0 for all s < sgold . the highest level of consumption is reached ∗ for sgold . 1].451: Introduction to Economic Growth for example. To obtain the steady state relationship between c∗ and s. there exists a unique savings rate and the corresponding capital-labor ratio which will maximize steady-state consumption. Therefore. The relationship between consumption and the savings rate takes the form plotted in the next ﬁgure. of the case where s = 1!). when f 0 (k∗ (s)) = δ . Moreover.

2.451: Introduction to Economic Growth the consumption-maximizing savings rate denoted by sgold and the corresponding consumption per capita by cgold : Below this savings rate. individuals are investing too much and not consuming enough. This is the essence of what people refer to as dynamic ineﬃciency. however. and above this rate. the reason why such dynamic ineﬃciency will not arise once we endogenize consumption-saving decisions of individuals will be apparent to many of you already.14. so statements about “ineﬃciency” have to be considered with caution and skepticism. i. the society has too low a capital-labor ratio to maximize consumption.1. 35 . In fact. which we will encounter in greater detail in models of overlapping generations.e.. the capital-labor ratio is too high. recall that there is no explicit utility function here. that an equilibrium path does not refer simply to the steady state but to the entire path of capital stock.6 Transitional Dynamics in the Solow Model Proposition 2 establishes a unique steady state equilibrium. However. Recall.

at time t = 0.451: Introduction to Economic Growth output.13).13) starting from an arbitrary capital-labor ratio. Of special interest is the answer to the question of whether the economy will tend to this steady state starting from such an arbitrary capital-labor ratio. the economy starts with k (0) = K (0) /L as its initial value and then follows the law of motion given by the diﬀerence equation (2. since. k (0). Therefore. x (t + 1) = F (x (t)) . To determine what this equilibrium path looks like we need to study the “transitional dynamics” of the equilibrium diﬀerence equation (2. the total amount of capital at the beginning of the economy. the supply of labor L is ﬁxed. consumption and factor prices. is taken as a state variable. Before doing this.18) where x (t) ∈ Rn and F : Rn → Rn . i. (2.e.13) will take us to the unique steady state. x∗ = F (x∗ ). we ∗ any solution {x (t)}∞ t=0 . recall some deﬁnitions and key results from the theory of dynamical systems.18) with x (0) ∈ B (x ). for Theorem 2 Consider the following linear diﬀerence equation system x (t + 1) = Ax (t) 36 (2. and how it will behave along the transition path. K (0). Deﬁnition 4 An equilibrium point x∗ is (locally) asymptotically stable if there exists an ∗ open set B (x∗ ) 3 x∗ such that for any solution {x (t)}∞ t=0 to (2. Let x∗ be a zero (equilibrium) of this system. Thus the question is whether the diﬀerence equation (2. have x (t) → x∗ . x∗ is globally asymptotically stable if for all x (0) ∈ Rn . It is important to consider an arbitrary capital-labor ratio. Moreover.. we have x (t) → x . while for now.14. which means a ﬁxed point of the mapping F (·). as noted above.19) . Consider the nonlinear system of autonomous diﬀerence equations.

e. John Wiley & Sons. B (x∗ ) ⊂ Rn such that starting from any x (0) ∈ B (x∗ ). David Luenberger Introduction to Dynamic Systems: Theory Models and Applications. 1979. Then the diﬀerence equation (2. but the following is a standard local stability result. i. Suppose that all of the eigenvalues of A are strictly inside the unit circle (i. in the sense that there exists an open neighborhood of x∗ . the unique ∗ ∗ solution {x (t)}∞ t=0 satisﬁes x (t) → x where x is the steady state (zero) of the diﬀerence equation given by Ax∗ = x∗ . the absolute value of the real parts of the eigenvalues is strictly less than 1). F (x∗ ) = x∗ . 1994. Theorem 3 Consider the following nonlinear autonomous system x (t + 1) = F [x (t)] (2. in the sense that starting from any x (0) ∈ Rn .19) is globally asymptotically stable.14. much less can be said about nonlinear systems. where x (t) ∈ Rn for all t and A is an n × n matrix. and a version of it for diﬀerential equations is in Carl Simon and Lawrence Bloom Mathematics for Economists. Norton...18). Then the diﬀerence equation (2. for example. The proof of this theorem can be found in any textbook on dynamical systems. 37 . we have x (t) → x∗ . Unfortunately.20) where F :Rn → Rn and suppose that F is continuously diﬀerentiable. Deﬁne A =∇F (x∗ ) . and suppose that all of the eigenvalues of A are strictly inside the unit circle.20) is locally asymptotically stable.451: Introduction to Economic Growth with initial value x (0). Let x∗ be a zero of this system. Next let us return to be the nonlinear autonomous system (2.e. with initial value x (0).

1). and thus [sf 0 (k∗ ) + (1 − δ )] ∈ (0.451: Introduction to Economic Growth Therefore.21) around k ∗ . From (2.13) is asymptotically stable. Now recall that f (·) is concave from Assumption 1 and satisﬁes f (0) = 0 from Assumption 2. the nonlinear diﬀerence equation x (t + 1) = g (x (t)) is locally asymptotically stable if |g 0 (x∗ )| < 1.13): Proposition 5 Suppose that Assumptions 1 and 2 hold.14. we can have local stability results. Then. 38 .22) where the second line uses the fact that f (0) = 0.14). Moreover. diﬀerentiable at x∗ where g (x∗ ) = x∗ . k (t) → k∗ . Since from (2. (2. Proof. let g : R → R be a continuous function. for nonlinear systems. (2.21) with a unique zero at k∗ . and starting from any k (0) > 0. we have k (t + 1) ' [sf 0 (k∗ ) + (1 − δ )] (k(t) − k∗ ). (2. For any strictly concave function.13). establishing local asymptotic stability for the Solow model from Corollary 1. Now linearizing (2. An immediate corollary of these results is: Corollary 1 Let x (t) ∈ R. Now let us apply this result to (2. then the linear diﬀerence equation x (t + 1) = ax (t) + b is asymptotically stable (in the sense that x (t) → x∗ = b/ (1 − a)) if |a| < 1. we have k (t + 1) = sf (k (t)) + (1 − δ) k (t) .22) implies that δ = sf 0 (k∗ ) /k∗ > f 0 (k∗ ). δk∗ = sf (k ∗ ). we have that f (k) > f (0) + kf 0 (k) = kf 0 (k ) . then the equilibrium of the Solow growth model described by the diﬀerence equation (2.

then {w (t)}∞ t=0 is ∗ an increasing sequence and {R (t)}∞ t=0 is a decreasing sequence. Conversely. the opposite results apply. If k (0) > k .451: Introduction to Economic Growth Moreover. if the economy starts with too little capital relative to its labor supply. there will be capital deepening (capital accumulation relative to labor). and in the process the wage rate will decline and the rate of return to capital will increase.14. The next ﬁgure shows this process diagrammatically. and k (0) < k ∗ . and the wage rate will increase. Consequently. if it starts with too much capital. This stability result is easier to see diagrammatically. we have k (t + 1) − k (t) > 0 and for all k (0) > k ∗ . thus must be globally stable. (2. we have k (t + 1) − k (t) < 0. it will decumulate capital. the solution to (2. and as a result the marginal product of capital will fall given the diminishing returns to capital feature embedded in Assumption 1. Intuitively.21). The following corollary is then immediate: Corollary 2 Suppose that Assumptions 1 and 2 hold. ∗ {k (t)}∞ t=0 always approaches k . which is shown in the next ﬁgure.21) also implies that for all k ∗ > k (0) > 0. emphasizing that the trade-oﬀ is between the replacement of the capital stock per eﬀective labor due to depreciation (and perhaps population growth and technological change) and the capital to eﬀective labor ratio: 39 .

it has no growth however. The Solow model typically incorporates economic growth by allowing technological change.451: Introduction to Economic Growth Therefore. the Solow growth model has a number of nice properties. asymptotic stability. 40 . The steady state is the point at which there is no growth in the capital-labor ratio. it is useful to look at the mapping between discrete time and continuous time. Before doing this. unique steady state. no more capital deepening. and no growth in income per capita. however. So far. and simple and intuitive comparative statics.14.

Moreover. months or years.2 2. Now divide both sides of this equation by ∆t. you should also convince yourself that this approximation could in fact be quite bad if you take a very nonlinear function g. and take limits 41 . for any ∆t ∈ [0. This suggests that perhaps it may be more convenient to look at dynamics by making the time unit as small as possible.2. (2. 1]. which should not be too bad if the distance between t and t + 1 is not very large. x (t + 1)] (however.e. In-between it is a linear approximation. weeks.. the absolute growth in x is given by g (x (t)). Let us start with a simple diﬀerence equation x (t + 1) − x (t) = g (x (t)) .451: Introduction to Economic Growth 2. the time unit is not important.23) This equation states that between time t and t + 1. When ∆t = 0.1 The Solow Model in Continuous Time From Diﬀerence to Diﬀerential Equations Recall from the discussion above that the time periods could refer to days. The continuous time setup in general has a number of advantages. for which the behavior changes signiﬁcantly between x (t) and x (t + 1)).14. it gives (2. In some sense. For us. especially in the presence of uncertainty. this equation is just an identity. by going to continuous time. they are useful particularly because a lot of growth theory is cast in continuous time. since some pathological results of discrete time disappear in continuous time (see Problem Set 1). When ∆t = 1. so that g (x) ' g (x (t)) for all x ∈ [x (t) . i. continuous time models have more ﬂexibility both in doing dynamics and for providing explicit form solutions. Let us now consider the following approximation x (t + ∆t) − x (t) ' ∆t · g (x (t)) .23).

23) for the case in which the distance between t and t + 1 is small. i.2.14. L (t) = exp (nt) L (0) . population growth plays an important role. Recall that here x ˙ (t) denotes the time derivative ∂x (t) /∂t. w (t) is the ﬂow of wages that the worker receives for an instant etc.24) The purpose of doing so is that in many of the classical analyses of economic growth. while consumption is given by (2. so it is useful to see how it aﬀects things here. let us now introduce population growth into this model. which will be done below. Also..4) and (2.. ∆t→0 ∆t lim as a diﬀerential equation representing the same dynamics as the diﬀerence equation (2. We are not introducing technological progress yet.e. Savings are again given by S (t) = sY (t) . 2.2 The Fundamental Equation of the Solow Model in Continuous Time We can now repeat all of the analysis so far using the continuous time representation. and assume that the labor force L (t) grows proportionally. (2.). so we continue to have (2.5) as the factor prices. but now these refer to instantaneous rental rates (i.451: Introduction to Economic Growth to obtain x (t + ∆t) − x (t) =x ˙ (t) ' g (x (t)) .9) above.e. 42 . Nothing has changed on the production side.

L (t) . from the limiting argument in the previous subsection. we obtain the fundamental law of motion of the Solow model in continuous time for the capital-labor ratio as ˙ (t) = sf (k (t)) − (n + δ ) k (t) . K Now using the deﬁnition of k (t) as the capital-labor ratio and the constant returns to scale properties of the production function. we will refer to the steady-state equilibrium capital-labor ratio as k∗ . C (t) .5). labor. A(t)] − δK (t) . output levels. L (t) . consumption levels. k Therefore we have: Deﬁnition 5 In the basic Solow model in continuous time with population growth at the rate n. w (t) .9). k (t) K (t) The law of motion of the capital stock. C (t) is given by (2.25).24). As before.4) and (2. L (t) satisﬁes (2.14. no technological progress and an initial capital stock K (0). Y (t) . and w (t) and R (t) are given by (2. 43 (2.25) K (t) . L (t) . As before.1). R (t)]∞ t=0 such that K (t) satisﬁes (2. a steady-state equilibrium involves k (t) remaining constant.451: Introduction to Economic Growth Recall that k (t) ≡ which implies that ˙ (t) ˙ (t) K k = − n. Y (t) is given by (2. an equilibrium path is a sequence of capital stocks. is given by: ˙ (t) = sF [K (t) . wages and rental rates [K (t) .

per capita output is given 44 . which is given by a slight modiﬁcation of (2.26). ∞) and is given by (2.451: Introduction to Economic Growth It is easy to verify that the equilibrium diﬀerential equation (2. going from discrete to continuous time has not changed any of the basic economic features of the model.14.25) has a unique zero at k∗ . k∗ s (2. and again the steady state can be plotted in the familiar ﬁgure used above (now with the population growth rate featuring in there as well): We immediately obtain: Proposition 6 Consider the basic Solow growth model in continuous time and suppose that Assumptions 1 and 2 hold.14) above to incorporate population growth: n+δ f (k∗ ) = . Then there exists a unique steady state equilibrium where the capital-labor ratio is equal to k∗ ∈ (0.26) In other words.

we simply need to remember the equivalents of the above theorems for diﬀerential equations. which only accumulates slowly. n) ∂y ∗ (a. δ. δ. s. A higher population growth rate means there is more labor to use the existing amount of capital. n) ∂y ∗ (a. δ. Then we have ∂k∗ (a. n) when the underlying parameters are given by a. s. δ. δ. s. and consequently the equilibrium capital-labor ratio ends up lower. > 0. n) > 0. n) and the steady-state level of output by y ∗ (a. n) ∂k∗ (a. Moreover. also reduces the capital-labor ratio and income per capita. s. s. s. s. s. s. ∂a ∂s ∂δ ∂n The new result relative to the earlier comparative static proposition is that now a higher population growth rate. δ. n) ∂y ∗ (a. Denote the steadyProposition 7 Suppose Assumptions 1 and 2 hold and f (k) = af state equilibrium level of the capital-labor ratio by k ∗ (a. To do this in detail. s and δ . n) > 0. This result implies that countries with higher population growth rates will have lower incomes per person (or per worker). In particular we have: 45 . s. and < 0. > 0. The reason for this is simple. δ.14. n) ∂k∗ (a.451: Introduction to Economic Growth by y ∗ = f (k∗ ) and per capita consumption is given by c∗ = (1 − s) f (k∗ ) . again let ˜ (k) . n. δ. The stability analysis is also unchanged. f (k) = af Then we have ˜ (k). and <0 ∂a ∂s ∂δ ∂n ∂y ∗ (a. n) ∂k∗ (a. δ. δ.

Deﬁne A =∇F (x∗ ) . Corollary 3 Let x (t) ∈ R.14. with continuous time.28) where F : Rn → Rn and suppose that F is continuously diﬀerentiable. where x (t) ∈ Rn for all t and A is an n × n matrix. and suppose that all of the eigenvalues of A have negative real parts. i. Suppose that all of the eigenvalues of A have negative real parts. the nonlinear diﬀerential equation x g (x (t)) is a locally asymptotically stable if g0 (x∗ ) < 0.451: Introduction to Economic Growth Theorem 4 Consider the following linear diﬀerential equation system x ˙ (t) = Ax (t) (2. Then. Let x∗ be a zero of this system. x (t) → x∗ where x∗ is the steady state (zero) of the system given by Ax∗ = 0.e. in the sense that there exists an open neighborhood of x∗ . let g : R → R be continuous and ˙ (t) = diﬀerentiable at x∗ where g (x∗ ) = 0. we have x (t) → x∗ . with initial value x (0).27) with initial value x (0). Theorem 5 Consider the following nonlinear autonomous diﬀerential equation x ˙ (t) = F [x (t)] (2. we also have another useful theorem: 46 . then the linear diﬀerence equation x ˙ (t) = ax (t) is asymptotically stable (in the sense that x (t) → 0) if a < 0.27) is asymptotically stable. Moreover. in the sense that starting from any x (0) ∈ Rn .. Finally. Then the diﬀerential equation (2. F (x∗ ) = 0. Then the diﬀerential equation (2.28) is locally asymptotically stable. B (x∗ ) ⊂ Rn such that starting from any x (0) ∈ B (x∗ ).

the Cobb-Douglas production function: Example 1 Supposed the aggregate production function is given by F [K. Moreover. It is a good vehicle to illustrate issues. then the basic Solow growth model in continuous time with no population growth and technological change is asymptotically stable. L] = AK α L1−α with 0 < α < 1. In view of these results. Proof. k (t) → k∗ . This production function is very easy to work with. Proposition 5 immediately generalizes: Proposition 8 Suppose that Assumptions 1 and 2 hold. sf (k) −(n + δ ) k < 0. Notice that the equivalent of Theorem 6 is not true in discrete time. and suppose that there exists a unique x∗ such that g (x∗ ) = 0. but it also has many special features that are far from general.451: Introduction to Economic Growth Theorem 6 Let g : R → R be a continuous function. and starting from any k (0) > 0.14. in particular because it has an elasticity of substitution equal to 1 between capital and labor. suppose g (x) < 0 for all x > x∗ and g (x) > 0 for all ˙ (t) = g (x (t)) is a (globally) asymptotically x < x∗ . It is also useful at this point to look at one of the most common examples of the production function used in macroeconomics. You should remember from basic micro theory that the Cobb-Douglas production function is extremely special. Then the nonlinear diﬀerential equation x stable. The proof of stability is now simpler and follows immediately from Theorem 6 by noting that whenever k < k ∗ . but you should not think that all production functions are Cobb-Douglas! 47 . and starting with any x (0). x (t) → x∗ . sf (k) −(n + δ ) k > 0 and whenever k > k ∗ . and this will be illustrated by one of the problems in Problem Set 1.

14. we have that f (k) = Ak α . 48 µ sA n+δ . It can be immediately calculated that.451: Introduction to Economic Growth One very important feature of the Cobb-Douglas production function is that factor shares are constant. let x (t) ≡ k (t)1−α . the share of labor is αL (t) = 1 − α. so the equilibrium law of motion of the capital labor ratio can be written in terms of x (t) as x ˙ (t) = (1 − α) sA − (1 − α) (n + δ) x (t) . so the steady state is given again from (2.26) (with population growth at the rate n) as A (k∗ )α−1 = or k = ∗ n+δ s 1 ¶ 1− α which is a very nice and simple interpretable form for the steady-state capital-labor ratio. In particular. we have: ˙ (t) = sA [k (t)]α − (n + δ ) k (t) k with initial condition k (0). Similarly. With this production function. L (t)) K (t) = Y (t) αA [K (t)]α−1 [L (t)]1−α K (t) = A [K (t)]α [L (t)]1−α = α. we have the share of capital is constant irrespective of the capital-labor ratio: αK (t) = R (t) K (t) Y (t) FK (K (t). with competitive factor markets. To solve this equation. Transitional dynamics are also straightforward in this case. .

and in fact. n+δ δ This solution illustrates that starting from any k (0).3 A First Look at Sustained Growth Before discussing technological progress. adjustment of the capital-labor ratio back to its steady-state level can be very very slow. let us relax Assumptions 1 and 2 (which do not allow α = 1). 49 (2.451: Introduction to Economic Growth which is a linear diﬀerential equation. This is intuitive: a higher α implies less diminishing returns to capital. Similarly a smaller δ means less replacement of depreciated capital and a smaller n means slower population growth. To do this. the equilibrium k (t) → k∗ = (sA/ (n + δ ))1/(1−α) . A (t)] = AK (t) . 2. L (t) . it is useful to see how the model we have developed so far can generate sustained growth (without technological progress). the simplest model of sustained growth essentially takes α = 1 in terms of the Cobb-Douglas production function above.2. which slows down dynamics. with a general solution ∙ ¸ sA sA x (t) = + x (0) − exp (− (1 − α) (n + δ ) t) n+δ n+δ or in terms of the capital-labor ratio 1 ½ ∙ ¸ ¾ 1− α sA sA 1−α k (t) = + [k (0)] exp (− (1 − α) (n + δ ) t) − . A very slow adjustment towards a steady-state has the ﬂavor of “sustained growth” rather than the system settling down to a stationary point quickly. In fact. and suppose that F [K (t) .14. both of those slowing down the adjustment of capital per worker and thus transitional dynamics. The Cobb-Douglas example above already shows that when α is close to 1. the rate of adjustment is related to (1 − α) (n + δ ).29) .

the fundamental law of motion of the capital stock is given by (again with population growth given by (2. The economy always grows 50 .14. This immediately establishes the following proposition: Proposition 9 Consider the Solow growth model with the production function (2.24)): ˙ (t) k = sA − δ − n. but also shows that in this simplest form. In particular. and given (2.30) to Problem Set 1. starting with a capital-labor ratio k (0) > 0. This proposition not only establishes the possibility of endogenous growth. for example.451: Introduction to Economic Growth where A > 0 is a constant. (2. the economy has k (t) = exp ((sA − δ − n) t) k (0) and y (t) = exp ((sA − δ − n) t) Ak (0) . there is sustained growth in the capital-labor ratio. Then in equilibrium. there are no transitional dynamics. if sA − δ − n > 0. there is sustained growth of income per capita at the rate sA − δ − n. The results I would like to highlight apply with a more general constant returns to scale production function.29) and suppose that sA − δ − n > 0.29). there is sustained growth in income per capita. With this production function.30) but it is simpler to illustrate the main insights with (2. k (t) Therefore. and in its simplest form output does not even depend on labor. leaving the analysis of the richer production function (2. This is the so-called “AK” model. L (t) .29). F [K (t) . A (t)] = AK (t) + BL (t) .

L (t) .3. has progressed tremendously over the past 200 years. The next ﬁgure shows this equilibrium diagrammatically. We now introduce changes in A (t) to capture improvements in the technological know-how of the economy.14. irrespective of what level of capital-labor issue it starts from. denoting the growth rate of the economy (and the capital-labor ratio by γ K ): 2. In 51 . and even more tremendously over the past 1000 or 10. There is little doubt that what human societies know to produce.1 Solow Model with Technological Progress Balanced Growth The models analyzed so far did not feature technological progress.3 2. and how eﬃciently they can produce them.451: Introduction to Economic Growth at a constant rate sA − δ − n. An attractive way of introducing economic growth is to allow technological progress. A (t)] is too general to achieve our objective.000 years. The question is how to do this. At some level we will see that the production function F [K (t) .

there is no trend. with this general structure. These are sometimes referred to as the Kaldor facts. Despite fairly large ﬂuctuations. We are ignoring the share of land here as we did in the analysis so far: land is not a major factor of production. The next picture. 100% 90% Labor and capital share in total value added 80% 70% 60% 50% 40% 30% 20% 10% 0% 1929 1934 1939 1944 1949 1954 1959 1964 1969 1974 1979 1984 1989 1994 Labor Capital Capital and Labor Share in the U. for example. This 52 .451: Introduction to Economic Growth particular. we may not have balanced growth. while the labor share is about 2/3. while income per capita increases.S. This and the relative constancy of capital-output ratios until the 1970s have made many economists prefer models with balanced growth to those without. By balanced growth. (Since the 1970s capital-output ratios may or may not be constant depending on how you measure them). we mean a path of the economy in which. note that the capital share in national income is about 1/3. GDP. Also for future reference. shows the evolution of the share of capital in national income in the United States.14. the capital-labor ratio and the distribution of income between capital and labor is roughly constant.

In reality. So it is an advantage to have models featuring balanced growth. F [K (t) . In any case. but of course it is very special. but for this course. manufacturing ﬁrst increasing and then shrinking. A (t)]? First we could have ˜ [K (t) . Ultimately. the most important characteristic of balanced growth is that it is much easier to handle than non-balanced growth. 2. we will largely focus on models with balanced growth. with agriculture shrinking. This is known as Hicks-neutral technological progress. L (t)] .14.3.2 Neutral Technological Progress What are some convenient special forms of the general production function F [K (t) . L (t) . For example. For us. and we should think about how incorporating land into this picture changes the patterns. this pattern of factor distribution of income. This production function does a good job in certain circumstances. A (t)] = A (t) F so that technological progress simply multiplies output. we would like to have models that combine certain quasi-balanced features with these types of structural transformations embedded in them. technological progress simply corresponds to a relabeling of the isoquants (without any change in their shape). often makes them choose an aggregate production function of the form AK 1/3 L2/3 as an approximation to reality (especially since it ensures that factor shares are constant by construction). in this case if we think of the isoquants in the L-K space. These are interesting frontiers of research. the share of diﬀerent sectors changes systematically over the growth process. growth has many non-balanced features. 53 . Intuitively. combined with economists’ desire to work with simple models. L (t) .451: Introduction to Economic Growth is clearly not the case for the poor countries.

54 . in the form ˜ [A (t) K (t) . A (t) L (t)] . In particular. Of course. so we could have a vector valued at index of technology and a production function that looks like ˜ [AK (t) K (t) .14. A (t)] = F whereby an increase in technology increases output as if the economy had more labor. L (t) . and it is also somewhat troubling. L (t) . Equivalently. F [K (t) . We now state and prove the relevant theorem here. AL (t) L (t)] . This is a very surprising result. we can have labor-augmenting or Harrod-neutral technological progress ˜ [K (t) . F [K (t) .451: Introduction to Economic Growth Another alternative is to have capital-augmenting or Solow-neutral technological progress. the slope of the isoquants are constant along rays with constant capital-output ratio. since we have no idea why technological progress should take this form. in practice technological change can be a mixture of these. A (t)] = F This is referred to as capital-augmenting progress. L (t) . although all of these forms of technological progress look equally plausible ex ante. This type of technological progress corresponds to the isoquants shifting with technological progress in a way that they have constant slope at a given labor-output ratio. balanced growth forces us to one of these types of neutral technological progress. L (t)] . Finally. F [K (t) . A (t)] = AH (t) F It turns out that. balanced growth necessitates that all technological progress be labor augmenting or Harrod-neutral. because a higher A (t) is equivalent to the economy having more capital.

K Suppose also that there is a constant growth rate of population. L (t) . y ˙ (t) /y (t) = g > 0. let us focus on continuous time models. and then also give a more heuristic proof.e. The key elements of balanced growth. we have that αL (t) + αK (t) = 1.e. and ˙ (t) A = g. 1) as t → ∞.14.. Y ∗ (t) = F where *’s denote asymptotic steady-state values. are the constancy of factor shares and the constancy of the capital-output ratio. L (t) . i.. i. by factor shares.. αK (t) = α (x∗ ) ∈ (0. For simplicity and without loss of any generality. Y (t) Y (t) By Assumption 1 and Theorem 1. and factor shares are nonzero and constant. we mean αL (t) ≡ w (t) L (t) R (t) K (t) and αK (t) ≡ . K (t) /Y (t).3. A (t)] − C (t) − δK (t) . A (t) 55 .451: Introduction to Economic Growth 2. the production function can be represented as: ˜ [K ∗ (t) . i. If a balanced growth path exists with constant capital-output ratio and per capita growth rate. A (t) L (t)] . then asymptotically.e. Since there is only labor and capital in this model. The following theorem was ﬁrst stated and proved by Uzawa. as suggested by the discussion above. A (t)] and capital accumulation equation ˙ (t) = F [K (t) . L (t) = exp (nt) L (0). Here I present a version of Uzawa’s proof along the lines of the more recent paper by Jones and Scrimgeour (2005). Theorem 7 (Uzawa) Consider a growth model with a constant returns to scale aggregate production function F [K (t) .3 The Steady-State Technological Progress Theorem A version of the following theorem was ﬁrst proved by Uzawa in 1961.

which only depends on time. also. ∂ log x (t) 1 − α (x∗ ) Integrating both sides and noting that the right hand side does not depend on time. and the share of capital in national income is potentially a function of this capital-output ratio. Therefore. ξ (x∗ ) is invertible in the neighborhood of x∗ . Now let x (t) ≡ K (t) /Y (t).451: Introduction to Economic Growth Proof.A(t)] αK (t) .A(t)]K (t) F [K (t).L(t). we have the following partial diﬀerential equation: ∂ log y (t) α (x∗ ) = . we have Z α (x∗ ) dx∗ log y (t) = a (t) + 1 − α (x∗ ) x∗ for some function a (t). asymptotically.L(t). we have y (t) = A (t) ξ (x∗ ) . Taking exponents. Notice.14. and by hypothesis asymptotically αK (t) = α (x∗ ) where x∗ refers to the steady state value of K/Y . with inverse . ³R ´ α(x∗ ) dx∗ ∗ . 1 − αK (t) ´−1 where the last line uses the deﬁnition of αK (t). for future use that where A (t) ≡ exp (a (t)) and ξ (x ) ≡ exp 1−α(x∗ ) x∗ denoted by ξ −1 (y/A) 56 from the inverse function theorem.A(t)] ∂ log K (t) 1 ´−1 −1 −1 FK [K (t).L(t). Let us look at the following derivative v ≡ = ∂ log y (t) ∂ log (k (t) /y (t)) 1 ∂ log K (t) ∂ log Y (t) −1 = ³ = ³ = 1 ∂ log F [K (t).

which is the Cobb-Douglas production function. by deﬁnition. we must have A (t) ≡ exp (gt) A (0). A (t) A (t) ¸ K ∗ (t) Y (t) = A (t) L (t) f .14. Finally. the share of either capital or labor will be increasing over time. grow at the same rate. For a more heuristic reasoning. AL (t) L (t). otherwise. Capital accumulation implies that K (t) will grow at the same rate as AL (t) L (t). note that.451: Introduction to Economic Growth Since ξ (x∗ ) is constant and y ˙ (t) /y (t) = g. A (t) L (t)] . A (t) L (t) ∗ and thus ∙ which. AK (t) K (t). Y ∗ (t) = F completing the proof. consider production function of the form F [AK (t) K (t) . which can only be the case when total capital inputs. Thus balanced growth can only be possible if AK (t) is asymptotically constant. k (t) = x (t) y (t). which implies asymptotically (in steady state) that µ ¶ y ∗ (t) −1 y ∗ (t) k∗ (t) = ξ A (t) A (t) A (t) ∙ ∗ ¸ y (t) = f −1 A (t) or ∙ ∗ ¸ y ∗ (t) k (t) =f . where we can have Y (t) = [AK (t) K (t)]α [AL (t)L(t)]1−α 57 . is another way of writing ˜ [K ∗ (t) . Balanced growth requires factor shares to be constant. under constant returns to scale. There is one exception to this. AL (t) L (t)]. and total labor inputs.

K (t) AK (t) K (t) ¶ µ L (t) .e.e. AL (t) L (t)]. K (t) K (t) and in steady state. AK (0) and AL (0) to1.451: Introduction to Economic Growth and both AK (t) and AL (t) could grow asymptotically. is possible to give the more heuristic proof of Theorem 7. i. AK (t) and AL (t) are growing asymptotically at the rates gH . and since we are interested in asymptotic states. while maintaining balanced growth. technological change can be represented as purely labor augmenting. we have Y (t) /K (t) con˙ (t) /K (t) = g . Alternative Proof of Theorem 7: Suppose that Y (t) = AH (t) F [AK (t) K (t) . gK and gL . in other words. which is what Theorem 7 requires. It is quite straightit should have a representation of the form Y ∗ (t) = F forward to see that in this Cobb-Douglas example we can deﬁne A (t) = [AK (t)]α/(1−α) AL (t). Based on these ideas. Then we can write that asymptotically ¸ ∙ AL (t) L (t) Y (t) = exp ((gH + gK ) t) F 1.14. notice that Theorem 7 does not require that Y ∗ (t) = F ˜ [K ∗ (t) . so K 58 . Notice ﬁnally that this theorem does not state that technological change has to be labor augmenting all the time. ˜ [K ∗ (t) . according to the hypotheses of the theorem. suppose that AH (t)... but that However. capital grows at the same rate as total output. i. But it requires that it has to be labor augmenting asymptotically. along the balanced growth path. and the production function can be represented as Y (t) = [K (t)]α [A(t)L(t)]1−α . Normalize AH (0). Combined stant. A (t) L (t)]. A (t) L (t)]. ≡ exp ((gH + gK ) t) f exp ((gL − gK ) t) K (t) Now we also have ˙ (t) Y (t) K =s − δ.

K (t) But from this equation Y (t) /K (t) can remain constant only under the one of the two following circumstances: 1.e.14. let us assume that it takes this form throughout. and g = gL + n.. which is only possible when f (x) = xβ for some β. Moreover.31) . From Theorem 7. exp ((gH + gK ) t) is constant and exp ((gL − gK + n − g) t) is constant. A (t) 59 (2. exp ((gH + gK ) t) increases exactly at the same rate as f (exp ((gL − gK + n − g ) t)) decreases. if we impose Assumption 1 (or just CRS and positive marginal products) then we get β ∈ (0. 1).451: Introduction to Economic Growth with the hypothesis that L (t) = exp (nt) L (0).Then. A (t) L (t)] . This completes the alternative proof of Theorem 7.e. i. I will only present the analysis for continuous time (the discrete time case is equivalent). For simplicity. ¥ 2. Y (t) = exp ((gH + gK ) t) f (exp ((gL − gK + n − g ) t)) .4 The Solow Growth Model with Technological Progress: Continuous Time Now we are ready to analyze the Solow growth model with technological progress. gH = gK = 0. 2. we know that the production function must take the form F [K (t) .3. with purely labor-augmenting technological progress asymptotically. this then implies (for L(0) normalized to 1). ˙ (t) A = g. suppose that there is technological progress at the rate g .. i.

we can deﬁne k (t) ≡ K (t) . K (2. Since eﬀective units of labor are given by A (t) L (t).e.34) Income per capita is y (t) ≡ Y (t) /L (t). (2. A (t) L (t)] − δK (t) .14. y (t) = A (t) y ˆ (t) .451: Introduction to Economic Growth and population growth at the rate n. ˙ (t) L = n. L (t) Again using the constant savings rate we have ˙ (t) = sF [K (t) .1 A (t) L (t) ≡ f (k (t)) ..32) The simplest way of analyzing this economy is again to express everything in terms of a normalized variable. and F exhibits constant returns to scale in its two arguments (by virtue of exhibiting constant returns to scale in capital and labor). 60 .33) Now diﬀerentiating this expression with respect to time. A (t) L (t) (2. i. we obtain ˙ (t) ˙ (t) K k = −g−n k (t) K (t) The quantity of output per unit of eﬀective labor can be written as y ˆ (t) ≡ Y (t) A (t) L (t) ∙ ¸ K (t) = F .

we have: Proposition 10 Consider the basic Solow growth model in continuous time.32) into (2. Denote the balanced growth path level of eﬀective capital-labor ratio by k∗ (A (0) . Consequently.451: Introduction to Economic Growth ˙ (t) from (2.33). Proposition 11 Suppose Assumptions 1 and 2 hold and let A (0) be the initial level of technology. k (t) K (t) Now using (2. ∞) and is given by f (k∗ ) δ+g+n = . and deﬁne the eﬀective capital-labor ratio as in (2. ˙ (t) k sf (k (t)) = − δ − g − n. A (t) L (t)] = − δ − g − n. A (t).33). Suppose that Assumptions 1 and 2 hold. The comparative static results are also similar to before. with Harrodneutral technological progress at the rate g and population growth at the rate n. A (0) (since the level of technology later. (2.25). we have Now substituting for K ˙ (t) k sF [K (t) . with the additional comparative static with respect to the initial level of the labor-augmenting technology.34). ∗ k s Per capita output and consumption grow at the rate g . n) 61 . Then there exists a unique steady state equilibrium where the eﬀective capital-labor ratio is equal to k∗ ∈ (0.14. k (t) k (t) (2. is completely determined by A (0) given the assumption in (2.35) which is very similar to the law of motion of the capital-labor ratio in the continuous time model.31)). s. δ. An equilibrium in this model is deﬁned similarly to before.

451: Introduction to Economic Growth and the level of income per capita by y ∗ (A (0) .14. δ. n. The major diﬀerence. s. ∂A (0) ∂s ∂n ∂δ and also ∂y ∗ (A (0) . n) ∂k∗ (A (0) . < 0 and < 0. t) ∂y ∗ (A (0) . s. t) > 0. < 0 and < 0. n) = 0. δ. δ. s. s. n) ∂k∗ (A (0) . 62 . ∂A (0) ∂s ∂n ∂δ Finally. The growth rate is exactly the same as the exogenous growth rate of the technology stock.e. n) ∂k∗ (A (0) . and starting from any k (0) > 0. Proposition 12 Suppose that Assumptions 1 and 2 hold. is that now the model generates growth in income per capita. Then we have ∂k∗ (A (0) . δ. the eﬀective capital-labor ratio converges to a steady-state value k∗ . t) ∂y ∗ (A (0) . Therefore. However the disadvantage is that this growth is driven entirely exogenously. δ.. t) ∂y ∗ (A (0) . s. n. δ. of course. n. the comparative statics and dynamics are very similar to the model without technological progress (and without population growth). then the Solow growth model with Harrod-neutral technological progress and population growth in continuous time is asymptotically stable. > 0. > 0. δ. The model does not specify where this technology stock comes from and how fast it grows. s. s. we also have very similar transitional dynamics. n. δ. k (t) → k∗ . n. so can be mapped to the data much better. δ. i. s. t) (the latter is a function of time since it is growing over time). s.

e. growth accounting. A (t)] .1 Growth Accounting Let us go back to the most general form of the aggregate production function given by (2.Chapter 3 The Solow Model and the Data One of the important uses of the aggregate production function approach and the basic Solow model is that they provide us with a simple vehicle to look at the data. i. both at growth over time and income-level diﬀerences (and growth rate diﬀerences) across countries. I start here with over-time changes. which involves looking at cross-country diﬀerences. whereby Y (t) = F [K (t) .. 3. and then will move to the more important application for the purposes of this course. L (t) .1). 63 .

In particular. we need to make further assumptions. we have x = g − αK gK − αL gL . Y (t) = F then we have ˙ 1 A [g − αK gK − αL gL ] . This is the fundamental growth accounting equation. if we assume that the production function takes the standard labor-augmenting form ˜ [K (t) . output growth. and also deﬁning gL ≡ L/L ˙ ˙ /Y .451: Introduction to Economic Growth Diﬀerentiate this function with respect to time on both sides to obtain (dropping timedependence) ˙ FK K K ˙ ˙ ˙ FA A A FL L L Y = + + . labor force growth and capital stock growth. denoting an estimate by “^”. gK ≡ K/K and Recalling the deﬁnition of factor shares above. we have the estimate of TFP growth as: x ˆ = g − αK gK − αL gL . = A αL 64 . ˙ If we are interested in A/A rather than x. For example. Y Y A Y K Y L ˙ . This equation lets us estimate the contribution of technological progress to economic growth from factor shares. and denoting g ≡ Y x≡ ˙ FA A A Y A as the contribution of technology to growth.14. This contribution from technological progress is also referred to as Total Factor Productivity (TFP) or sometimes as Multi Factor Productivity. A (t) L (t)] .

t+1 = gt.t. there is a problem.t+1 − α where α ¯ K.t+1 ≡ and α ¯ L. across industries and across ﬁrms. Applying this method. This is still an active research area. for example over a year (or sometimes with the better data. In this case.t.451: Introduction to Economic Growth ˙ but this equation is not particularly useful. instead of instantaneous changes. It can be shown that this could lead to serious biases.t. The most common way of dealing with this is to use factor shares calculated as the average of the two points in time. This has been a landmark ﬁnding. we have ¯ K. since A/A is not something we are inherently interested in.t+1 gK. for a change between times t and t + 1. partly because there are conceptual issues about how far one should go in adjusting the quality of inputs. of course. In practice. We will return to these issues again below. better computers can translate into more capital. In continuous time. but at the end of the day better computers are a result of better technology. Since then. reducing the TFP residual. 65 αK.14. factor shares can change. Much more interesting is precisely x ˆ. this equation is exact. since over the time horizon in question.t. most notably Dale Jorgensen.t+1 gL. have attempted to reduce the amount due to the “residual” technology by adjusting for the quality of labor and capital inputs.t. For example. focusing the attention of economists on sources of technology diﬀerences over time. we look at changes over discrete time periods.t+1 − α ¯ L.t+1 2 .t+1 is deﬁned similarly. Solow found that much of economic growth over the 20th century was due to technological progress. x ˆt.t + αK. Therefore in discrete time. across nations.t. many economists. perhaps over a quarter or a month).t+1 .

In other words. 3. Since our purpose here is to look at cross-country income diﬀerences.451: Introduction to Economic Growth 3. I start with this case of no interdependence. and then look at the empirical evidence. which is that the world consists of a cross-section of countries which do not interact. these countries do not trade ﬁnancial assets. Therefore. among them Mankiw.14. Barro (1991) and much of Barro and Sala-i-Martin (2004). but interdependences arising from technology ﬂows and international trade will be discussed below. a basic estimation which does not take human capital into account proved to be inadequate. These countries inhabit the world. adopted by many authors. I ﬁrst develop this model brieﬂy. (3.2. Mankiw. However. Here already there is a major (and at some level a very problematic assumption). The simplest way of doing this is to follow the approach of Mankiw. but they are all islands onto themselves. goods. or there is no slow diﬀusion of technology across these countries. These authors basically estimated a cross country regression inspired by the above model. from the beginning. Romer and Weil (1992).1) 66 .2 Solow Model and Cross-Country Income Diﬀerences We are now in a position to take the basic Solow model to the data. Romer and Weil (1992).1 Solow Model with Human Capital Suppose that output in country j is given by β α Yj = Kj Hj (Aj Lj )1−α−β . Romer and Weil (1992) used an augmented Solow also incorporating human capital. I present the model for a cross-section of countries.

14.1) rather than Yj = 1−α (Aj Hj )α with Hj interpreted as eﬃciency units of labor. Romer and Weil (1992). Y is total output. which enables them to take the augmented-Solow model to the data is the following: Common technology advances assumption: Aj (t) = Aj exp (gt) . This is in part motivated by the relative stability of 67 . j denotes country. g .6) below).e. Aj Lj Aj Lj kj ≡ and deﬁne yj ≡ Yj /Lj as output per worker. A key assumption of Mankiw. In fact. L is labor. but they share the same common technology growth rate.2) Suppose also that population grows at a constant rate nj in country j . we should solve the model. α + β ≤ 1. That is. β ≥ 0.. this Kj latter approach is much more in line with the Becker model of human capital. We have α. This model cannot be easily taken to the data because we have no idea what Aj is. Then β α hj . we can use the usual trick of the neoclassical growth model of transforming variables to per capita eﬀective units: Kj Hj and hj ≡ . equation (3. But before seeing why this is. yj = Aj kj (3. First. countries may diﬀer according to their technology level. The important assumption here is that human capital is taken to be a diﬀerent factor of production rather than simply augmenting labor (i. H is human capital. and writing the model in this way is not without loss of any generality (as we will see below). see (3. A is labor-augmenting technical change.451: Introduction to Economic Growth where I have dropped time to simplify notation.

Next. population growth 68 .4) and solving yields the following steady-state ˙ j = 0 and h setting k values of physical capital and human capital ratios to eﬀective labor: ∗ kj ⎛Ã = ⎝ ⎛Ã = ⎝ sk j nj + g + δ k sk j nj + g + δ k !1−α Ã !β Ã sh j nj + g + δh sh j h∗ j nj + g + δ Now substituting back into (3. Thus ˙ j = 0 in (3.2) and taking logs. consider constant savings rates for human and physical capital. in steady state.5) This is an equation which can be estimated using cross-country data if we have measures of k sh j .451: Introduction to Economic Growth the world income distribution discussed earlier. h j Aj (3. and the world income distribution would become more and more dispersed. In addition.3) (3. In the absence of this assumption.14.3) and (3.4) As in our baseline models. we obtain ! ! Ã Ã sh sk β α j j + ln yj = ln Aj + gt + ln ln 1−α−β 1−α−β nj + g + δ h nj + g + δ k 1 !1−β ⎞ 1−α −β ⎠ . both kj and hj have to be constant. Then µ ¶ ¡ ¢ yj k ˙ − nj + g + δ k kj kj = sj A µ j¶ ¡ ¢ ˙ j = sh yj − nj + g + δ h hj . h 1 !α ⎞ 1−α −β ⎠ (3. as a direct generalization of the standard Solow model: k h h ˙ j = sk ˙ K j Yj − δ Kj and Hj = sj Yj − δ Hj where δ ’s denote constant depreciation rates. countries would grow at diﬀerent rates. we can use investment rates (investments/GDP) for sj .

This is what Mankiw.. because the term ln Aj is unobserved to the econometrician. If human capital is not included.. Therefore implicitly. with all of these assumptions. This is shown in the next table.. Romer and Weil do (or they estimate a version of this with δ h = δ k ). and could be correlated with all of the other right hand side variables. is this a good proxy for investment in human capital?.. the ﬁt is not very good and the estimates are not reasonable. Mankiw. Romer and Weil make another crucial assumption. However.5). 69 . Mankiw.5) can still not be estimated. and the standard depreciation rates for δ k . equation (3. The estimation is a success for the augmented-Solow model.451: Introduction to Economic Growth rates nj .14. Romer and Weil estimate equation (3.]. With these assumptions. considerably stronger than the “common technology advances” assumption: Orthogonal technology assumption: Aj = εj A where εj is orthogonal to all other country variables. They approximate sh j using the fraction of the working age population enrolled in school [.

the coeﬃcient in front of the investment/GDP ratio should be β/ (1 − β ). 70 . Now the parameter estimates imply α ≈ 1/3. which is far too high bearing in mind that given the factor distribution of income we expect the exponent of capital in the production function to be closer to 1/3.78. β ≈ 1/3 and R2 ≈ .14. thus the estimate suggests β ' 0.6.451: Introduction to Economic Growth Without human capital. But for the augmented model with human capital. the ﬁt is very good as shown in the next table.

The estimate of α is consistent with a capital share of one-third in national income. and the R2 implies that almost 80 percent of the diﬀerences in income per capita can be explained by investment decisions (human and physical capital diﬀerences). When Aj varies across countries.14. these results provide strong support for the augmented Solow model.2. The orthogonal technology assumption is too strong. 3. Romer and Weil Approach But there are two major (and related) problems with this approach: 1.451: Introduction to Economic Growth At face value. 71 .2 Problems with the Mankiw.

. The coeﬃcient on sh j is too large. one would need very very large human capital externalities 72 . In practice. This variable ranges from 0. and E is years of schooling.24. recall that Mankiw.24) − 1 ≈ 9 times. the predicted log diﬀerence in incomes between these two countries is α (ln 12 − ln (0.1).4)) ≈ 2. More explicitly. 2. the diﬀerence in average years of schooling between any two countries over this time period is less than 12. Here φ is estimated to be between 0. The labor literature suggests that additional years of schooling is associated with a 6 to 10 percent increase in individual earnings (e.g.14. This implies that a worker with one more year of schooling is typically about 6 to 10 percent more productive. Romer and Weil use the fraction of the working age population enrolled in school.451: Introduction to Economic Growth k it will plausibly be correlated with our measures of sh j and sj .06 and 0. consider the individual level Mincer regression ln wi = Xi0 γ + φEi where w is wage income.4 to over 12 in the sample of countries used for this regression. Their estimates therefore imply that a country with approximately 12 for this variable should have income per capita about 9 times that of a country with sh j = 1! (This is holding all other variables constant). Xi is a set of demographic controls. So in the absence of human capital externalities. To see this. so there will be an omitted variable bias leading to overestimates of α and β as well as an exaggeration of the R2 . 1−α−β and exp (2. a country with 12 more years of average schooling should be at most twice as rich instead of 9 times as rich! Even allowing for human capital externalities.

but in one economy all workers have E1 years of schooling. open capital accounts). show that they are rather small). consider a simple competitive economy. all workers. irrespective of their level of schooling. both economies will function at the same physical to human capital ratio. To understand this last point. Taking logs of this equation. they have E2 > E1 schooling. Now consider two economies with the same technology. we immediately obtain Yi = A (1 − α)(1−α)/α r−(1−α)/α exp (φEi ) . How large should the income gap between these two countries be? Using the fact that with the same interest rate. Wages are equal to marginal product. so w (h) = α (1 − α)(1−α)/α Ar−(1−α)/α h So wages are linear in human capital due to constant returns to scale.451: Introduction to Economic Growth in order to get this type of results (existing estimates of human capital externalities. First-order condition from ﬁrm maximization gives r = (1 − α) (Ah/k )α . while in the other. and human capital is a function of schooling. 73 . with the standard exponential form hi = exp (φEi ). with the slope coeﬃcient on education measuring the relationship between education and human capital. the same technology. Acemoglu and Angrist. Suppose that each ﬁrm has a production function y = k 1−α (Ah)α Firms face cost of capital r. 2000. for example. will work exactly at the same physical to human capital ratio. In other words. the same interest rate (for example. we end up with the standard log linear wage equation ln wi = cst + φEi .14.

approximately 0. we obtain log Y = 0. as a function of average human capital in the economy. the increase in the individual’s wage as a 74 . and φ is about 6 percent. and the quantitative eﬀect of capital (as a proxy for interest rates) is plausible. we obtain that log Y2 − log Y1 = φ (E2 − E1 ). So if one economy has on average one year more of schooling. In this regression. the rate of return to human capital in the Mincer regressions would only reﬂect the private return–that is. In this case. there is still a very large eﬀect of education on income. there are much larger diﬀerences.14. its income should be 6 percent higher.408 (0. This relationship between education and income may reﬂect human capital externalities. This result is not simply explained by the fact that interest rates vary across countries. In particular. For example. so including the (log) capital output ratio would be one way to control for interest rate diﬀerences. the correlation between income and schooling is “too strong” relative to what we should expect on the basis of micro evidence.451: Introduction to Economic Growth Or.178) log ¡K ¢ Y E So. Running this regression with 1985 data.5 taking α as 2/3. In the data.266 (0.027) In other words.313 (0. the “eﬀect” of schooling on income is much larger than the 6-10 percent diﬀerence expected.033) E + 0. Notice that we can write r = (1 − α) Y /K . A. For example a cross-country regression of income per capita on average years of schooling in 1985 gives log Y = 0. the log capital-output should have a coeﬃcient of (1 − α) /α. we might have the productivity term. taking logs.

or because some third factor aﬀects both human capital and technology.6) with Hj interpreted as eﬃciency units of labor. existing evidence indicates that human capital externalities are limited. Therefore. as noted above. This is the approach ﬁrst taken by Bils and Klenow. the private plus the external eﬀect of schooling. 3. Consider the following production function 1−α (Aj Hj )α Yj = Kj (3. But regressions using aggregate data would capture the total eﬀect of an increase in human capital on income–that is. and then by Klenow and Rodriguez and Hall and Jones. The disadvantage is that certain assumptions on functional forms have to be taken much more seriously.451: Introduction to Economic Growth function of his own human capital. Aj ’s. holding average human capital constant. However. The advantage of the calibration approach is that the omitted variable bias underlying the estimates of Mankiw. and we explicitly have to assume no human capital externalities.2. Romer and Weil will be less important (since microlevel evidence is being used to anchor the contribution of human capital). Such a pattern of correlation may arise because human capital responds to technology.14.3 The Macro Mincer Approach (Bils-Klenow-Rodriguez-HallJones) A related approach is to use calibration/levels accounting rather than regression analysis and make use of the ﬁndings of Mincer (micro wage) regressions. The alternative interpretation of the patterns is that there are diﬀerences in technologies. one possibility is that there are large human capital externalities. and these are correlated with human capital diﬀerences. Assume the following Mincer-type relation75 . Here let me follow Hall and Jones.

S. accounting for as much as 50 percent of the actual diﬀerences in output per worker.14. for example. We can use diﬀerent values for φ (E ) and construct alternative estimates of Hj . which can be constructed using standard perpetual inventory methods. as ¡ S ¢2/3 ˆj = K 1/3 AU Y t Hj j and compare these predicted incomes with actual incomes. 76 . They ﬁnd: 1. Diﬀerences in physical and human capital still matter a lot. Alternatively. Once we have a series for Hj and one for Kj . 2. Hall and Jones (1999) use a piecewise linear speciﬁcation for φ (E ) based on work by Psacharapoulos from less developed countries (showing returns to earlier years of schooling that are greater than to higher education). But there are also signiﬁcant productivity diﬀerences.451: Introduction to Economic Growth ship between human capital and education Hj = X E exp {φ (E )} Lj (E ) where φ (E ) is the rate of return to E years of schooling and Lj (E ) is the number of individuals in country j with E years of schooling. we can construct “predicted” incomes.) as Aj t = S AU t Ã Ytj YtU S !3/2 µ KtU S Ktj ¶1/2 µ HtU S Htj ¶ Hall and Jones perform this exercise using output per worker rather than income per capita. we could back out country-speciﬁc technology terms (relative to the U.

14.451: Introduction to Economic Growth The next ﬁgure and the table show a summary of their results: 77 .

j +1 = gj. some of the assumptions of these calibration exercise can be relaxed.14.j. ranked the countries according to their capital-labor ratio (or capital-output ratio).j. 78 . Essentially.j +1 gK. one could do “levels accounting”. we can write x ˆj.451: Introduction to Economic Growth The conclusion of this calibration exercise is therefore very similar to the one that followed from the regression analysis presented in the previous section. Naturally.j +1 .j. For example instead of assuming at Cobb-Douglas production function.j +1 − α ¯ K. and then use the equivalent of the growth accounting equation above.j +1 − α ¯ Lj. in particular.j +1 gL.

Of course. one unit of labor (or one college graduate) in the U. This is what Treﬂer does to test an augmented version of the Heckscher-Ohlin approach to international trade. but allows for factor-speciﬁc productivity diﬀerences across countries.S. An alternative is to use additional data. It is important that technology diﬀerences take the form of factor productivity diﬀerences. productivity/technology diﬀerences are obtained as “residuals” from a calibration exercise.14. With this method. as the base country. so we have to trust the functional form assumptions used in this exercise. there are no diﬀerences in industry technologies) and share the same homothetic preferences (in particular. Although Treﬂer does not emphasize the implications of his ﬁndings for productivity diﬀerences.3 An Alternative Approach to Estimating Productivity Diﬀerences (Treﬂer) In the above approach. The same applies to 79 . In particular.j +1 is a proportional diﬀerence of labor supply between the two countries.e.j. they allocate consumption expenditures across goods in the same manner). gL. all countries share the same technology (i. for example the United States. and x ˆj. thus gK. Other than these factor-speciﬁc diﬀerences.. for this we need to have good measures of factor shares in diﬀerent countries which are not always available. a byproduct of his analysis is a series of estimates for diﬀerences in factor productivities across countries. we can calculate relative technology diﬀerences across countries. 3. and taking one of the countries. could be more productive than one unit of labor (or one college graduate) in Nigeria.451: Introduction to Economic Growth where j stands for country. Treﬂer starts from the standard Heckscher-Ohlin model of international trade.j +1 is the TFP diﬀerence.j.j +1 is the proportional diﬀerence in capital stock between countries j and j + 1.

With data for any pair of countries.7) where Vjf is that endowment of factor f in country j . in the absence of any trading frictions and with identical (or homothetic) preferences. f ’s. and sj is the share of country j in world consumption (this uses the assumption that all countries have the same homothetic preferences). It turns out that the series implied by (3. since capital-augmenting technology diﬀerences are allowed and the elasticity of substitution between diﬀerent factors is not assumed to be equal to 1. we should also have f wj πf j = f wj 0 πf j0 (3.7) and (3.8) are very similar.14. So from this equation we can obtain an estimate of the diﬀerences in factor productivities. where wj on factor prices.8) f is the price of factor f in country j .1). is exports of country j . the Xj a unique sequence of π f j ’s taking one of the countries as the base. This speciﬁcation of course is more general than the production function in (3. The major contribution of Treﬂer’s paper is to note that if there is factor price equalization. equation (3. A standard equation in international trade is that. j and j 0 . The following ﬁgure shows his estimates: 80 . N is the total number of countries. so there appears to be some validity to this approach.451: Introduction to Economic Growth capital. π f j is the factor productivity of factor f in country j . the net export of factor f embedded in the f . this may be viewed simply as an untested strong hypothesis.7) solves for Given estimates of the net export of factor contents. Xj f Xj f πf j Vj N X i=1 f πf i Vi = − sj (3. we can therefore construct alternative series for π f j ’s. At this level.

81 . These numbers imply that there are very large diﬀerences in labor productivity. For example. we can presume that there is some information in the numbers that Treﬂer obtains.14. but much smaller diﬀerences in capital productivity. capital productivity diﬀerences are much more limited than labor productivity diﬀerences. labor in Pakistan is 1/25th as productive as labor in the United States. capital in Pakistan is only half as productive as capital in the United States.451: Introduction to Economic Growth Given this validation. and some substantial. For example. In contrast.

14.451: Introduction to Economic Growth 82 .

themselves.Chapter 4 Fundamental Determinants of Diﬀerences in Income 4. less physical capital. the observation that a country is poorer than another because it has worse technology. or other eﬃciency diﬀerences in the use of the factors. the framework we have does a very good job of helping us understand the proximate causes of income diﬀerences. At this level. human capital diﬀerences and technology diﬀerences. less human capital? This question 83 . less physical capital and less human capital immediately poses the next question: why does it have worse technology. These technology diﬀerences. The same procedure also helps us understand the proximate causes of the process of economic growth. However.1 From Proximate to Fundamental Causes The use of the Solow model and the production function approach illustrated how cross country income diﬀerences can be understood as resulting from physical capital diﬀerences. may represent actual diﬀerences in the technologies used by countries.

and at some level. in some sense. but less speciﬁc (one way of operationalizing it may be by using the multiple equilibrium models we will discuss below). and also in clarifying the mechanics of the process of growth. Growth theory is useful in highlighting the proximate causes. in providing us with a framework for thinking about the fundamental causes. A version of this hypothesis where a small diﬀerence caused by luck may lead to large persistent diﬀerences is also diﬃcult to reconcile with the data given the reversal documented above. But we have to take this additional step of looking for fundamental causes. about the fundamental causes of income per capita (and growth) diﬀerences across countries. It is diﬃcult to operationalize this approach.2 Hypotheses Why do some countries invest more in physical and human capital and possess better technologies? There are four sets of broad hypotheses: 1. otherwise what we have learned will be only partial. it is quite similar to the other hypotheses. Nevertheless. A version of this hypothesis where such diﬀerences are transitory is clearly not supported by the evidence presented so far. So I will place less emphasis on the importance of luck. 4.451: Introduction to Economic Growth is. Luck: some countries just turned out to be lucky. so that we can more carefully evaluate diﬀerent theories and approaches. which points out to very persistent diﬀerences over long periods. some of the theories presented below will show how small diﬀerences in initial condi84 .14.

For example.14. domesticated animals. Guns. 2.. Germs and Steel. Gunnar Myrdal (1968): “climate exerts everywhere a powerful inﬂuence on all forms of life. diﬀerences in geographic.. This idea dates back to Machiavelli and Montesquieu. he states that: “. These diﬀerences stemmed ultimately from Eurasia’s much longer history of densely populated. should take into account the climate and its impacts on soil. so far as they can be explained at all. Diamond argues that diﬀerences in the nature and history of food production. climatic and ecological characteristics across countries.” societies dependent on food production (1997. p. Geography: This view is becoming very popular recently. animals. on living conditions in economic development. In the economics circles. to a large extent. are due to the types of crops.proximate factors behind Europe’s conquest of the Americas were the diﬀerences in all aspects of technology.” The recent bestseller by Jared Diamond. The most common is the view that climate has a direct eﬀect on income through its inﬂuence on work eﬀort.” and that “serious study of the problems of underdevelopment. and technology. in turn. 358).. Alfred Marshall (1890) similarly wrote: “vigor depends partly on race qualities: but these. It claims that diﬀerences in economic performance reﬂect. humans and physical assets– in short. all of which are geographically determined characteristics... Jeﬀ Sachs has been pushing for this view. suggests that the timing of the Neolithic revolution has had a long lasting eﬀect by determining which societies were the ﬁrst ones to develop strong armies.451: Introduction to Economic Growth tions can lead to large ultimate diﬀerences. He argues that “Certain parts of the world are geographically 85 . seem to be chieﬂy due to climate”. vegetation.. and the axis of agricultural technology diﬀusion in diﬀerent continents.

32). such as climate) having an eﬀect on the long-run distribution of income across countries: 86 . Finally. Geographical advantages might include access to key natural resources. advantageous conditions for agriculture.” (2000. He further suggests that “Tropical agriculture faces several problems that lead to reduced productivity of perennial crops in general and of staple food crops in particular” (2000. advantageous conditions for human health. 32). 33-34). The following ﬁgure shows the geographical distribution of income per capita. pp.451: Introduction to Economic Growth favored. p. Sachs argues that the greater population in the temperate areas over the past centuries led to more rapid advances in technologies appropriate for these areas relative to technologies necessary for development in the tropics (2001. which is consistent with some geographic factors. p. 3 and 2000. proximity to other successful economies. p. 30). and that “The burden of infectious disease is similarly higher in the tropics than in the temperate zones” (2000.14. access to the coastline and sea– navigable rivers. p.

or equal opportunity. diﬀerences in economic performance largely reﬂect diﬀerences in the organization of society. 4. There are many versions of this hypothesis.14.451: Introduction to Economic Growth 3. or speciﬁc government policies are important for investment and eﬃciency (of course. others suggest that limited government. whether these policies are adopted is in turn determined by other factors). Societies that provide incentives and opportunities for investment will be richer than those that fail to do so. Culture and social capital: this view instead emphasizes whether societies are able to engender the values conducive to entrepreneurship or cooperation among agents. “Institutions”: according to this view. some of them suggesting that institutions that support property rights and rule of law are important. 87 .

in the culture view. Finally. the institutions view emphasizes much more the importance of conﬂict between diﬀerent groups or individuals as a determinant of social outcomes. the variables in X that enter signiﬁcantly can be interpreted as determinants of cross-country diﬀerences in growth. This is borne out both by growth regressions. and level regressions. since diﬀerences in physical and 88 . culture or social capital. There are two major diﬀerences between the institutions view and the culture view.451: Introduction to Economic Growth Popular versions of this story include the thesis by Max Weber on the importance of religion for capitalism. institutions and culture? Measures of each are strongly correlated with income per capita or other determinants of income. to a ﬁrst approximation. that is responsible for prosperity. whereas there is a more cooperative undertone to the culture view (especially in the social capital versions of this view). returning to growth regressions of the type (1. First. many versions of the culture view. Instead. it is the social organization of the society. and the recent work by Robert Putnam on social capital and co-operation (which is in turn related to some early work by Banﬁeld on lack of corporation in the South of Italy). For example. These regression analyses ﬁnd a variety of variables to be important in explaining growth. Second. cannot be changed. is changeable. But. such as those of Max Weber or David Landes.2). which. There is a very large literature on regressions of this sort. First. in the institutions view. emphasize religion or other predetermined factors as crucial determinants of individuals’ approach to life and economic success.14. Can we say anything about the relative importance of geography. this does not inform us much about the ultimate sources of diﬀerences in economic performance. at least in theory. investment rates in physical and human capital are found to be important.

the expansion of European overseas empire provides a “natural experiment”. Johnson and Robinson (2002).3 Europe’s Expansion and Colonial Origins of Institutions As discussed above. openness. The geography explanation predicts persistence in income. since the particular historical development of the world economy may have brought about a correlation between institutional development and geography. the role of government. In this context. reﬂecting the importance of other omitted factors. 89 . 4.14. This reasoning is invalid. ﬁnancial development. political instability. since the geographic. Much of the correlation may be no more than just that–spurious correlation. This lack of causality could be important especially when we think about the broad hypotheses outlined above. so one might be tempted to think that the correlation between these variables and income is more likely to spurious. institutions and culture are endogenous. and therefore give more importance to geography. in Acemoglu. a major shock could disrupt persistence. institutions. we looked at the horserace between geography and institutions. and demographics are typically included in these types of empirical analyses and found to be important.451: Introduction to Economic Growth human capital investments must be in turn caused by other factors. Europeans aﬀected the institutions of many societies through their colonization. Although the institutions view also suggests persistence. ecological and climatic factors that should matter are changing only little over periods as long as 500 years. geography. or even create a reversal. In particular. Among these other factors. share of natural resources. The big problem with all this literature is that there is very little attempt to formally establish causality.

while the geography view predicts persistence between 1500 and today among the former European colonies. strongly suggest that there was a reversal in relative rankings across this set of countries. At one extreme. European powers set up “extractive states”. and exploited this theory to derive a possible source of exogenous variation. Johnson and Robinson (2001). with strong emphasis on private property and checks against government power. this observation does not give us a direct estimate of the eﬀect of institutions/social organization on economic performance. This discussion suggests that geographic or climactic diﬀerences across countries are not of ﬁrst-order importance in shaping diﬀerences in income we observe today. The data. Our theory rests on three premises: 1. Primary examples of this include Australia.14. as discussed above. Therefore. New Zealand. it turns out that Europeans introduced worse institutions–in the sense of institutions discouraging investment– in previously prosperous places. To go beyond a simple horserace of geography versus institutions. At the other extreme. Canada. There were diﬀerent types of colonization policies which created diﬀerent sets of institutions. the institutions view suggests the possibility of a reversal.451: Introduction to Economic Growth More strikingly. many Europeans migrated and settled in a number of colonies. exempliﬁed by the Belgian colonization of the Congo. we need a source of exogenous variation in institutions. these geographic factors appear to be less important than other factors. we proposed a theory of institutional diﬀerences among countries colonized by Europeans. and the 90 . In Acemoglu. The settlers in many areas tried to replicate European institutions. These institutions did not introduce much protection for private property. nor did they provide checks and balances against government expropriation. At the very least. and to estimate the impact of institutions on economic performance. Nevertheless.

The ﬁrst shows the cross-sectional relationship between income per capita and a measure of economic institutions. Summarizing this schematically: (potential) settler mortality ⇒ settlements ⇒ early institutions ⇒ current institutions ⇒ current performance The results show a large eﬀect of institutions on income. This is one of many potential variables capturing the institutional features of a country that can be used. 91 . thus intimately related to economic incentives that are highlighted by the institutions approach. extractive policies were more likely. In places where the disease environment was not favorable to European settlement. and generate no evidence that geography matters. Based on these three premises. protection against expropriation risk. The colonial state and institutions.451: Introduction to Economic Growth United States. 2. at least to some extent. persisted.14. The following two ﬁgures summarize most of the ﬁndings. 3. Its advantage is that it is directly about protection of property rights. we use the mortality rates expected by the ﬁrst European settlers in the colonies as an instrument for current institutions in these countries. The colonization strategy was inﬂuenced by the feasibility of settlements.

451: Introduction to Economic Growth 10 ARG HKG MLT BHS CHL VEN URY MEX GAB MYS ZAF CRI COL TTO BRA IDN USA SGP AUS CAN NZL Log GDP per capita. and the third shows the reduced form between income per capita and settler mortality.14. 1995 PAN GTM 8 TUN ECU PER DOM DZA PRY JAM EGYMAR BOL GUY AGO LKA HND NIC CMR GIN CIV COG SEN GHA PAK SDN VNM TGO HTI KEN UGA BGD NGA ZAR BFA MDG NER MLI SLV SLE ETH TZA IND GMB 6 4 4 6 8 Average Expropriation Risk 1985-95 10 The second shows the ﬁrst-stage relationship between log (potential) settler mortality and protection against expropriation risk (so that higher scores correspond to better protection against expropriation by government or elites. 92 . or generally to better property rights protection). The latter two ﬁgures together give the two-stage least squares estimate of the eﬀect of broad economic institutions on long-run income per capita diﬀerences. PPP.

there is always the possibility that the instrument is not excludable).14. and likely due to the institutional channel (but like all instrumental variable strategies.451: Introduction to Economic Growth 10 Average Expropriation Risk 1985-95 NZL AUS USA CAN SGP 8 HKG MYS MLT ZAF IND BRA CHL IDN BHS MEX TTO COL VEN MAR JAM CRI URY PRY EGY ECU DZA TUN VNM ARG DOM LKA KEN SEN PAN PER BOL HND NIC BGD GTM SLV GAB GMB CIV TGO TZA CMR GIN GHA SLE AGO NER COG UGA BFA MDG MLI NGA 6 PAK GUY ETH 4 2 SDN HTI ZAR 4 6 Log of Settler Mortality 8 10 AUS NZL USA SGP HKG CAN MLT CHL BHS BRB ARG VEN URY MEX GAB PAN COL CRI TTO BRA TUN ECU PER DZA DOM BLZ GTM PRY JAM IDN MAR EGY SLV BOL GUY AGO LKA HND NIC CMR GIN CIV MRT SEN COG GHA PAK IND SDN VNM TGO CAF HTI BEN LAO KEN UGA BGD ZAR BFA TCD NERMDG BDI RWA TZA SLE ETH MUS Log GDP per capita. Johnson and Robinson (2001) conduct a variety of checks to show that this relationship is robust. PPP. 1995 MYS ZAF FJI 8 GMB NGA MLI 6 4 2 4 6 Log of Settler Mortality 8 Acemoglu. 93 .

Johnson and Robinson (2001). religion and the identity of colonial power. But throughout. we will look deeper into a range of models in order to understand how diﬀerences in various policies. technologies. But this is not conclusive. which control for proxies of cultural diﬀerences. 94 . suggest that it is political and economic institutions not culture that matter more.451: Introduction to Economic Growth Even taking this evidence at face value. can we distinguish between culture and institutions? Some of the results in Acemoglu. preferences and institutions translate into growth rates and cross-country diﬀerences. Thus the rest of the course will be about the mechanics of economic growth.14. you may want to bear in mind how these mechanics may relate to fundamental causes as discussed in this chapter. For the rest of the course. Future and “smarter” work is needed to make progress in distinguishing between culturebased and institutional explanations.

Part II Neoclassical Growth 95 .

.

97 . we discuss the basic neoclassical approaches to economic growth.451: Introduction to Economic Growth In this part. focusing on models with exogenous technological progress. so we can make meaningful statements about savings rates being endogenous and also think of welfare of consumers. The most important advance over what we have seen so far is that these models explicitly incorporate consumer preferences and consumer behavior.14.

14.451: Introduction to Economic Growth 98 .

These households can be truly inﬁnitely lived. it would be much more satisfactory to specify the preference orderings of individuals as in standard general equilibrium theory and go from there. let us consider an economy consisting of a unit measure of inﬁnitely-lived households. To prepare for this. Then the problem would be one in which each household i has an instantaneous utility function given by ui (ci (t)) where ui : R → R is increasing and concave and ci (t) is the consumption of household i– this means that the individual does not derive any utility from the consumption of other households. Instead.Chapter 5 Towards Neoclassical Growth At this point. so that 99 . we will assume that individuals discount the future “proportionally” (also referred to as “exponentially”). or could consist of overlapping generations with full (or partial) altruism linking generations within the household. Throughout. The entire Solow growth model was predicated on a constant savings rate. let us take a step back. so consumption externalities are ruled out.

What this means is that we will think that the preference side of the economy can be represented as if there were a single consumer making the consumption and saving decisions (and labor supply decisions when these are endogenized). and the same sequence of eﬀective labor endowments {h (t)}∞ t=0 . we can have diﬀerences in households’ income processes.1 Representative Consumer Instead of the more general framework mentioned above.14. we will look at economies that admit a representative consumer. and also the same discount factor β . for each household we could have eﬀective labor ∞ endowments of {hi (t)}∞ t=0 . where β i ∈ (0. the standard approach in macroeconomics and economic growth is to assume the existence of a representative consumer. One way of having a representative consumer is to assume that each household has the same utility function u (ci (t)) where u : R → R is increasing and concave and ci (t) is the consumption of household i. The advantage of this approach is that the economy indeed has a representative 100 . 1) is the discount factor of household i. 5. their preferences at time t = 0 are given by ∞ X t=0 βt i ui (ci (t)) . this problem is too hard. it would be impossible to go beyond that. To avoid the complexities involved in this general formulation. for example. In addition. at this level of generality.451: Introduction to Economic Growth in discrete time and ignoring uncertainty. Even though we may be able to establish some existence results. Unfortunately. thus a sequence of labor income of {w (t) hi (t)}t=0 where w (t) is the equilibrium wage rate per unit of eﬀective labor.

E. 101 . and c (t) is the consumption level of the representative household. For ε > 0. but no normative meaning. let Pε = p ∈RN + :pj /pj 0 ≥ ε for all j and j . Proof. though as the next theorem shows. with the representative consumer assumption in discrete time. but the aggregate behavior can be represented as if it were the outcome of the maximization of a representative consumer. we have that the preference side can be represented as the following maximization problem starting at time t = 0: max ∞ X t=0 β t u (c (t)) . so the representative consumer has a normative meaning as well as a positive meaning.14. any continuous function x : Pε → RN + that satisﬁes Walras’ Law and homogeneity of degree 0 can be an aggregate excess demand function. In other words. Yet alternatively. Then any ε > 0. the representative consumer will have positive meaning. In this case. each with potentially diﬀerent preferences. Proposition 17.3.451: Introduction to Economic Growth consumer. and we can use the same preferences to evaluate aggregate welfare. we could assume that there is heterogeneity among households. where β ∈ (0. Winston and Green (1995). See Debreu (1974) or Mas-Colell. most models do not admit representative consumers: Theorem 8 (Debreu-Mantel-Sonnenschein) Consider an exchange economy with a ﬁnite number N < ∞ of commodities and H < ∞ households. In any case. Let p be the vector of prices and x (p) be the vector of aggregate excess demands © ª 0 for these commodities at the price vector p. we can represent the savings and consumption decisions as if they are coming from a representative consumer. 1) is the common discount factor of all the households. This is an extremely convenient assumption.

when there is a special form of quasi-linearity in the preferences. this result is partly an outcome of very strong income eﬀects.14. then these preferences can be aggregated to be represented by those of a representative consumer. Theorem 9 (Gorman) Consider an economy with a ﬁnite number N < ∞ of commodities and H < ∞ households. Nevertheless. . H . y ) = PH H X i=1 ai (p) + b (p) y. yi ). and is left to you as an exercise. Proof.451: Introduction to Economic Growth Essentially. This is therefore a negative result warning us against the use of models with representative consumers. satisfying the weak axiom of revealed preference. or possessing a negative-semi-deﬁnite Jacobian) for aggregate (market) excess demand functions. Here the following aggregation theorem is particularly useful. In this context. which speciﬁes the household’s (ordinal) utility as a function of the price vector p and the household’s income (wealth) yi . Special but approximately realistic preference functions. where y ≡ i=1 yi is aggregate income. this theorem states that in general. with indirect utility v (p. aggregating them to have representation for a representative consumer is possible. Suppose that the preferences of household i lead to an indirect utility function of the form vi (p. To state this theorem... Therefore. The proof follows from basic micro theory. recall that an indirect utility function for household i is vi (p. the fact that there are optimizing individuals in the background imposes no restriction (such as being downward sloping. enable us to rule out arbitrary aggregate excess demand functions. it is interesting to consider the CRRA (constant relative risk aversion) 102 . as well as restrictions on the distribution of income across individuals.. yi ) = ai (p) + b (p) yi for i = 1.

1) is the discount factor of the households.14. U= P ∞ t ⎩ β ln c ( t ) if θ = 1 t=0 where θ is the coeﬃcient of relative risk aversion and also the inverse of the intertemporal elasticity of substitution. the assumption that the economy admits a representative consumer is not as restrictive as in models in which we wish to analyze growth without making the balanced growth assumption.2) where ρ > 0 is now the discount rate of the individuals. 103 . 5. (5.2 Problem Formulation Let us now make the representative consumer assumption. if we wish to impose balanced growth. In continuous time.451: Introduction to Economic Growth utility function in the inﬁnite-horizon economy ⎧ ⎨ P∞ β t c(t)1−θ −1 if θ 6= 1 and θ ≥ 0 t=0 1−θ . This class of utility functions satisfy the conditions of Theorem 9. Therefore. which regulates how willing individuals are to substitute consumption over time. this utility function becomes Z ∞ exp (−ρt) u (c (t)) dt 0 (5. Suppose that each household’s utility function in discrete time starting at time t = 0 is (ignoring uncertainty) ∞ X t=0 β t u (c (t)) .1) where β ∈ (0. because they are the unique class of utility functions that are consistent with balanced growth. We will see below that CRRA preferences have a special role in models of economic growth.

.451: Introduction to Economic Growth Where does the exponential form of the discounting in (5. imagine we are trying to calculate the value of $1 in T periods. and then took the limits in the numerator and denominator to obtain the second equality.2) come from? At some level. so the link should be apparent. we wish to calculate v (T ) ≡ lim v (T | ∆t) ≡ lim (1 + ∆t · r)T /∆t . Let the interest rate in each subinterval be equal to ∆t · r. Now we want to take the continuous time limit by letting ∆t → 0. Let us next write this as ln (1 + ∆t · r) r/ (1 + ∆t · r) = lim = rT ∆t→0 ∆t→0 ∆t/T 1/T lim where I used l’Hopital’s rule to obtain the ﬁrst equality.e. we would be changing the interest rate. T ] into T /∆t equally-sized subintervals. we can write v (T ) ≡ exp lim ln (1 + ∆t · r) ¸ ∙∆t→0 T ln (1 + ∆t · r) = exp lim ∆t→0 ∆t h T /∆t i However. the term in square brackets has a limit of the form 0/0. i.14. we called discounting in the discrete time case also “exponential”. v (T ) = exp (rT ) . Clearly the value of $1 in T periods at this interest rate is given by v (T | ∆t) ≡ (1 + ∆t · r)T /∆t . It is important that the quantity r is multiplied by ∆t. Therefore. and divide the interval [0. ∆t→0 ∆t→0 Since the limit operator is continuous. otherwise as we vary ∆t. But to see it more precisely. 104 .

which is. so we could not talk of preferences explicitly). equal to the total expenditure of household i. Other inner products and subvectors are deﬁned similarly. 5.3 Welfare Theorems Ultimately. Also denote the vector of endowments across households by ω and the vector of utilities by u. Here I give sketch proofs. and p · xi as the inner product of the vector of prices and the vector of consumption of household i. is worth exp (−rT ) today. $1 in T periods. Winston and Green (1995). denote the vector of prices for a ﬁnite dimensional commodity vector by p. we know that there should be an intimate connection between Pareto optima and competitive equilibria (so far we were not able to exploit these connections. the vector of production across commodities and ﬁrms by q and the vector of consumption across commodities and households by x. To remember these theorems. Recall also that by a competitive economy. Both of these theorems are proved in the most elegant fashion in Debreu’s Theory of Value for ﬁnite commodity spaces. so discounting in continuous time takes the exponential form. we refer to an environment without any externalities and where all commodities are traded competitively (recall that here goods at diﬀerent dates are diﬀerent commodities). and easier versions of the proofs are contained in Mas-Colell. with ρ as the discount rate. and use xi as the vector of consumption of household i. The same reasoning applies to discounting utility.14. by deﬁnition. 105 . Denote the set of households by I . each household denoted by i. we are interested in equilibrium growth. since utility functions were not speciﬁed.451: Introduction to Economic Growth With the same reasoning. Then we have the following two important theorems. But in competitive economies such as those analyzed so far.

3) where yi (p∗ ) is the income of household i at price vector p∗ deﬁned above. x∗ ) exists.451: Introduction to Economic Growth Theorem 10 (First Welfare Theorem) Consider a competitive economy with a ﬁnite number of individuals with preferences satisfying non-satiation and a ﬁnite number of commodities and an endowment vector ω. This would contradict the hypothesis that xi is utility maximizing at the price vector p∗ .. and again by non-satiation ∗ reach higher utility than that given by x∗ i . q∗ .e. xi + ε for ε small enough.. i. household i could choose more of each commodity. xi Âi x∗ i. Then it must be the case that for all households i ∈ I . q∗ . p · xi ≥ yi (p∗ ) (5. xi ºi x∗ i and for at least one i0 ∈ I .e. We have P that i πif = 1 by virtue of being shares. q∗ . x∗ ) is a competitive equilibrium. Then it is Pareto optimal.14.. x∗ ). xi is weakly preferred to x∗ i . Since (p∗ . Suppose to obtain a contradiction that there exists (p. q∗ . the new allocation is strictly preferred to x∗ i . Suppose a competitive equilibrium (p∗ . where π if is the share of proﬁts of ﬁrm f held by household i. 106 . i. it must be the case that for all i ∈ I . Proof.e. x) which Pareto dominates (p∗ . Suppose not. i. ∗ ∗ We know that by non-satiation p · x∗ i = yi (p ). q. x∗ ) being a competitive equilibrium implies that each ∗ household maximizes its utility by choosing x∗ i at the price vector p and income level P yi (p∗ ) ≡ p∗ · ω i + f π if · p∗ · q∗ f . then if p · xi < yi (p ). and q∗ f is the competitive equilibrium production vector of ﬁrm f . First recall that (p∗ .

5) for all feasible qf . i∈I i∈I f which contradicts (5. However.5). Notice that the proof of the ﬁrst welfare theorem only uses the summation of the values of commodities at a given price vector. Moreover. but the fact that the sums above exist is essential for the proof. the competitive equilibrium allocation (p∗ . by the fact that q∗ f is a proﬁt-maximizing vector at prices p .451: Introduction to Economic Growth Moreover.4) p∗ ·xi > X i∈I yi (p∗ ) = = since we have that P X i∈I X i∈I Ã Ã p∗ · ω i + p∗ · ω i + X f π if · p∗ · q∗ f p∗ · q∗ f ! . x∗ ) is not Pareto dominated by any other feasible allocation. we must have Ã ! X X X ωi + xi = qf . and is thus Pareto optimal. we have X i∈I (5.3) and (5. ! X f i ∗ πif = 1. Now summing (5. thus X i∈I X f p∗ · q∗ f ≥ Ã X f p∗ · qf ! p∗ ·xi > X i∈I p∗ · ω i + X f p∗ · qf (5. Naturally. by feasibility of an allocation. it must be that p · xi0 > yi0 (p∗ ) . No convexity assumption is necessary. Consequently. for all feasible qf .4) over I . Finiteness of the number of commodities and number of individuals was suﬃcient to guarantee the existence of the sums. q∗ . for i0 ∈ I . the 107 .14.

they may possibly fail to exist. q∗∗ . there exists an endowment vector ω ∗ . Although this is standard practice. in which case the proof. that is. a capital accumulation. It states that any Pareto optimal allocation can be decentralized as a competitive allocation. saving and consumption path that is Pareto optimal given the preferences of a representative household. Then provided that all production sets and preferences are convex. there is a technical problem here.14. Theorem 11 (Second Welfare Theorem) Consider a Pareto optimal allocation yielding utility vector u∗∗ to households. it is also the more important one. In many ways. Proof. while social welfare maximizing allocations are more straightforward. but with inﬁnite number of commodities. This hyperplane gives relative prices that can decentralize the competitive equilibrium at an appropriately chosen endowment vector. This is especially useful in dynamic models where sometimes competitive equilibria can be quite diﬃcult to characterize or even to specify. and even perhaps the First Welfare Theorem. x∗∗ ) is a point of tangency between the aggregate production possibilities set and the aggregate preference set. 108 . q∗∗ . (idea) Given convexity of preference and production sets. Motivated by this. x∗∗ ) yields exactly the utility vector u∗∗ . This motivates many macroeconomists to look for the set of Pareto optimal allocations instead of explicitly characterizing competitive equilibria. Then by the standard separating hyperplane theorem.451: Introduction to Economic Growth sums may exist under other conditions. since the classical welfare theorem apply when there are ﬁnite number of commodities. such that the resulting competitive equilibrium (p∗∗ . we could start by looking at optimal growth. The second welfare theorem is the harder theorem because of the convexity requirement. there exists a hyperplane separating these two sets. (p∗∗ . both the which are convex. may not apply.

let us suppose that the Second Welfare Theorem applies in this environment.4 Optimal Growth in Discrete Time Let us continue to consider an economy characterized by an aggregate production function. Given this. The problem arises for the First Welfare Theorem because of the issue of existence of sums as discussed above. k (t) ≥ 0 and given k (0) > 0. This is the optimal growth approach. For now. and a representative consumer (household).6) . in this optimal growth problem. having an inﬁnite-dimensional commodity space does not create a problem for the Second Welfare Theorem. we can start on the analysis of economic growth at optimizing agents by looking at the social planner’s choice of an allocation that maximizes the representative household’s lifetime discounted utility.14. In fact. The optimal growth problem in discrete time with no population growth or technological progress can be written as follows: max ∞ X t=0 {c(t).451: Introduction to Economic Growth whereas in growth models there is an inﬁnite number of commodities. as long as convexity continues to hold. and the exact conditions will be discussed below. 5. the social planner chooses an entire sequence of consumption levels and capital stocks in order to maximize the discounted sum 109 (5. The welfare theorems can be extended to inﬁnite number of commodities under certain circumstances. In other words.k(t)}∞ t=0 β t u (c (t)) subject to k (t + 1) = f (k (t)) − c (t) + (1 − δ ) k (t) .

meaning that it links tomorrow’s assets to today’s assets. Even if our purpose were not to characterize the Pareto optimal allocations. but this gives a single initial condition.7) is the ﬂow budget constraint. this will come from of the optimality of a dynamic plan in the form of a transversality condition. In particular. (5. The constraint (5. This can be ensured by imposing a lifetime budget constraint. The constraint. This maximization problem can be solved in a number of diﬀerent ways. Here we need an additional condition so that this ﬂow budget constraint eventually converges (i. so we need to augment it with another condition as we will see later. for example. (5.6) embeds the capital accumulation equation together with the production function.. 110 . Instead. each household would be solving the following problem: max ∞ X t=0 {c(t).e. where a (t) denotes the assets of the household at time t and r (t) is the rate of return on assets and w (t) is wage in come. so that a (t) should not go to negative inﬁnity). But the most convenient and common way of approaching it is by using dynamic programming. by setting up an inﬁnite dimensional Lagrangian. We will see later that we need another boundary condition but not in the form of an initial condition. but to ﬁnd equilibrium. we would have to solve a problem similar to this. We have also speciﬁed that the initial level of capital stock is k (0).k(t)}∞ t=0 β t u (c (t)) subject to a (t + 1) = r (t) a (t) − c (t) + w (t) .7) given a (0).451: Introduction to Economic Growth of the utility of the representative consumer.14. but the ﬂow budget constraint is often more convenient to work with.

we have [c(t).8) k (t) ≥ 0 and given k (0).451: Introduction to Economic Growth 5. this problem lacks one boundary condition which will come from the transversality condition.14.k(t)]t=0 max∞ Z ∞ 0 exp (−ρt) u (c (t)) dt subject to ˙ (t) = f (k (t)) − c (t) − δk (t) k (5. 111 . The most convenient way of characterizing the solution to this problem is via optimal control. We next discuss dynamic programming and optimal control brieﬂy. Once again.5 Optimal Growth in Continuous Time The formulation of the optimal growth problem in continuous time is very similar. In particular.

451: Introduction to Economic Growth 112 .14.

which are not essential for the purposes of this course.Chapter 6 Dynamic Programming and Optimal Growth Here I provide a very brief overview of inﬁnite horizon optimization in discrete time. I also include some technical details. in particular of stationary dynamic programming. but may be useful for those of you who want to understand some of the tools better. 113 .

Here F is the payoﬀ function. depending on xt . It only depends on xt and xt+1 . we will have K = 1.451: Introduction to Economic Growth 6. xt+1 will also directly become the state variable in the next time period. xt+1 ) subject to xt+1 ∈ Γ(xt ). so that xt ∈ R.1 Brief Review of Dynamic Programming Using abstract but simple notation. In many economic applications. since there is no guarantee that the maximal value is attained by any feasible plan. and xt+1 . In this simple formulation. which corresponds to the control variable. Notice that this problem is stationary in the sense that the payoﬀ function F is not time-dependent. the canonical dynamic optimization program in discrete time can be written as Problem A1 ∗ : sup {xt+1 }∞ t=0 ∞ X t=0 v (x0 ) = β t F (xt . Here I used “sup ” rather than max. The constraint on the problem is written as xt+1 ∈ Γ(xt ) where Γ:X ⇒X is a correspondence determining what type of xt+1 is allowed given the state variable xt .14. 114 . which is the state variable. for all t ≥ 0 where xt ∈ X ⊂ RK for some K ≥ 1. x0 given.

meaning the value of pursuing the optimal strategy starting with initial condition x0 . which can be thought of as the value function.14.. k (t) ≥ 0 and given k (0). To map this problem into the form here. let us recall the optimal growth problem from above: max ∞ ∞ X t=0 {c(t). then because there is no discounted structure. If instead we had a much more general problem. sup {xt+1 }∞ t=0 F (x0 . Then use the constraint to write: c (t) = f (k (t)) − k (t + 1) + (1 − δ ) k (t) . in the sense that the original plan that maximizes the initial objective function is not necessarily what an individual would like to stick to if he or she is carrying out the optimization period by period. let xt = k (t) and xt+1 = k (t + 1). For concreteness. Moreover. In many ways.451: Introduction to Economic Growth Of particular importance is the function v ∗ (x0 ).. This is the class of problems in which dynamic programming will be most useful. for example. . dynamic programming could not be used (at least in its simplest form). it is also the essence of dynamic programming. it can be noted that problems that do not have an exponential discounted structure pose another problem for us: they are not time-consistent.). Time consistency is both a very natural property and one that makes the mathematical analysis much simpler.k(t)}t=0 β t u (c (t)) subject to k (t + 1) = f (k (t)) − c (t) + (1 − δ ) k (t) . 115 . x1 . Notice also that I have already simpliﬁed life by writing the objective function as a discounted sum.

Problem A1. the relevant functional equation can be written as Problem A2 : y ∈Γ(x) v (x) = sup [F (x. where L∞ is the vector space of inﬁnite sequences that are bounded with the k·k∞ norm.k(t)}∞ t=0 max ∞ X t=0 β t u (f (k (t)) − k (t + 1) + (1 − δ) k (t)) subject to k (t) ≥ 0 (which is the simplest form of a constraint correspondence Γ). Suppose Problem A1 has a maximum with optimal sequence denoted by {x∗ t }t=0 starting with x0 Then by deﬁnition.. Such problems sometimes have nice features. for all x ∈ X. y ) + βv(y )] .e. It is also often easier to characterize analytically or numerically. x∗ 1) +β + ∗ β j F (x∗ j +1 . v (x0 ) = = = ∗ ∞ X t=0 ∗ β t F (x∗ t .451: Introduction to Economic Growth and substitute into the objective function to obtain: {c(t). xt+1 ) ∞ X F (x0 . similar to the logic of comparing today to tomorrow. x∗ 1) F (x0 . but often are diﬃcult to characterize both analytically and numerically. xj +2 ) j =0 βv ∗ (x∗ 1) 116 . The basic idea of dynamic programming is to turn the sequence problem into a functional equation. {xt }t=0 ∈ L . (6. also referred to as the sequence problem. is one of choosing an inﬁnite ∞ ∞ sequence {xt }∞ t=0 from some (vector) space of inﬁnite sequences (for example. This often gives better economic insights.1) In fact. one of ﬁnding a function rather than a sequence. In this particular case. which I will denote throughout by the simpler notation k·k). this form of the problem suggests itself naturally from the formulation Problem ∞ A1. i.14.

Therefore. Dynamic programming exploits this principle and provides us with a set of powerful tools to analyze optimization in discrete time inﬁnite horizon problems. ∞ instead of ﬁnding a sequence {xt }∞ t=0 ∈ L .1) as well. Problem A2 is commonly referred to as the Bellman equation. Essentially. after Richard Bellman. as the notation makes it clear with the sup (or max) deﬁning the function. including the contraction mapping theorem were earlier used by Lloyd Shapley in his work on stochastic games). g : X → X determining what value of xt+1 to choose for a given value of the state variable xt . and the optimal continuation path. this is often referred to as the recursive formulation. an optimal plan can be broken into two parts. Second. we will try to ﬁnd a function v . but I will return to some of these issues below. [In general. more formally. what is optimal to do today. in the sense that it is on the right hand side of (6. Part of the theory of dynamic programming is about specifying the conditions under which Problems A1 and A2 are equivalent. First. who introduced dynamic programming to operations research and engineering applications (though identical tools and reasonings. These are not central for us to focus upon here. because the function v is deﬁned recursively. What makes this formulation useful is that the solution will often be a time invariant policy function. a control reaching the 117 . v (x) is a function.14. A couple of points are immediately worth noting. there are two complications: ﬁrst. that satisﬁes (6. v:X→R Diﬀerently from other maximization problems. here maximization itself deﬁnes the function v.1).451: Introduction to Economic Growth This equation encapsulates the basic idea of dynamic programming: the principle of optimality.

This equation simply follows from the fact that g (x) is the optimal policy. but a policy correspondence g : X ⇒ X . but also some of its properties. These are not essential for understanding the application of these tools to economic growth models. and assume that g (·) is single valued.2. we may not have a policy function. thus a function–conditions to guarantee this are provided below]. once the value function v is determined. the policy function is given straightforwardly. ρ) be a metric space and T : S → S be an operator mapping S into 118 . 6.2 6. second. which is one way of determining the policy function. as we will see. so reaches the maximal value v (x). for all x ∈ X. ρ) is a metric space. Let me avoid these complications for now. The usefulness of the recursive formulation as in (6. because there may be more than one maximizers for a given state variable. if S is a space and ρ is a metric deﬁned over this space with the usual properties (loosely corresponding to “distance” between elements of S ). Moreover. which was the reason why we originally used the notation sup. Deﬁnition 6 Let (S. In particular.1) comes from the fact that there are some powerful tools which not only establish existence of the solution.451: Introduction to Economic Growth optimal value may not exist. g (x)) + βv (g (x))] .14.1 Digression: Technical Details Contraction Mappings We say that (S. but they are useful for working with these tools in general and with growth models in particular. by deﬁnition it must be the case that v (x) = [F (x.

Then T : S → S is a contraction if for some β ∈ (0. T y ) ≤ βρ(x. This implies that ρ(ν 2 . 2. 1).14. In other words. for all x. Now take ν 0 ∈ S . such that ν n+1 = T ν n so that ν n = T nν 0. . S = [a. 1).. |T x − T y | ≤ β < 1. Recall also that a metric space (S. i. b]. Deﬁnition 7 A ﬁxed point of T is any element of S satisfying T x = x. |x − y | all x. Then there exists a unique v ˆ ∈ S such that Tν ˆ=ν ˆ. ν 0 ).e. ρ) is complete if every Cauchy sequence in S converges to an element in S . and a sequence {ν n }∞ n=0 with each element in S . Note T n x = T (T n−1 x) for any n = 1.. y ∈ S. and T : S → S be a contraction. a unique ﬁxed point.451: Introduction to Economic Growth itself. For example. ρ) be a complete metric space. y ) = |x − y |. T ν 0 ) ≤ βρ(ν 1 .. y ∈ S with x 6= y. ν 1 ) = ρ(T ν 1 . 119 . T is a contraction mapping (with modulus β ) if for some β ∈ (0. with usual metric of this space ρ(x.. Proof. ρ(T x. Theorem 12 (Contraction Mapping Theorem) Let (S. y ). a contraction mapping brings elements of the space S “closer” to each other. let us take a simple interval of the real line as our space.

ν n+1 ) + ρ(ν n+1 . ≤ βρ(ˆ ν . + β n+1 + β n ρ(ν 1 . Since S is complete. and the second line the deﬁnition of the contraction. ν 0 ). 2. Moreover. ν n ) ≤ β n ρ(ν 1 . ν 0 ) ¤ £ = β n β m−n−1 + . ν ˆ) = 0. (6... m → ∞. ν m and ν n are getting closer... by induction. so {ν n }∞ n=0 is a Cauchy sequence. 120 n = 1. ν 0 ). ν n ) ¤ £ ≤ β m−1 + . ρ(ν m . T n ν 0 ) + ρ(T n ν 0 . .14. + β + 1 ρ(ν 1 .451: Introduction to Economic Growth where the last inequality uses the contraction property of T .2) . which implies that ρ(T ν ˆ. ν ˆ) ≤ ρ(T ν ˆ. T n−1 ν 0 ) + ρ(T n ν 0 . ≤ 1−β where the ﬁrst line uses the triangle inequality (which is true by deﬁnition for any metric). we have ρ(T ν ˆ. + ρ(ν n+2 . and the second line uses (6.. thus a ﬁxed point exists. ν ˆ) ˆ).2). ν 0 ) βn ρ(ν 1 . Now note that for any ν 0 ∈ S and any n ∈ N. for any m > n. The above argument shows that both of the terms on the right tend to zero as n → ∞. ν n ) ≤ ρ(ν m . establishing that T ν ˆ=ν ˆ.. Hence. The last line implies that as n... we have ρ(ν n+1 . ν where the ﬁrst line again uses the triangle inequality. this establishes that νn → ν ˆ ∈ S. ν m−1 ) + .

ρ) is a complete metric space and S 0 is a closed subset of S . such that T ν = ν and T ν ˆ=ν ˆ with ν ˆ 6= ν . ρ) be a complete metric space. proving the ﬁrst claim in the theorem. T ν The second part of this theorem is very important to prove results such as strict concavity or that a function is strictly increasing. proving the second part. if ˆ ∈ S 00 . Since β < 1. then ν Proof. ˆ ∈ S 00 . Before doing this. ν ˆ ∈ S 0 . Since S 0 is this sequence is in S 0 by the fact that T (S 0 ) ⊆ S 0 . T n ν 0 → ν closed. then by virtue of the fact that ν ˆ ∈ S 0. greatly facilitating the analysis of such dynamic models. so ν ˆ ∈ S 00 . ρ) is also a complete metric space.451: Introduction to Economic Growth Uniqueness is proved by contradiction. Theorem 13 Let (S. Take an arbitrary ν 0 ∈ S 0 . so in particular to the space of functions. 121 . and consider the sequence {T n ν 0 }∞ n=0 . let us consider another useful result. ν ) = βa. Each element of ˆ from Theorem 12. The use of the contraction mapping theorem is that it can be applied to any metric space. First. We will see below that this is often straightforward. ν ∈ S .1) deﬁnes a contraction mapping.1) will establish the existence of a unique value function v . This is because the set of strictly concave functions or the strictly increasing functions are not closed. Suppose that there exist ν ˆ. T : S → S be a contraction mapping with T ν ˆ = ν ˆ. If in addition we have that T (S 0 ) ⊆ S 00 . then ν ˆ ∈ S 0 . Applying it to equation (6. for this we have to prove that the recursion in (6. T ν ) ≤ βρ(ˆ ν . this yields a contradiction. Moreover. ν ) = ρ(T ν ˆ.14. proving uniqueness. The second part of the theorem enables us to avoid this complication. Naturally. This implies 0 < a = ρ (ˆ ν . T (S 0 ) ⊆ S 00 ⊆ S 0 . recall that if (S. and T (S 0 ) ⊆ S 0 . then (S 0 . If S 0 is a closed subset of S .

Applying the same argument in reverse establishes T g ≤ T f + β kf − g k . By deﬁnition for any f. 2. deﬁned on X Suppose that T : B (X ) → B (X ) is an operator satisfying the following two conditions: 1. especially in the context of dynamic programming. the following theorem is useful. we have T f ≤ T (g + kf − g k) ≤ T g + β kf − g k . Then. (monotonicity) For any f. Then: Theorem 14 (Blackwell’s suﬃcient conditions for a contraction) Let X ⊆ RK . Proof. Let us use the notation (f + a)(x) = f (x) + a for some a ∈ R. a ≥ 0. f ≤ g + kf − gk . where again k·k is the sup norm. for all f ∈ B (X ). g ∈ B (X ) and f (x) ≤ g (x) for all x ∈ X implies (T f )(x) ≤ (T g )(x) for all x ∈ X . Let f ≤ g stand for f (x) ≤ g (x) for all x ∈ X . T is a contraction with modulus β . and B (X ) be the space of bounded functions f : X → R. 122 . (discounting) There exists β ∈ (0. g ∈ B (X ). x ∈ X. 1) such that [T (f + a)](x) ≤ (T f )(x) + βa. Now applying the operator T on both sides.451: Introduction to Economic Growth How do we check that a mapping is a contraction? Here.14. where the ﬁrst inequality uses monotonicity and the second discounting.

xt+1 ) exists.14. proving that T is a contraction.3) is a solution to Problem A1. under some boundedness conditions.. t = 0. outlined at the beginning. . ..2 Application of Contraction Mappings to Dynamic Programming Let us now apply the above tools to the problem of dynamic programming..451: Introduction to Economic Growth Combining these two inequalities yields kT f − T gk ≤ β kf − g k . Consider a sequence {xt+1 }∞ t=0 which attains the supremum of Problem A1. We will now show that this sequence will satisfy the recursive equation of dynamic programming v(xt ) = F (xt . 2. and for all x0 ∈ X and x ∈ Π(x0 ).) ∈ Π(x0 ). we will establish some equivalence results between the solutions to Problem A1 and Problem A2.. and assume: Assumption 3 Γ (x) is nonempty for all x ∈ X . for all t = 0. any sequence that is a solution to (6. 1. To prepare for these results. x1 . in the sense that it attains its supremum. 123 . 1. (6. let us deﬁne the set of feasible sequences or plans starting with initial value x0 : Π(x0 ) = {{xt+1 }∞ t=0 : xt+1 ∈ Γ(xt )..}. 6. xt+1 ) + βv (xt+1 ). . P t limn→∞ n t=0 β F (xt . In other words..2.. Let us denote a typical element of the set by x = (x0 .3) and moreover.

y ) + βv ∗ (y ) + ε. as: (R Thus v∗ (x0 ) is the supremum in Problem A1 (i..14. If |v ∗ (x0 )| < ∞.4) 2. where R ¯ is the extended real line Next deﬁne the supremum function v∗ : X → R v∗ (x0 ) = sup u(x). then u(x) = −∞. for all x ∈ Π(x0 ). (6. then v∗ (x0 ) ≥ F (x0 .e. if |v∗ (x0 )| < ∞. then there exists a sequence {xk } in Π(x0 ) such that limk→∞ u(xk ) = +∞. v∗ (x0 ) ≤ F (x0 . if v ∗ (x0 ) = +∞. v∗ (x0 ) ≤ u(x) + ε.451: Introduction to Economic Growth ¯ . x∈Π(x0 ) ¯ = R ∪ {+∞. Conversely. and for any ε > 0. Note that it follows by deﬁnition that v ∗ is the unique function satisfying the following three conditions for Problem A1.. if the following three conditions for FE hold: 1. and 3.3)). SP: 1. then v ∗ (x0 ) ≥ u(x). (6. some x ∈ Π(x0 ). and for any ε > 0.6) . −∞}). y ) + βv ∗ (y ). the value of the program in Problem A1). we will say that v ∗ is a solution to Problem A2 (and thus satisﬁes the functional equation (6. or the sequence problem. if v∗ (x0 ) = −∞. (6.5) all x ∈ Π(x0 ). (6.7) all y ∈ Γ(x0 ). 124 some y ∈ Γ(x0 ).

x1 ) + βu(x0 ) with x0 = (x1 . and β satisfy Assumption 3. x2 . then F (x0 . x1 . Theorem 15 Let X. if v∗ (x0 ) = −∞. Γ. F.9) Then for any x0 ∈ X and any lim n X t=0 β t F (xt .. .451: Introduction to Economic Growth 2.8) 3. y k ) + βv ∗ (y k ) = +∞. xt+1 ) n X t=0 = F (x0 .). x1 ) + β lim n→∞ β t F (xt+1 . y ) + βv∗ (y ) = −∞. if v∗ (x0 ) = +∞. F. and β satisfy Assumption 3. (6. for any x0 ∈ X and any x ∈ Π(x0 ). Proof. Γ. It therefore formalizes the principle of optimality introduced more informally above. x1 ) + βu(x0 ).14. x = (x0 . Under Assumption 3. then there exists a sequence {y k } in Γ(x0 ) such that k→∞ ¤ £ lim F (x0 . u(x) = n→∞ all y ∈ Γ(x0 ).. u(x) = F (x0 .. Then the function v ∗ is a solution to Problem A2. . xt+2 ) = F (x0 . We now have the following simple lemma: Lemma 1 Let X. the current return and continuation value. (6.) ∈ Π(x0 ).. This lemma basically says that the utility from any feasible plan can be decomposed into two parts. 125 .

this establishes FE (6. If v ∗ (x0 ) = +∞. x1 . x1 ) + βu(x0 ) + ε.8) holds for the sequence {y k = xk 1 } in Γ(x0 ).). and 0k k ∗ k u(xk ) = F (x0 . x1 ) + βv ∗ (x1 ) − βε for any ε > 0. x1 .. xk 1 ) + βu(x ) ≤ F (x0 .. 126 all (x0 .. establishing FE (6. x2 . . and choose x0 ∈ X. If β = 0. x2 .. and it is suﬃcient to show that this implies that the FE conditions (6.5) there exists x0 = (x1 . ∗ it follows that FE (6.. Then SP conditions (6.4) that v∗ (x0 ) ≤ F (x0 . then there exists a sequence {xk } in Π(x0 ) such that lim u(xk ) = +∞. Suppose v ∗ (x0 ) is ﬁnite. Hence it follows from SP (6.451: Introduction to Economic Growth Proof. x2 . From SP (6. To establish FE (6. Since x1 ∈ Γ(x0 ). it follows that one can choose x = (x0 .7) hold. x1 ) + βv∗ (x1 ) + ε. . It then follows from SP (6. Then by SP (6. x1 ) + βu(x0 ) ≥ F (x0 . where x0 = (x1 .7). If v (x0 ) = −∞. Suppose that β > 0.) ∈ Π(x1 ) such that u(x0 ) ≥ v∗ (x1 ) − ε.7). . x1 ) + βv (x1 ).) ∈ Π(x0 ). ..14.. all k.5) hold.6).4) and Lemma 1 that v∗ (x0 ) ≥ u(x) = F (x0 . then u(x) = F (x0 .6) and (6.5) and Lemma 1. .4) and (6..) = x ∈ Π(x0 ). x1 ) + βu(x0 ) = −∞.) ∈ Π(x0 ).. the result is trivial. Note also that x = (x0 . x1 .. k→∞ all k. To establish (6. let x1 ∈ Γ(x0 ) and ε > 0 be given. choose x0 ∈ X and ε > 0. . Since xk 1 ∈ Γ(x0 ). so that v∗ (x0 ) ≤ u(x) + ε = F (x0 . x2 .6).

all x1 ∈ Γ(x0 ).. 127 . x1 ) + βF (x1 . all x0 ∈ Π(x1 ).6) and (6.. then (6. x2 . ≥ un (x) + β n+1 v(xn+1 ). . x1 ) + βv (x1 ) ≥ F (x0 . all x0 ∈ X. Proof. (6.451: Introduction to Economic Growth where x0 = (x1 . Hence v satisﬁes (6. F.10) implies that v cannot take on the values +∞ or −∞. Hence v ∗ (x1 ) = −∞..10). . and β satisfy Assumption 3. x2 ) + β 2 v (x2 ) . x1 .) ∈ Π(x0 ).. we have the following converse to this theorem: Theorem 16 Let X. Γ.7). Now taking the limit as n → ∞ and using the convergence property from (6. Since v is the solution to Problem A2. all (x0 .4) and (6.5).9) follows immediately. Since F is real-valued (thus does not take the values −∞ or +∞).4) for any x ∈ Π (x0 ). satisﬁes n→∞ If v is a solution to (FE) and lim β n v(xn ) = 0. (sketch) Condition (6.10) then v = v ∗ . .6) implies that for all x0 ∈ X and x ∈ Π(x0 ) v(x0 ) ≥ F (x0 . and it is suﬃcient to show that this implies v satisﬁes (6. (6. Under the additional boundedness condition.14. all x1 ∈ Γ(x0 ). . we obtain (6.). Since F is real-valued and β > 0. it follows that u(x0 ) = −∞.

F. we can make a lot of progress by studying solutions to Problem A2.11) 128 . (6. + β n δn+1 ) ≤ un (x) + β n+1 v(xn+1 ) + ε/2. we have v (x0 ) ≤ un (x) + β n+1 v(xn+1 ) + (δ 1 + βδ2 + .. t = 0. Then ∗ ∗ ∗ ∗ v∗ (x∗ t ) = F (xt . xt+1 ) + βv(xt+1 ) + δ t+1 . .. An important implication is that although Problem A2 may have many solutions. and β satisfy Assumption 3. completing the proof. In general. we obtain that for any x = (x0 . 1. xt+1 ) + βv (xt+1 ).14. but sometimes we need to impose (6. n = 1.10). . Since (6. . Naturally. choose an arbitrary sequence {δ n }∞ n=1 in R+ such P∞ n−1 that n=1 β δ n ≤ ε/2... 2..10) in order to pick the right solution (this is similar to sometimes working with necessary conditions for optimization. x1 . For this we have: Theorem 17 Let X. v(x0 ) ≤ u(x) + ε. 2.7) holds. Since (6.451: Introduction to Economic Growth Now for a given x0 ∈ X and ε > 0. our interest is mainly with optimal plans.) ∈ Π(x0 ). we can choose xt+1 ∈ Γ(xt ) so that v (xt ) ≤ F (xt . it follows that as n → ∞.. though of course then we need to impose the suﬃciency conditions). only one of those will satisfy the convergence condition (6. Let x∗ ∈ Π(x0 ) be a feasible plan that attains the supremum in Problem A1 starting with initial state x0 ..10) implies that for n suﬃciently large the second term is also less than ε/2. Γ. Using these inequalities. x2 ..

Then it follows by induction on (6. Therefore u(x∗0 ) = v(x∗ 1 ). Proof.13) Then x∗ attains the supremum in Problem A1 for initial state x0 .11) for t = 0. The above theorems are useful in showing the equivalence of Problem A1 and Problem A2.. Γ. Finally. all x ∈ Π(x∗ 1 ). F. . the reverse inequality holds.451: Introduction to Economic Growth Proof. ∗ ∗ ∗0 v∗ (x∗ 0 ) = u(x ) = F (x0 . Now the usefulness of the dynamic programming formulation in Problem A2..13).) ∈ Π(x0 ).. x3 . and with lim sup β t v∗ (x∗ t ) ≤ 0.13).) ∈ Π(x1 ) implies that (x0 .11). so that u(x∗0 ) ≥ u(x0 ). x1 ) + βu(x ) ≥ u(x) = F (x0 . and β satisfy Assumption 3.14. we ﬁnd that v∗ (x0 ) ≤ u(x∗ ).12) ∗ ∗ ∗ Now choose x1 = x∗ 1 . . x3 . and hence 129 .12) yields (6. Let x∗ ∈ Π(x0 ) be a feasible plan from x0 satisfying (6.11) and (6. x1 ) + βu(x0 ).11) that v ∗ (x0 ) = un (x∗ ) + β n+1 v∗ (x∗ n+1 ).. x2 . Continuing by induction establishes (6. establishing the result. x2 . t→∞ (6. 2. x1 . Substituting this into (6. the converse to this theorem is: Theorem 18 Let X . Then using (6. (6. n = 1. all x ∈ Π(x0 ). Since x∗ attains the supremum. ...12) still holds. Since x∗ ∈ Π(x0 ). (6. Since (x1 .11) for all t. Suppose that x∗ ∈ Π(x0 ) satisﬁes (6.

14) where β < 1. v = T v. now deﬁne the operator T such that (T f )(x) = max [F (x. y ) ∈ A.451: Introduction to Economic Growth of the contraction mapping theorem. y ∈Γ(x) (6. there exists some B < ∞. As before. the sup norm.15) A ﬁxed point of this operator. Γ is nonempty. So for this purpose.14. Moreover. but greatly simpliﬁes the analysis. y ) + βf (y )]. comes from the fact that its solution is often easy to characterize. We now make an additional assumption. Assumption 4 X is a compact subset of RK . 130 . y ) ∈ X × X : y ∈ Γ(x)} and F : A → R be bounded and continuous. kf k = supx∈X |f (x)|. all x ∈ X . such that |F (x. X is the possible set of values for the state variable and Γ : X ⇒ X is the correspondence describing the constraints on the problem. y )| < B for all (x. to see the usefulness of the contraction mapping theorem. Consequently. which is not necessary. since F is bounded over its eﬀective domain. This immediately implies that |v ∗ (x)| ≤ B/(1 − β ). with the natural norm on this space. establishing the desired results. Most importantly. will be a solution to (6. we can focus our attention on value functions in the space C (X ) of continuous bounded functions deﬁned on X . y ) + βv(y )] .14). y ∈Γ(x) (6. compact-valued and continuous. let A = {(x. Then we can derive the policy functions from the value function. The importance of Assumption 4 is that it will allow us to focus on the space of bounded functions. take the following version of the dynamic programming problem (Problem A2) v(x) = max [F (x. In particular.

Consequently. F. and let v be the unique solution to (6. We can next see how Theorem 13 enables us to establish more properties of the value function and the policy correspondence. with the sup norm. and Γ is monotone in the sense that x ≤ x0 implies Γ(x) ⊆ Γ(x0 ). Theorem 20 Let X. a unique v ∈ C (X ) satisfying (6. T : C (X ) → C (X ).15) is one of maximizing a bounded function over a compact set. and β satisfy Assumption 4 and let C (X ) be the space of bounded continuous functions f : X → R.. Γ. Γ.14) exists. Then the operator T maps C (X ) into itself. applying Theorem 12. i. Under the assumptions of Theorem 19. Therefore. v ∈ C (X ) satisfying (6. and β satisfy Assumptions 4 and 5. and has a unique ﬁxed point.14). Proof. In particular. G is compact valued and upper hemi-continuous. it has a solution. Formulated in this way. F.e. it is immediate that T is a contraction. y ) + βv (y )} .451: Introduction to Economic Growth Theorem 19 Let X. Then v is strictly increasing. (6. y ) is strictly increasing in each of its ﬁrst K arguments. Since the maximization problem on the right hand side of (6. T is well deﬁned and is easily seen to satisfy the suﬃcient conditions for a contraction in Theorem 14. let us assume Assumption 5 For each y . F (·.14).14. Proof. This follows immediately from Berge’s maximum theorem. Corollary 4 Let G : X → X deﬁned as G(x) = {y ∈ Γ(x) : v(x) = F (x. for example.16) be the policy function (correspondence). 131 .

Furthermore. continuous. x0 ∈ X. and let v satisfy (6. and β satisfy Assumptions 4. single-valued and all θ ∈ (0. F [θ(x. y 0 ) ∈ A. y 0 )] ≥ θF (x. all (x. 5 and 6. The proof again follows from Theorem 13. and let C 00 (X ) ⊂ C 0 (X ) be the set of strictly increasing functions. Let C 0 (X ) ⊂ C (X ) be the set of bounded. Assumption 5 immediately implies that for any nondecreasing f ..16).451: Introduction to Economic Growth Proof. y 0 ). let us impose Assumption 6 F is strictly concave.14. . y ∈ Γ(x) and y 0 ∈ Γ(x0 ) implies θy + (1 − θ)y 0 ∈ Γ[θx + (1 − θ)x0 ]. F. Γ is convex in the sense that for any 0 ≤ θ ≤ 1. the inequality is strict if x 6= x0 . Theorem 21 Let X. This assumption imposes enough concavity on the problem. and let G satisfy (6. Moreover. y ) + (1 − θ)(x0 . 1). and x. Let C 0 (X ) ⊂ C (X ) be the set of bounded. in particular. (x0 . it is suﬃcient to show that T [C 0 (X )] ⊆ C 00 (X ). nondecreasing functions on X .e. by Theorem 13. establishing the result. T f is increasing. Proof. y ). function. Since C 0 (X ) is a closed subset of the complete metric space C (X ). Γ. and let C 00 (X ) ⊂ C 0 (X ) be the set of strictly 132 Then v is strictly concave and G is a continuous. In addition. i.14). y ) + (1 − θ)F (x0 . continuous. (weakly) concave functions on X . it rules out “increasing returns” of any form.

5. Assumption 7 F is continuously diﬀerentiable on the interior of its domain A. To see this. θ ∈ (0. Hence G is a single-valued function.15) and the fact that yθ ∈ Γ(xθ ). by also assuming diﬀerentiability. Let yi ∈ Γ(xi ) attain (T f )(xi ). Γ(x) is convex. T [C 0 (X )] ⊆ C 00 (X ) would establish the results. Since. so that the unique ﬁxed point v is strictly concave. so that (T f )(xθ ) ≥ F (xθ . and xθ = θx0 + (1 − θ)x1 .451: Introduction to Economic Growth concave functions. Since C 0 (X ) is a closed subset of the complete metric space C (X ). let f ∈ C 0 (X ) and let x0 6= x1 . 6 and 7.16). Suppose also that x0 ∈IntX and G (x0 ) ∈IntΓ (x0 ).15) is attained at a unique y value. Furthermore. from Assumption 6. F is also concave and for each x ∈ X . y0 ) + βf (y0 )] + (1 − θ)[F (x1 . y1 ) + βf (y1 )] = θ(T f )(x0 ) + (1 − θ)(T f )(x1 ). by Theorem 13. the second line uses the hypothesis that f is concave and the concavity restriction on F from Assumption 6.14. then v is continuously diﬀerentiable at x0 . 1. 133 . we can also prove that the value function is diﬀerentiable. Theorem 22 Let X. Then Assumption 6 implies that yθ = θy0 + (1 − θ)y1 ∈ Γ(xθ ). for i = 0. and β satisfy Assumptions 4. let v satisfy (6. F. they establish T [C 0 (X )] ⊆ C 00 (X ). and its continuity follows from the fact that it is upper hemi-continuous. Since these relationships are true for any f ∈ C 0 (X ). it follows that the maximum in (6. 1). Γ. where the ﬁrst line is a simple implication of (6. yθ ) + βf (yθ ) > θ [F (x0 . Finally.14) and G satisfy (6.

17). for any subgradient p of −v at x0 must satisfy p · (x − x0 ) ≥ v (x) − v(x0 ) ≥ W (x) − W (x0 ). it follows that G(x0 ) ∈IntΓ(x). For this note that v (·) is concave.451: Introduction to Economic Growth Proof. where the ﬁrst inequality uses the deﬁnition of a subgradient and the second uses the fact that W (x) ≤ v(x). G is a function (i. and again by a standard result in convex analysis. it follows that W (x) ≤ max [F (x.17) with equality at x0 . 134 . for all x ∈ D. Since W is diﬀerentiable at x0 . is diﬀerentiable as desired. Moreover. Since F is concave (Assumption 6) and diﬀerentiable (Assumption 7). y∈Γ(x) for all x ∈ D (6. thus v (·). From Theorem 21. Deﬁne W (·) on D by W (x) = F [x. Moreover. since G(x0 ) ∈ Γ(x) for all x ∈ D. it possesses subgradients.. we show that (6. Now. any convex function with a unique subgradient at an interior point x0 is diﬀerentiable at x0 . This establishes that −v (·). p is unique. and by a standard result in convex analysis.14. Moreover.e. single valued). with equality at x0 as established in (6. since G(x0 ) ∈IntΓ(x0 ) and Γ is continuous. y ) + βv (y )] = v (x). thus −v (·) is convex. it follows that W (·) is concave and diﬀerentiable. G(x0 )] + βv[G(x0 )].17) implies that v (·) is diﬀerentiable. for all x in some neighborhood D of x0 .

451: Introduction to Economic Growth 6. let us assume (as proved under some conditions above) that the value function v is diﬀerentiable (we take the payoﬀ function F to be diﬀerentiable everywhere). it requires the sum of the marginal gain today from increasing y and the discounted marginal gain from increasing y on the value of all future returns to be equal to zero. For example.18) We know that the solution to our problem has to satisfy this functional equation.19) requires the current cost of increasing y to be compensated by higher values tomorrow. equation (6. y ∗ ) + βv 0 (y ∗ ) = 0. in other words. Let us ﬁrst focus on the case where both x and y are real numbers. y ) + βv(y )] . In this case. we can think of F as being decreasing in y and increasing in x (recall for example the representation of the basic growth model with F (x. Moreover. the constraints on the problem are not binding. y ∈Γ(x) (6.1 Basic Equations Now consider the functional equation v(x) = max [F (x. consider y ∈IntΓ (x). Then we can write a convenient Euler equation for this problem (again using ∗’s to denote optimal values) as ∇y F (x∗ .3 Back to the Fundamentals of Dynamic Programming 6. for all x ∈ X. we have the simpler condition: ∂F (x∗ . Then.19) This is very intuitive.14. Moreover.3. 135 . ∂y (6. y ∗ ) + β ∇y v (y ∗ ) = 0. y ) corresponding to u (f (x) − y + (1 − δ ) x)–or u (f (k (t)) − k (t + 1) + (1 − δ ) k (t))).

y ∗ ) .451: Introduction to Economic Growth In the context of growth. given the optimality condition (6. where ∇z f denotes the gradient vector of function f with respect to the vector z ..20) together with (6. xt+2 ) +β = 0. This is a very nice condition. we could write this with the time subscripts as ∗ ∗ ∂F (x∗ ∂F (x∗ t . g (y ∗ )) +β =0 ∂y ∂x where ∂x denotes the derivative with respect to the ﬁrst argument and ∂y with respect to the second argument. we have the more intuitive equation: v0 (x) = ∂F (x.21) However. which we do not know. this Euler equation is not suﬃcient for optimality.e. but it involves v 0 (y ).19). this corresponds to current cost of reducing consumption to be compensated by higher consumption tomorrow. Now in the one-dimensional case. y ∗ ). i. In addition we need the transversality condition. the derivative of the value function. Here we can use the equivalent of the Envelope Theorem for dynamic programming.14. Alternatively. xt+1 ) t+1 . ∂x (6. combining (6.18): ∇x v (x) = ∇x F (x. In the more general case this is equivalent to: t→∞ ∗ ∗ lim β t ∇xt F (x∗ t . and diﬀerentiate (6. i. xt+1 ) · xt = 0 136 . ∇x y or ∂y/∂x can be ignored.. and g (x) is the optimal policy given state variable x. we have the following very useful condition: ∂F (x∗ .e. y ∗ ) ∂F (y ∗ .20) These equations follow from the fact that x does not appear directly anywhere else (and its eﬀects through y . ∂xt+1 ∂xt+1 (6. In the case of one-dimensional variables.19)).

22). . we can note the following theorem: Theorem 23 Let X ⊂ RK + . If we establish that ∆ is nonnegative for any feasible nonnegative sequence {xt }. . concave. Then the sequence xt+1 t=0 . xt+1 ) · (xt − xt ) + Fy (xt .22) In words. We will see why this transversality condition makes sense shortly. we have ∆ ≥ lim T X t=0 ∗ ∗ ∗ ∗ ∗ β t [Fx (x∗ t . then we will have established {x∗ t } yields no lower utility than any feasible {xt }. xt+1 )] T →∞ as the diﬀerence of the objective function between the feasible sequences {x∗ t } and {xt }. let {x∗ t } be a feasible (nonnegative) sequence satisfying (6. t = 0. thus it must be optimal. In the one-dimensional case. xt+1 ) · x∗ t = 0.14. Now by deﬁnition of a concave function.451: Introduction to Economic Growth where “·” denotes the inner product operator. 1. is optimal for Problem A1 given x0 . if it satisﬁes (6.21) and (6. and β satisfy Assumptions 4. F. xt+1 ) − F (xt . this condition requires that the product of the marginal return from the state variable x times the value of this state variable does not increase asymptotically at a rate faster than 1/β .21) and (6. Proof. . Assumptions 4. and suppose that X. 6 and 7 imply that F is continuous.22) and {xt } another feasible (nonnegative) sequence. so let us deﬁne ∆ ≡ lim T X t=0 ∗ β t [F (x∗ t . . xt+1 ) · (xt+1 − xt+1 )] T →∞ 137 . 6 and © ∗ ª∞ ∗ 7. Let x0 be given. with x∗ t+1 ∈IntΓ(xt ). we have the simpler transversality condition: lim β t ∗ ∂F (x∗ t . 5. But for now. Γ. ∂xt t→∞ (6. and diﬀerentiable.

22). let us again assume that the optimal solution lies in 138 . xT +1 ) · (xT − xT ) ∗ ∗ ≥ − lim β T Fx (x∗ T . xT +1 ) · xT . let F (xT . substituting from (6. T →∞ t=0 Since {x∗ t } satisﬁes (6. establishing the desired 6. we have a ﬁnite-dimensional optimization problem and we can simply look at ﬁrst-order conditions. Fx ≥ 0 and xt ≥ 0. let us suppose that xt is one dimensional and that there is a ﬁnite horizon T . Also. xT +1 ) · (xT +1 − xT +1 ) . In this case.. xt+2 )] · (xt+1 − xt+1 ) + β Fy (xT .e. Moreover. xt+1 ) subject to xt+1 ≥ 0 with x0 as given. Therefore.21).21) into the last term and then using (6.451: Introduction to Economic Growth Since x∗ 0 − x0 = 0. F is increasing in x.14. the terms in the summation are all zero. since the world ends after date T .22) gives ∗ ∗ ∆ ≥ − lim β T Fx (x∗ T . T →∞ T →∞ where the last line uses the fact that from Assumption 5.3. i. Moreover. with xT +1 as the state variable left after the last period (this utility could be thought of as the “salvage value” for example). Then the problem becomes {xt+1 }T t=0 T X t=0 max β t F (xt . xt+1 ) + βFx (xt+1 . result. ∆ ≥ 0 then immediately follows from (6. xT +1 ) be the last period’s utility. rearranging terms gives (T −1 ) X T ∗ ∗ ∗ ∗ ∗ ∗ ∗ β t [Fy (x∗ ∆ ≥ lim t .2 Dynamic Programming Versus the Sequence Problem To get more insights into dynamic programming. all t. let us return to the sequence problem.

β or for any 0 ≤ t ≤ T − 1. xt+2 ) ∂xt+1 = 0.e.. ∂xT +1 ∗ Therefore.451: Introduction to Economic Growth the interior of the constraint set.e. y ) = u (f (x) + (1 − δ) x − y ) . In addition. ∗ ∗ ∂F (x∗ ∂F (x∗ t . recall that F (x. i. we have the following boundary condition T x∗ T +1 ≥ 0. returning to the growth example for a second.23) Intuitively. utility could be improved by consuming that capital either at the last date or at some earlier date. we have for any 0 ≤ t ≤ T − 1.14. Given these. this boundary condition requires that x∗ T +1 should be positive only if an interior value of it maximizes the salvage value at the end. ∂xT +1 (6. xT +1 ) = −u0 (c∗ T ) < 0. If any of it were left. we have ∗ ∂F (x∗ T . and β ∗ ∂F (x∗ T . which are identical to the Euler equations for the inﬁnite-horizon case. Now in this case at the last date T . x∗ t > 0. 139 . there will be no capital left at the end of the world. the ﬁrstorder conditions of this ﬁnite-dimensional problem are exactly the same as the above Euler equation.. xt+1 ) ∂xt+1 +β ∗ ∗ t+1 ∂F (xt+1 . ∂xt+1 ∂xt+1 ∗ ∗ t ∂F (xt . we must have kT +1 = 0. i. In particular. with the mapping x = k and y = k+1 . xt+2 ) +β = 0. for xT +1 . Again. xt+1 ) t+1 . xT +1 ) ∗ xT +1 = 0. so that we do not have to worry about boundary conditions and complementary-slackness type conditions. This is very intuitive.

xT +2 ) ∗ xT +1 = 0. but we generally need a boundary condition at inﬁnity. changing the timing: lim β T ∗ ∂F (x∗ T . heuristically we can derive the transversality condition as an extension of condition (6. xT +1 ) ∗ xT = 0.451: Introduction to Economic Growth Now. xT +1 ) ∗ xT +1 = 0.e. xT +2 ) +β = 0. we have the Euler equation ∗ ∗ ∂F (x∗ ∂F (x∗ T . ∂xT +1 Moreover. i. is to consider the above sequence problem with T → ∞. as T → ∞.. xT +1 ) T +1 . xT +1 ) ∗ xT +1 = 0.23) to T → ∞. which would be one of multiple potential conditions. which implies T →∞ lim β T ∗ ∂F (x∗ T . we have − lim β T +1 T →∞ ∗ ∂F (x∗ T +1 . ∂xT +1 or canceling the negative sign. ∂xT +1 T →∞ which emphasizes that there is no unique transversality condition. and without loss of any generality. 140 . {xt+1 }∞ t=0 max ∞ X t=0 β t F (xt . Therefore. xt+1 ). ∂xT T →∞ which is exactly the transversality condition as (6. This issue will return when we look at optimal control in continuous time. This derivation also emphasizes that alternatively we could have had the transversality condition as lim β T ∗ ∂F (x∗ T . ∂xT +1 ∂xT +1 thus substituting into the previous equation.22). a slightly diﬀerent (and more heuristic) way of obtaining Theorem 23. Take this limit.14.

4 Optimal Growth in Discrete Time We are now in a position to apply the methods developed so far to the problem of optimal growth. or heuristically. with given k (0). 6. ∂xt+1 ∂xt+1 and now the transversality condition (6. But this assumption helps us avoiding inessential technical 141 (6. concavity or even continuity is enough for most of the results. xt+2 ) +β = 0 for all t ≥ 0.25) . We continue to make the standard assumptions on the production function as in Assumptions 1 and 2. In fact.24) subject to k (t + 1) = f (k (t)) + (1 − δ ) k (t) − c (t) and k (t) ≥ 0. which can be established by using a variational argument. In addition. as the limit of the boundary condition as derived above. ∞)→ R is continuously diﬀerentiable and strictly concave. xt+1 ) t+1 . we assume that: Assumption 8 u : [c. In this section. This is considerably stronger than what we need.k(t)}∞ t=0 β t u (c (t)) (6. we obtain the Euler equation: ∗ ∗ ∂F (x∗ ∂F (x∗ t .14.451: Introduction to Economic Growth By taking the limit of the above ﬁnite-dimensional conditions. Recall the optimal growth problem as max ∞ X t=0 {c(t).22) is also necessary. I will limit myself to optimal growth.

the following proposition immediately follows: Proposition 13 Given Assumptions 1. Suppose. Optimality of the solution to the value function (6. The ﬁrst step is to write the optimal growth problem as a (stationary) dynamic programming problem. since.26) for the problem (6. in particular Theorems 15-22. This follows immediately from what we have done so far: V (k) = max {u (c) + βV [f (k) + (1 − δ ) k − c]} c∈Γ(k) (6. s (k) > s (k0 ). Given the above theorems. That V (k) exists follows from Theorem 19. that s (k) is decreasing. Proof. s (k ) is feasible when the capital stock is k0 . 142 . and the fact that it is increasing and strictly concave.26) with Γ (k) given by the interval [c.24) and (6.e.451: Introduction to Economic Growth details.24) and (6. to arrive at a contradiction. s (k 0 ) is feasible at capital stock k. Moreover. has a stationary solution characterized by the value function V (k) and consumption function c (k). Thus we only have to show that s (k) is nondecreasing. with the policy correspondence being a policy function follows from Theorem 21. 2 and 8. the optimal growth model as speciﬁed in (6. there exists k and k 0 > k such that s (k) > s (k0 ). Since k0 > k.25) follows from Theorems 15-18. where s (k) = f (k) + (1 − δ ) k − c (k). i. V (k) is strictly increasing and concave in k and s (k) is nondecreasing. Moreover.14. f (k) + (1 − δ ) k ] given the nonnegativity of the capital stock. The lower bound on consumption is imposed to have a compact set of consumption possibilities.25).. by hypothesis. This can be proved by contradiction. The amount s (k) is the capital stock of the next period.

contradicting (6. Assumption 2 (the Inada conditions) imply that savings and consumption levels have to be interior.27). But clearly.451: Introduction to Economic Growth By optimality and feasibility. Or denoting z ≡ f (k) + (1 − δ ) k and x ≡ s (k ) and similarly for z 0 and x0 . which combined with the fact that z 0 > z and that u is strictly concave and increasing implies that u (z − x0 ) − u (z − x) > u (z 0 − x0 ) − u (z 0 − x) . Combining and rearranging these.14. (z − x0 ) − (z − x) = (z 0 − x0 ) − (z 0 − x) . thus Theorem 22 applies and immediately establishes: 143 (6. we have u (z − x0 ) − u (z − x) ≤ u (z 0 − x0 ) − u (z 0 − x) . we have u (f (k) + (1 − δ ) k − s (k)) − u (f (k) + (1 − δ ) k − s (k0 )) ≥ β [V (s (k0 )) − V (s (k))] ≥ u (f (k0 ) + (1 − δ) k0 − s (k)) −u (f (k0 ) + (1 − δ) k0 − s (k0 )) . In addition. This establishes that s (k) must be nondecreasing everywhere. we must have: V (k) = u (f (k) + (1 − δ ) k − s (k)) + βV (s (k)) ≥ u (f (k) + (1 − δ ) k − s (k0 )) + βV (s (k0 )) V (k 0 ) = u (f (k0 ) + (1 − δ ) k0 − s (k0 )) + βV (s (k0 )) ≥ u (f (k0 ) + (1 − δ ) k0 − s (k)) + βV (s (k )) .27) .

we have the steady state capital-labor ratio as β [f 0 (k∗ ) + (1 − δ )] = 1. Consequently. the value function V (k) deﬁned above is diﬀerentiable. A steady state is deﬁned as usual as an allocation in which the capital-labor ratio and consumption do not depend on time. we can look at the Euler equations.14.451: Introduction to Economic Growth Proposition 14 Given Assumptions 1. Thus we have 144 . Consequently. we have the familiar-looking condition u0 (ct ) = β [f 0 (kt+1 ) + (1 − δ )] u0 (ct+1 ) . but simply on technology. let us write the recursive formulation as V (k) = max {u (f (k) + (1 − δ ) k − s) + βV [s]} s∈Γ(k) In this case the Euler equation takes the simple form: u0 (c) = βV 0 (s) where s denotes the next date’s capital stock. from Theorem 23. since f (·) is strictly concave. Moreover. Applying the envelope condition.28) which is a remarkable result. we have V 0 (k) = [f 0 (k) + (1 − δ )] u0 (c) . so again denoting this by *. depreciation and the discount factor. because it shows that the steady state capital-labor ratio does not depend on preferences. We will obtain an analogue of this result in the continuous-time neoclassical model as well. 2 and 8. To do this. (6. k ∗ is uniquely deﬁned.

145 . Consequently. is unique and ﬁnite by Assumption 2. Since the steady-state is unique. s (kt ) for kt 6= k∗ cannot satisfy k∗ = s (k ∗ ). if k0 < k∗ .451: Introduction to Economic Growth Proposition 15 In the neoclassical optimal growth model speciﬁed in (6. in the optimal growth model there exists a unique steady state and the economy monotonically converges to the unique steady state. simply use the fact that kt+1 = s (kt ) with s (·) deﬁned in Proposition 13. Next. then the equilibrium consumption sequence ct ↑ c∗ and if k0 > k∗ . since this is the unique steady state. Uniqueness and existence were established above. and starting from any initial k0 . s (kt ) is s k an increasing sequence in a compact set.14. i. the economy monotonically converges to this unique steady state.e. Finally. Moreover. thus s (kt ) → s be equal to k∗ .28).. completing the proof. then the equilibrium capital stock sequence kt ↓ k ∗ .25) with Assumptions 1. we can also show that consumption also monotonically increases (or decreases) along the path of adjustments to the unique-steady state: Proposition 16 c (k ) deﬁned in Proposition 13 is nondecreasing. Consequently. However. To establish monotonic convergence. there exists a unique steady-state capital-labor ratio k∗ given by (6. then the equilibrium capital stock sequence kt ↑ k∗ and if k0 > k∗ . if k0 < k∗ . then ct ↓ c∗ . note s (kt ) is nonnegative and can never exceed ¡ ¢ ¡ ¡ ¢¢ ¯ =f s k ¯ . thus s (kt ) → k∗ . 2 and 8. for example by accumulating more and more capital (if it starts with a too low capital-labor ratio). any limit point of s (kt ) must set necessarily converges. thus it must be increasing. A monotonically increasing sequence in a compact ¯ for some s ¯. Proof. which exists. and was shown to be nondecreasing. where c∗ is given by c∗ = f (k∗ ) − δk ∗ .24) and (6.

it is also useful to see how this optimal growth allocation can be decentralized. Suppose they all start with capital stock k0 . in this particular case we can use the second welfare theorem to show that the optimal growth allocation is also a competitive equilibrium. i. and provide a more detailed discussion of the equilibrium growth in the context of the continuous time model. with utility function given by u (c) as above. Households rent their capital to ﬁrms. including incorporating population growth and technological change. since this depends on the utility function.14. Suppose that all households are identical. We will return to all of these issues. But for now. In addition. the steady state capital-labor ratio and steady state income level do not depend on the savings rate anyway. It is straightforward to see that households will receive a rental price of Rt = f 0 (kt ) because of competitive market prices.e. But interestingly and very diﬀerently from the Solow growth model.451: Introduction to Economic Growth The proof of Proposition 16 is left as an exercise to you. and normalize their measure to 1. and we can do the usual exercises we performed with the Solow growth model.29) for renting one unit of capital at time t in terms of date t + 1 goods. 6. They will therefore face a gross rate of return equal to rt = [f 0 (kt ) + (1 − δ )] receive the wage rate of wt = f (kt ) − kt f 0 (kt )..5 Competitive Equilibrium Growth To show that the Pareto optimal growth allocation can be decentralized is very straightforward. This treatment shows that the optimal growth model is very tractable. they will . There is no immediate counterpart of a savings rate. 146 (6. The other side of the economy are competitive ﬁrms.

I will discuss this in greater detail below. Both are exactly as in the optimal growth problem. a similar argument establishes that the whole competitive equilibrium path is identical to 147 .e. and also subject to a no Ponzi constraint which requires the individual asset holdings not to go to minus inﬁnity. equations (6. it suﬃces to see that by exactly the same Euler equation type arguments.28) and (6. Next.451: Introduction to Economic Growth Now consider the maximization problem of the representative household: max ∞ X t=0 {ct .. therefore. we have u0 (ct ) = rt+1 βu0 (ct+1 ) .14. Imposing steady state implies that ct = ct+1 . where at denotes asset holdings at time t. For now.29). so the capital-labor ratio of the competitive equilibrium is given by β [f 0 (kt+1 ) + (1 − δ )] = 1.29). market clearing immediately implies that rt+1 is given by (6. The steady state is given by β [f 0 (k∗ ) + (1 − δ )] = 1. i. In fact.at }∞ t=0 β t u (ct ) subject to the ﬂow budget constraint at+1 = rt at + wt − ct . we must have rt+1 β = 1.

not surprising in view of the second (and ﬁrst) welfare theorems we saw above.14. This is.451: Introduction to Economic Growth the optimal growth path. 148 . We will discuss many of the implications of competitive economic growth in the neoclassical model once we go through the continuous time version as well. of course.

y : [t0 .Chapter 7 Brief Review of Optimal Control The continuous time problem brings a number of new issues. This requires us to review some basic ideas from the calculus of variation and from optimal control. The main reason is that even with a ﬁnite horizon. 149 . I will start with the ﬁnite-horizon problem and the simplest treatment (which is much more similar to calculus of variation than optimal control). and then provide the more powerful theorems from optimal control. but most of the tools and ideas that are necessary for this course are very straightforward. the maximization is with respect to an inﬁnite-dimensional object (in fact an entire function. t1 ] → R). to give you the basic idea.

It may often hit the boundary of the feasible set etc. These features make it diﬃcult for us to know what type of optimal policy to look for. whose behavior is governed by the diﬀerential equation (7.1 7.14.3) (7. y (t)) ≡ f (t.1. 2. (7. The diﬃculty of this problem arises from two features: 1. For example. y (t)) and y (t) ∈ Y (t) for all t. t1 can be a choice variable as well. The constraint takes an unusual form of a diﬀerential equation.x1 0 (7. or it could extend to inﬁnity as we will see later).2). We are choosing a function: y : [0.1) subject to x ˙ (t) = g (t. y may be a very discontinuous function. x (t) . x (t) . y (t)) dt x(t). x (0) = x0 and x (t1 ) = x1 .2) Here x (t) ∈ R is the state variable. This is the simplest optimal control problem because it has boundary conditions that regulate when the planning horizon ends (more generally. t1 ] → Y rather than a vector or a ﬁnite dimensional object. 150 . we assume that f and g are continuously diﬀerentiable functions. In addition. y (t) ∈ Y (t) ⊂ R is the control variable.1 Finite-Horizon Optimal Control The Fundamental Problem Consider the following ﬁnite-horizon continuous time problem Z t1 max J (x (t) .y (t).451: Introduction to Economic Growth 7.

ε) . t1 ] and with x (0. This implies that x (t. let us suppose that ∃ a continuous function y ˆ (·) deﬁned over [0. ε)) for all t ∈ [0.2 Variational Arguments Before going into greater detail.e. we can always ﬁnd εη > 0 such that for any η (·) function y ˆ (t) + εη (t) ∈ IntY (t) for all ε < εη . Now consider the following variation y (t. Therefore. However. t1 ] with y ˆ (t) ∈ IntY (t) which achieves the optimum in this problem. of course. ε) ∈ / Y (t) for some t. ε) = x0 . in analogy with regular calculus. which can be done by using the variational principle of the calculus of variation. and deﬁne x (t. and a continuous function over a compact set [0. let us ﬁx an arbitrary η (·). Thus we can conduct variational arguments for small ε’s. the argument that there is no gain from a variation for small ε’s is essentially what we need. But. let us try to understand the essence of the problem. ε). y (t. The problem. 151 (7.4) . since y ˆ (t) ∈IntY (t). To prepare for these arguments. because given η (t). ε) as the path of the state variable corresponding to the path of control variable y (t. where η (t) is an arbitrary ﬁxed continuous function. by varying ε.14. For this purpose. We refer to this as a variation.1.451: Introduction to Economic Growth 7. x (t.. is that some of these may be infeasible. we obtain diﬀerent sequences of controls. ε) = g (t. y (t. we are ruling out both the boundary conditions and discontinuities. ε) = y ˆ (t) + εη (t) . i. ε) is given by: x ˙ (t. t1 ] is bounded.

ε) . x (t. t1 ]: g (t. ε)]] dt. ε)] dt = 0. rewrite the equation (7. with a similar interpretation to the Lagrange multipliers in regular (constrained) optimization.4). t1 ˙ (t) x (t. chosen suitably. ε)) + λ (t) x (t. y (t. ε) dt 0 −λ (t1 ) x (t1 . ε)) + λ (t) [g (t. ε) . ε)) is feasible. will be the costate variable. ε) . so for all t ∈ [0. x (t. y (t. x (t. ε) dt. and that for ε < εη . x (t. we have Z Z t1 λ (t) x ˙ (t. x (t. ε)) dt. we obtain: Z t1 h i ˙ Φ (ε) ≡ f (t. y (t. ε) .6) to (7. t1 ] → R. ε)) − x ˙ (t. we have Φ (ε) ≤ Φ (0) for all ε < εη . Next. 152 . Now add (7.451: Introduction to Economic Growth Now deﬁne for ε < εη : Φ (ε) ≡ Z t1 f (t. y (t. ε) .14. ε) dt.5) to obtain: Φ (ε) ≡ Z t1 0 [f (t. (7. it must be the case that Z t1 λ (t) [g (t. ε)) − x ˙ (t. ε) − λ (0) x0 − 0 λ (t) x ˙ (t.6) 0 The function λ (·). x (t. ε) ≡ 0. y (t. y (t. Now for any continuously diﬀerentiable function λ : [0. λ 0 Substituting this back. x (t. ε)) + λ (t) g (t. Start by considering the integral Integrating this by parts. ε) dt = λ (t1 ) x (t1 . y (t. (7. ε) + λ (0) x0 . ε) . ε) (and thus x (t. ε)) − x ˙ (t. ε) .5) 0 By the fact that y ˆ (t) is optimal. R t1 0 Now we want to evaluate this term. y (t.

y ˆ (t)) + λ t1 [fy (t. fy etc. y 153 (7. x (t) . x (t) . y ˆ (t).. ε) . ε)) + λ (t) gx (t. 0) dt fx (t. since f and g are continuously diﬀerentiable. ε) . x (t) . Now at evaluating this expression at ε = 0.14. i. if there exists some function η (t) for which Φ0 (0) 6= 0. y ˆ (t)) + λ (t) gx (t. This can only be possible if the second integral is equal to zero for all η (t). only if Z t1 Z h i ˙ (t) xε (t. y ˆ (t)) + λ (t) gy (t. ε) . As with the standard ﬁnite-dimensional optimization. y (t..e. y ˆ (t))] η (t) dt 0 [fy (t. x (t.7) . we need to have Φ0 (0) ≡ 0 for all η (t) . ε) dt fx (t. y (t. x (t) . diﬀerentiation gives Φ (ε) ≡ 0 Z t1 0 + −λ (t1 ) xε (t1 . x (t) . ε) . x (t. ε)) + λ (t) gy (t. ε))] η (t) dt 0 Z t1 0 + −λ (t1 ) xε (t1 . we have Φ (0) ≡ 0 Z h i ˙ (t) xε (t. Therefore. y (t. ε) . x (t. y ˆ (t))] η (t) dt = 0 for all η (t) . fy (t. ε) in ε by construction. Denoting their derivatives by xε and yε . x (t. and so is y (t. y (t. this means that the value of the program can be improved. 0) . x (t) . where x (t) denotes the path of the state variable corresponding to the optimal plan. This is feasible by Leibniz’s rule. y ˆ (t)) + λ (t) gy (t. x (t) . y ˆ (t)) ≡ 0 for all t ∈ [0. fx . and the derivatives of f and g by ft . t1 ] . ε)) + λ t1 [fy (t. x (t) . 0 which is only possible if ˆ (t)) + λ (t) gy (t.451: Introduction to Economic Growth Now diﬀerentiate this term with respect to ε.

8) should remind you of a Lagrangian maximization. with f and g continuously diﬀerentiable. x (t) .7) and (7.8) 7. so we need to have the ﬁrst integral identically equal to zero. This derivation (from calculus of variation) therefore has established the following theorem: Theorem 24 (Necessary Conditions) Consider the problem of maximizing (7. xε is also arbitrary.2). with f and g continuously diﬀerentiable. (7. λ) be 154 (7. Let H (t. By analogy with the Lagrangian. has an interior solution y ˆ (t) ∈IntY (t) with corresponding path of state variable x (t). x. y. t1 ] such that (7.2) and (7. a much more economical way of expressing Theorem 24 is to construct the equivalent of the Lagrangian in this case. x (t) .2) and (7. then there exists a continuously diﬀerentiable costate function λ (·) deﬁned over t ∈ [0. Then we have Theorem 25 (Simpliﬁed Maximum Principle) Consider the problem of maximizing (7.3). y. y ˆ (t)) + λ (t) gx (t. the Hamiltonian: H (t. x. x (t) .451: Introduction to Economic Growth By the same reasoning.1. or ˙ (t) = − [fx (t.1) subject to (7.3).3 Simpliﬁed Maximum Principle The conditions (7. x (t) .14. (7.1) subject to (7. λ) ≡ f (t.7) and (7.9) . has an interior solution y ˆ (t) ∈IntY (t) with corresponding path of state variable x (t). y (t)) + λ (t) gy (t.8) hold. y ˆ (t))] λ and therefore λ (t1 ) = 0. y (t)) .

While Theorem 25 gives necessary conditions. t1 ] .12) Theorem 25 is a simpliﬁed version of the celebrated Maximum Principle. Hy (t.9). λ (t)) = 0 for all t ∈ [0. t1 ] . y ˙ (t) = −Hx (t. as in regular optimization problems. The following theorem provides conditions for the necessary conditions to also be suﬃcient to characterize the optimal plan. 155 . and λ (t1 ) = 0. we ﬁnd the optimal solution by looking jointly for a set of “multipliers” and the optimal path of the control and state variables. λ (t)) for all t ∈ [0. y ˆ (t) . λ x ˙ (t) = Hλ (t. Then the optimal control y ˆ (t) and the corresponding path of the state variable x (t) satisfy the following necessary conditions: ˆ (t) . x (t) .451: Introduction to Economic Growth given by (7. these may not be suﬃcient. the costate variables are informative about the value of relaxing the constraint. Suﬃciency is again guaranteed by imposing concavity. λ (t)) for all t ∈ [0. Again as in the usual constrained maximization problems. and the more general version will be given below. Here λ (t) is the value of an inﬁnitesimal increase in x (t) at time t. there is no value to having more x. t1 ] . After the planning horizon.11) (7. x (t) . Here the multipliers are referred to as the costate variables. For now. 3. and x (0) = x0 . 2.14. With this interpretation.10) (7. it makes sense that λ (t1 ) = 0 is part of the necessary conditions. x (t) . (7. y ˆ (t) . This is therefore the ﬁnite-horizon equivalent of the transversality condition we encountered above. a couple of features are worth noting: 1. As in the usual constrained maximization problems.

with f and g continuously diﬀerentiable. If M (t. λ).2) and (7.1). then y ˆ (t) and the corresponding x (t) achieve the unique global maximum of (7.1) subject to (7. so they are omitted. x. deﬁne M (t.10)-(7.12).451: Introduction to Economic Growth Theorem 26 (Mangasarian Suﬃcient Conditions) Consider the problem of maximizing (7.3).1) subject to (7. λ) ≡ H (t. then y the corresponding x (t) achieve the unique global maximum of (7.1) subject to (7. x.1).10)-(7. Deﬁne H (t. Suppose also that for the resulting ˆ (t) and costate variable λ (t). t1 ]. y ˆ (t) . λ) is jointly concave in (x. since given the concavity/convexity of the g (·) function. the concavity of the Hamiltonian will depend on the sign on the costate variable λ (t). The proofs of these theorems are long and not necessary for what will follow. x.14. and suppose that an interior solution y ˆ (t) ∈IntY (t) and the corresponding path of state variable x (t) satisfy (7. Given the resulting costate variable λ (t). x. y. t1 ]. y. is diﬃcult to apply. λ) as in (7. y. with corresponding costate variable λ (t).2) and (7. The following lemma (again proof omitted) provides some information on the sign of λ (t): Lemma 2 Suppose that y ˆ (t) and the corresponding x (t) are the optimal solutions to maximizing (7. H (t. Then we have that 156 . y ) for all t ∈ [0. x. λ) as in (7.12).9).3). Deﬁne H (t. with f and g continuously diﬀerentiable. λ) is concave in x for all t ∈ [0. even Theorem 27.3). An alternative set of suﬃcient conditions are provided by Arrow: Theorem 27 (Arrow Suﬃcient Conditions) Consider the problem of maximizing (7. and suppose that an interior solution y ˆ (t) ∈IntY (t) and the corresponding path of state variable x (t) satisfy (7. As stated Theorem 26.9).2) and (7. x.

there is an intimate relationship between the sign of the multiplier and the returns from increasing the stock of the state variable. y (t)) ≡ 157 Z t1 x(t).x1 f (t. etc. The vector-values theorems are direct generalizations of the ones presented above. λ (t)) > 0 for all t ∈ [0.y(t). then the Hamiltonian given in (7. If fx (t. λ (t)) < 0 for all t ∈ [0. The usefulness of Lemma 2 comes from the fact that if λ (t) > 0 for all t (which follows from fx > 0 for all t). t1 ]. t1 ). x (t) . If fx (t.9) is a concave function of x and y for given λ (t) when f and g are concave functions. but here we need the eﬀect of the state variable to be positive everywhere in order for the multiplier to be positive. y Therefore. t1 ).4 Generalizations The above theorems can be immediately generalized to the case in which the state variable and the controls are vectors rather than scalars. x (t) . The constrained case requires constraint qualiﬁcation conditions as in the standard ﬁnite-dimensional optimization case. 3. ˆ (t) . These are slightly more messy to express.14. 2. and also to the case in which there are constraints. In particular. x (t) . Therefore.1. y (t)) dt (7. then λ (t) = 0 for all t ∈ [0. I will not state these theorems.451: Introduction to Economic Growth 1. then λ (t) < 0 for all t ∈ [0. t1 ]. let max J (x (t) . t1 ). t1 ]. x (t) .13) 0 . y ˆ (t) . and since we will make no use of the constrained maximization problems. as in standard maximization problems. then λ (t) > 0 for all t ∈ [0. λ (t)) = 0 for all t ∈ [0. If fx (t. the suﬃcient conditions in Theorem 26 are very straightforward to check (though often quite restrictive). y ˆ (t) . and are useful in growth models with multiple capital goods. 7.

17) (7.15).13) subject to (7. x (t) . We then have: Theorem 28 (Maximum Principle) Consider the problem of maximizing (7. we have straightforward generalizations of the suﬃciency conditions: Theorem 29 (Mangasarian Suﬃcient Conditions) Consider the problem of maximizing (7. ˆ (t) . y Moreover. y λ ˆ (t) . x ˙ (t) = ∇λ H (t. t1 ] .14) and (7. (7. x (t) . λ) ≡ f (t. λ (t)) for all t ∈ [0.15). x (t) .14) and (7. λ (t)) = 0 for all t ∈ [0. we again assume that f and g are continuously diﬀerentiable functions.19) . with f and g continuously diﬀerentiable. Then the optimal control y ˆ (t) and the corresponding path of the state variable x (t) satisfy the following necessary conditions: ∇y H (t. x (t) . with f and g continuously diﬀerentiable. Deﬁne 158 (7. y (t)) . y. x. In addition. Let H (t. and y (t) ∈ Y (t) for all t. y ˆ (t) . x (t) . t1 ] and λ (t1 ) = 0.14. y (t)) + λ (t) gy (t.16) where λ (t) ∈ RK . x. x (0) = x0 and x (t1 ) = x1 .18) (7. has an interior solution y ˆ (t) ∈IntY (t) with corresponding path of state variable x (t).13) subject to (7. y. y (t)) .451: Introduction to Economic Growth subject to x ˙ (t) = g (t.14) Here x (t) ∈ RK for some K ≥ 1 is the state variable and again y (t) ∈ Y (t) ⊂ RN for some N ≥ 1 is the control variable. t1 ] and x (0) = x0 . ˙ (t) = −∇x H (t.15) (7. x (t) . (7. λ) be given by H (t. λ (t)) for all t ∈ [0.

and equally important for our purposes. y) for all t ∈ [0.14. then y and the corresponding x (t) achieve the unique global maximum of (7. H (t. and suppose that an interior solution y ˆ (t) ∈IntY (t) and the corresponding path of state variable x (t) satisfy (7.19). First. and suppose that an interior solution y ˆ (t) ∈IntY (t) and the corresponding path of state variable x (t) satisfy (7.13) subject to (7. Second. λ) as in (7.1.15). Deﬁne H (t. λ).13). This is done in the next section. whereas analysis of growth models requires us to solve inﬁnite horizon problems.14) and (7. y. This is in general a very strong assumption. Theorem 30 (Arrow Suﬃcient Conditions) Consider the problem of maximizing (7. y.19). x. To deal with both of these issues. x.13). 159 . λ) as in (7. λ) ≡ H (t. we need to look at the more modern theory of optimal control. y. deﬁne M (t. y ˆ (t) .5 Limitations The limitations of what we have done so far are obvious. Suppose also that for the resulting ˆ (t) costate variable λ (t).17)-(7. with f and g continuously diﬀerentiable. then y ˆ (t) and the corresponding x (t) achieve the unique global maximum of (7. t1 ]. x. we have assumed that a continuous and interior solution to the optimal control problem exists. 7.17)-(7. t1 ]. we have so far looked at the ﬁnite horizon case. If M (t.16).16). λ) is concave in x for all t ∈ [0. λ) is jointly concave in (x. x. x.451: Introduction to Economic Growth H (t. Suppose also that for the resulting costate variable λ (t). x.

21) (7. x (t) .y(t) max J (x (t) . which are similar to those in the discrete time analysis. and there is no choice of endpoint x1 .2) and (7.1 Inﬁnite-Horizon Optimal Control The Basic Problem: Necessary and Suﬃcient Conditions Consider the inﬁnite-horizon following version of problem of maximizing (7. since discontinuous controls are allowed as long as they are piecewise continuous.1) subject to (7.3). t→∞ (7.21) given y (t) (since x (t) is given by a continuous diﬀerential equation.2 7. y (t)) dt (7.2. we call a pair (x (t) . There are a number of technical diﬃculties when dealing with the inﬁnite-horizon case. For this problem. Notice that this is a signiﬁcant generalization of the above approach. x (0) = x0 and lim x (t) ≥ x1 .22) The main diﬀerence is that now time runs to inﬁnity. Primary among those is the fact 160 . x (t) .451: Introduction to Economic Growth 7.14. simply requiring this function to be real-valued. the piecewise continuity of y (t) ensures the piecewise smoothness of x (t)). Z ∞ x(t). In addition. y (t)) admissible if y (t) is a piecewise continuous function of time and x (t) is a piecewise smooth function of time satisfying (7. and y (t) ∈ R for all t. y (t)) . I have simpliﬁed the problem by removing the feasibility set on the control y (t). y (t)) ≡ f (t.20) 0 subject to x ˙ (t) = g (t.

λ) be given by (7. There is no boundary condition in Theorem 31 corresponding to λ (t1 ) = 0 of Theorem 25. y λ ˆ (t) . λ) = 0 161 t→∞ . One might be tempted to impose a condition of the form t→∞ lim λ (t) ≥ 0 as the transversality condition.20) may not be ﬁnite. Let H (t.451: Introduction to Economic Growth that the value of the functional in (7. has an interior solution y ˆ (t) with corresponding path of state variable x (t). y ˆ (t) . ˙ (t) = −Hx (t. y t→∞ (7. x ˙ (t) = Hλ (t. x (t) .22). x. Consequently. y.14. λ (t)) for all t ∈ R+ . x (0) = x0 and lim x (t) ≥ x1 . y. x. x (t) . The main theorem for the inﬁnite-horizon optimal control problem is the following maximum principle: Theorem 31 Suppose that problem of maximizing (7. λ (t)) for all t ∈ R+ . A milder transversality condition of the form lim H (t. λ (t)) = 0 for all t ∈ R+ .9). We will deal with some of these issues below. We will see an example where this does not apply soon. Then the optimal control y ˆ (t) and the corresponding path of the state variable x (t) satisfy the following necessary conditions: Hy (t. x (t) . To do this we need an inﬁnite-horizon version of the transversality condition.20) subject to (7.25) Notice an important diﬀerence between Theorem 25 and the current theorem. with f and g continuously diﬀerentiable. the necessary conditions in Theorem 31 will not uniquely pin down a solution path.23) (7.24) (7. ˆ (t) .21) and (7. but this is not in general the case.

21) and (7.21) and (7. x.25). λ). there are immediate generalizations of the suﬃciency theorems to this case. Given the resulting costate variable λ (t). deﬁne M (t. y. H (t. This condition will disappear when we can impose a proper transversality condition. λ) as in (7. x. λ) ≡ H (t.22).20) subject to (7. x.451: Introduction to Economic Growth always applies. λ) is concave in x and limt→∞ λ (t) (x (t) − x ˜ (t)) ≤ 0 for all x ˜ (t) implied by an admissible control path y (t). x. with f and g continuously diﬀerentiable. and suppose that a solution y ˆ (t) and the corresponding path of state variable x (t) satisfy (7.9). Theorem 33 (Arrow Suﬃcient Conditions for Inﬁnite Horizon) Consider the problem of maximizing (7. λ) as in (7. x. Suppose also that for the resulting costate variable λ (t). Stronger transversality conditions apply when we put more structure on the problem. If M (t. Deﬁne H (t. that limt→∞ λ (t) (x (t) − x then y ˆ (t) and the corresponding x (t) achieve the unique global maximum of (7.22). Deﬁne H (t.23)-(7. but is not easy to check. with f and g continuously diﬀerentiable. 162 . Before we do this. y. then y ˆ (t) and the corresponding x (t) achieve the unique global maximum of (7. and suppose that a solution y ˆ (t) and the corresponding path of state variable x (t) satisfy (7.23)-(7.20). λ) is jointly concave in (x. y ) for all t ∈ R+ and ˜ (t)) ≤ 0 for all x ˜ (t) implied by an admissible control path y (t).25). x.20) subject to (7. y. Notice that both of these this eﬃciency theorems have the diﬃcult to check condition that limt→∞ λ (t) (x (t) − x ˜ (t)) ≤ 0 for all x ˜ (t) implied by an admissible control path y (t).14.9). Theorem 32 (Mangasarian Suﬃcient Conditions for Inﬁnite Horizon) Consider the problem of maximizing (7. y ˆ (t) .20).

163 . Example 2 Consider the following problem: Z ∞ [log (c (t)) − log c∗ ] dt max 0 subject to ˙ (t) = [k (t)]α − c (t) − δk (t) k k (0) = 1 and t→∞ lim k (t) ≥ 0 where c∗ ≡ [k∗ ]α − δk∗ and k∗ ≡ (α/δ )1/(1−α) . In other words. illustrates that there are in general no transversality conditions. and implies the following necessary conditions (dropping time dependence to simplify the notation): Hc = Hk 1 −λ=0 c ¡ ¢ ˙ = λ αk α−1 − δ = −λ. c. which is very close to the original Ramsey model. c∗ is the maximum level of consumption that can be achieved in this model. λ) = [log c − log c∗ ] + λ [kα − c − δk] . The Hamiltonian is straightforward to construct and takes the form H (k. This way of writing the objective function makes sure that the integral converges and takes a ﬁnite value (since c (t) cannot exceed c∗ forever).14.2.451: Introduction to Economic Growth 7.2 Lack of Transversality Conditions The following example.

451: Introduction to Economic Growth It can be veriﬁed that c (t) → c∗ satisﬁes the necessary conditions. y (t)) dt with ρ > 0. our interest is with the growth models where the utility is discounted exponentially. λ (t)) = 0.27) (7. however. c (t) . y (t)) ≡ 0 exp (−ρt) u (x (t) . comes from the fact that we did not impose enough structure on the functions f and g . t→∞ (7. taking the form: Z ∞ x(t).2. (7. y (t)) . t→∞ c∗ t→∞ Therefore. It can be veriﬁed. 7. Then the problem is a more special one. the equivalent of f .14. The Hamiltonian in this case would 164 . t→∞ so the weaker transversality condition holds. x (0) = x0 and lim x (t) ≥ x1 . implies that lim λ (t) = 1 > 0 and lim k (t) = k∗ . the equivalent of the standard ﬁnite-horizon transversality conditions do not hold. however.26) subject to x ˙ (t) = g (t. As discussed above. This. and must be part of any optimal path.y(t) max J (x (t) . and y (t) ∈ R for all t. depends on time only through exponential discounting.28) The special feature of this problem is that the payoﬀ function.3 Discounted Inﬁnite-Horizon Optimal Control Part of the diﬃculty. that along the optimal path lim H (k (t) . x (t) . especially regarding the absence of a transversality condition.

y (t)) H which is not explicitly a function of time. y ˆ (t) .14. we can work with the current-value Hamiltonian.32) . We have the following result. µ (t)) for all t ∈ R+ and lim [exp (−ρt) x (t) µ (t)] = 0. y (t)) = exp (−ρt) [u (x (t) . λ (t)) ≡ u (x (t) . y (t) . in this case. y (t)) + µ (t) g (t. where the second line deﬁnes µ (t) ≡ exp (ρt) λ (t). µ) be the current-value Hamiltonian given by (7.30) ˆ x (x (t) .26) subject to (7. x ˙ (t) = H 165 (7. but also shows the necessity of a transversality condition: Theorem 34 (Maximum Principle for Discounted Inﬁnite-Horizon Problems) Suppose that problem of maximizing (7. rather than working with the standard Hamiltonian. x (t) . x (t) . ρµ (t) − µ ˙ (t) = H (7.28).27) and (7. y ˆ (t) . λ (t)) = exp (−ρt) u (x (t) . y ˆ (t) . In fact. y (t))] . x (t) .29) (7.29). which states not only the necessary conditions similar to Theorem 31. Then the optimal control y H ˆ (t) and the corresponding path of the state variable x (t) satisfy the following necessary conditions: ˆ y (x (t) . µ (t)) = 0 for all t ∈ R+ . H t→∞ (7. y (t) . λ (t)) for all t ∈ R+ . This equation makes it clear that the Hamiltonian depends on time explicitly only through the exp (−ρt) term. y (t)) + µ (t) g (t. y. x (0) = x0 and lim x (t) ≥ x1 . with u and g continuously diﬀerentiable. has a solution y ˆ (t) with corresponding path of state variable x (t). deﬁned as ˆ (x (t) . x (t) . Let ˆ (x. y (t)) + λ (t) g (t.451: Introduction to Economic Growth be: H (t.31) t→∞ ˆ λ (x (t) .

14.451: Introduction to Economic Growth The important feature of Theorem 34, which is the most useful theorem for the rest of the course, is that it also shows the transversality condition

t→∞

lim [exp (−ρt) x (t) µ (t)] = 0

is necessary. Notice that compared to the transversality condition before, there is the additional term exp (−ρt). This is because the transversality condition applies to the original costate variable λ (t), i.e., limt→∞ [x (t) λ (t)] = 0, and as shown above the current-value costate variable µ (t) is given by µ (t) = exp (ρt) λ (t) = 0. The suﬃciency theorems can also be strengthened now by incorporating the transversality condition and expressing the conditions in terms of the current-value Hamiltonian: Theorem 35 (Mangasarian Suﬃcient Conditions for Discounted Inﬁnite-Horizon Problems) Consider the problem of maximizing (7.26) subject to (7.27) and (7.28), with u ˆ (x, y, µ) as the current-value Hamiltonian as in and g continuously diﬀerentiable. Deﬁne H (7.29), and suppose that a solution y ˆ (t) and the corresponding path of state variable x (t) satisfy (7.30)-(7.32). Suppose also that for the resulting current-value costate variable µ (t), ˆ (x, y, µ) is jointly concave in (x, y ) for all t ∈ R+ , then y H ˆ (t) and the corresponding x (t) achieve the unique global maximum of (7.26). Theorem 36 (Arrow Suﬃcient Conditions for Discounted Inﬁnite-Horizon Problems) Consider the problem of maximizing (7.26) subject to (7.27) and (7.28), with u ˆ (x, y, µ) as the current-value Hamiltonian as and g continuously diﬀerentiable. Deﬁne H in (7.29), and suppose that a solution y ˆ (t) and the corresponding path of state variable x (t) satisfy (7.30)-(7.32). Given the resulting current-value costate variable µ (t), deﬁne ˆ (x, y M (t, x, µ) ≡ H ˆ, µ). If M (t, x, µ) is concave in x, then y ˆ (t) and the corresponding x (t) achieve the unique global maximum of (7.26). 166

**Chapter 8 The Neoclassical Growth Model
**

We are now ready to start our analysis of the standard neoclassical growth model (also known as the Ramsey, or Cass-Koopmans model). This model diﬀers from the Solow model only in explicitly modeling the consumer side and endogenizing savings (i.e., allowing consumer optimization). Beyond its use as a basic growth model, this model has become a workhorse for many areas of macroeconomics, including the analysis of ﬁscal policy, taxation, business cycles, and even monetary policy.

8.1

Preferences, Technology and Demographics

The economy is an inﬁnite-horizon economy in continuous time (the discrete-time version was analyzed above). We assume that the economy admits a representative household with instantaneous utility function u (c (t)) , and we make the following standard assumptions on this utility function: 167 (8.1)

14.451: Introduction to Economic Growth Assumption 9 u (c) is strictly increasing, twice continuously diﬀerentiable with derivatives u0 and u00 , and concave, and satisﬁes the following Inada type assumptions: lim u0 (c) = ∞ and lim u0 (c) = 0.

c→∞

c→0

Alternatively, we can think of the economy as consisting of a unit measure of identical households each with the instantaneous utility function given by (8.1). Population within each household grows at the rate n, starting with L (0) = 1, so that total population is L (t) = exp (nt) . All members of the household supply their labor inelastically. Consequently, we assume that each household maximizes overall utility U (0) at time t = 0 given by Z

∞

(8.2)

0

exp (− (ρ − n) t) u (c (t)) dt,

(8.3)

where c (t) is consumption per capita at time t, ρ is the subjective discount rate, and the eﬀective discount rate is ρ − n, since it is assumed that the household derives utility from the consumption of its additional members in the future as well. We assume throughout that Assumption 10 ρ > n. This assumption ensures that there is in fact discounting of future utility streams. Otherwise, (8.3) would have inﬁnite value, and standard optimization techniques would not be useful in determining what an optimal plan is (we would need to use over-taking type criteria etc.). More generally, there is something somewhat strange about models in which utility is 168

14.451: Introduction to Economic Growth equal to inﬁnity. Assumption 10 makes sure that in the model without growth, discounted utility is ﬁnite. When there is growth, we will strengthen this assumption. We start with an economy without any technological progress. Factor and product markets are competitive, and the production possibilities set of the economy is represented by the aggregate production function Y (t) = F [K (t) , L (t)] , which is a simpliﬁed version of the production function (2.1) used in the Solow growth model above, where the simpliﬁcation comes from the fact that there is no technology term. As in the analysis there, we impose the standard constant returns to scale and Inada assumptions embedded in Assumptions 1 and 2. The constant returns to scale feature enables us to work with the per capita production function f (·) such that, output per capita is given by y (t) ≡ Y (t) L (t) ∙ ¸ K (t) = F ,1 L (t) ≡ f (k (t)) , K (t) . L (t)

where, as before, k (t) ≡ (8.4)

Competitive factor markets then imply that, at all points in time, the rental rate of capital and the wage rate are given by: R (t) = FK [K (t), L(t)] = f 0 (k(t)). and w (t) = FL [K (t), L(t)] = f (k (t)) − k (t) f 0 (k(t)). 169 (8.6) (8.5)

14.451: Introduction to Economic Growth The household optimization side is more complicated, since each household will solve a continuous optimization problem in deciding how to use their assets and allocating consumption over time. To prepare for this, let us denote the asset holdings of the representative household at time t by A (t). Then we have the following law of motion for the total assets of the household ˙ (t) = r (t) A (t) + w (t) L (t) − c (t) L (t) A where c (t) is consumption per capita of the household, r (t) is the risk-free market rate of return on assets, and w (t) L (t) is the total labor income earnings of the household. Note that the r(t) is now a ﬂow return, not a gross return as used before. Deﬁning per capita assets as a (t) ≡ we obtain: a ˙ (t) = (r (t) − n) a (t) + w (t) − c (t) . (8.7) A (t) , L (t)

In practice, household assets can consist of capital stock, K (t), which they rent to ﬁrms and government bonds, B (t). In models with uncertainty, households would have a portfolio choice between the capital stock of the corporate sector and riskless bonds. Government bonds play an important role in models with uncertainty and heterogeneity, allowing households to smooth idiosyncratic shocks. But in representative household models without government, their only use is in pricing assets (for example riskless bonds versus equity etc.), since they have to be in zero net supply, i.e., total supply of bonds has to be B (t) = 0. Consequently, we will have that assets per capita are equal to the capital stock per capita (or the capital-labor ratio in the economy), i.e.: a (t) = k (t) . 170

14.451: Introduction to Economic Growth Moreover, since there is no uncertainty here and a depreciation rate of δ , the market rate of return on assets will be given by r (t) = R (t) − δ. (8.8)

The equation (8.7) is only a ﬂow constraint, and it is not suﬃcient to act as a proper budget constraint on the individual. To see this, consider a ﬁnite-horizon economy, ending at the time T . In this case, we could express the entire set of constraints on the household as a single budget constraint of the form: µZ T ¶ Z T c (t) L(t) exp r (s) ds dt + A (T ) 0 t µZ T ¶ µZ Z T = w (t) L (t) exp r (s) ds dt + A (0) exp

0 t

(8.9)

T

0

¶ r (s) ds ,

which requires the household’s discounted budget constraint to hold at time T (hence all income and expenditures are carried forward to date T units). Clearly, diﬀerentiating this expression and expanding L(t) gives (8.7). And yet (8.7) by itself does not guarantee that the level of A (T ) is such that this lifetime budget constraint holds. Therefore, in the ﬁnitehorizon, we would simply impose this lifetime budget constraint as a boundary condition. In the inﬁnite-horizon case, we need a similar boundary condition. This is generally referred to as the no-Ponzi-game condition, and takes the form µ Z t ¶ lim a (t) exp − (r (s) − n) ds ≥ 0.

t→∞ 0

(8.10)

This condition is stated as an inequality, to ensure that the individual does not asymptotically tend to a negative wealth. But we will see from the transversality condition of the individual problem that the individual would never want to have positive wealth asymptotically, so the no-Ponzi-game condition can be alternatively stated as: µ Z t ¶ lim a (t) exp − (r (s) − n) ds = 0.

t→∞ 0

(8.11)

171

14.451: Introduction to Economic Growth In what follows we will use (8.10), and then derive (8.11) using the transversality condition explicitly. The name no-Ponzi-game condition comes from the chain-letter schemes, which are sometimes called Ponzi games, where an individual can continuously borrow from a competitive ﬁnancial market (or more often, from unsuspecting souls that become part of the chain-letter scheme) and pay his or her previous debts using current borrowings. To understand where this form of the no-Ponzi-game condition comes from, multiply ´ ³ R T both sides of (8.9) by exp − 0 r (s) ds to obtain Z

T

=

Z

0 T

0

**µ Z t ¶ µ Z T ¶ c (t) L(t) exp − r (s) ds dt + exp − r (s) ds A (T ) 0 0 µ Z t ¶ w (t) L (t) exp − r (s) ds dt + A (0) ,
**

0

**then divide everything by L (0) and note that L(t) grows at the rate n, to obtain µ Z t ¶ µ Z T ¶ Z T c (t) exp − (r (s) − n) ds dt + exp − (r (s) − n) ds a (T ) 0 0 0 µ Z t ¶ Z T = w (t) exp − (r (s) − n) ds dt + a (0) .
**

0 0

**Now take the limit as T → ∞ and use the no-Ponzi-game condition (8.11) to obtain µ Z t ¶ µ Z t ¶ Z ∞ Z ∞ c (t) exp − (r (s) − n) ds dt = a (0) + w (t) exp − (r (s) − n) ds dt,
**

0 0 0 0

which essentially requires the discounted sum of expenditures to be equal to initial income plus the discounted sum of labor income. Therefore this equation is a direct extension of (8.9) to inﬁnite horizon. This derivation makes it clear that the no-Ponzi-game condition (8.11) essentially ensures that the individual’s lifetime budget constraint holds in inﬁnite horizon. 172

14.451: Introduction to Economic Growth

8.2

8.2.1

Characterization of Equilibrium

Deﬁnition of Equilibrium

We are now in a position to deﬁne an equilibrium in this dynamic economy. I will provide two deﬁnitions, the ﬁrst somewhat less formal, and second more useful in characterizing the equilibrium below. A competitive equilibrium of the Ramsey economy consists of paths of consumption, capital stock, wage rates and rental rates of capital, [C (t) , K (t) , w (t) , R (t)]∞ t=0 such that the representative household maximizes its utility given initial capital stock K (0) and the

∞ time path of prices [w (t) , R (t)]∞ t=0 , and the time path of prices [w (t) , R (t)]t=0 is such that

given the time path of capital stock and labor [K (t) , L (t)]∞ t=0 all markets clear. Notice that in equilibrium we need to determine the entire time path of real quantities and the associated prices. This is a very important point. In dynamic models whenever we talk of “equilibrium”, this refers to the entire path of quantities and prices. In some models, we will focus on the steady-state equilibrium, but equilibrium always refers to the entire path. Since everything can be equivalently deﬁned in terms of per capita variables, let me states the alternative deﬁnition in terms of those:

Deﬁnition 8 A competitive equilibrium of the Ramsey economy consists of paths of per capita consumption, capital-labor ratio, wage rates and rental rates of capital, [c (t) , k (t) , w (t) , R (t)]∞ t=0 such that the representative household maximizes (8.3) subject to (8.7) and (8.10) given initial capital-labor ratio k (0) and factor prices [w (t) , R (t)]∞ t=0 with the rate of return on assets r (t) given by (8.8), and factor prices [w (t) , R (t)]∞ t=0 are given by (8.5) and (8.6). 173

14.451: Introduction to Economic Growth

8.2.2

The Consumer Problem

Let us start with the problem of the representative consumer. From the deﬁnition of equilibrium we know that this is to maximize (8.3) subject to (8.7) and (8.11). Let us ignore (8.11) ﬁrst, and set up the current value Hamiltonian: ˆ (a, c, µ) = u (c (t)) + µ (t) [w (t) + (r (t) − n) a (t) − c (t)] , H with state variable a, control variable c and current-value costate variable µ. From Theorem 34, the following are necessary conditions: ˆ c (a, c, µ) = 0 = u0 (c (t)) − µ (t) , H ˆ a (a, c, µ) = −µ ˙ (t) + (ρ − n) µ (t) = µ (t) (r (t) − n) , H

t→∞

lim [exp (− (ρ − n) t) µ (t) a (t)] = 0.

and the transition equation. Notice that the transversality condition is written in terms of the current-value costate variable. ˆ (a, c, µ) is a concave function of (a, c), and thus from Theorem Moreover, for any µ (t), H 35, these conditions are suﬃcient for a solution. Rearranging the second condition, we have µ ˙ (t) = − (r (t) − ρ) , µ (t) (8.12)

which states that the multiplier changes depending on whether the rate of return on assets is currently greater than or less than the discount rate of the household. The ﬁrst condition, on the other hand, implies u0 (c (t)) = µ (t) . 174

note also that integrating (8.12).451: Introduction to Economic Growth To make more progress.we have σ u (t. s) = − As s ↓ t. s) → σ u (t) = − d log (c (s) /c (t)) . d log (u0 (c (s)) /u0 (c (t))) 1 u0 (c (t)) = . determines how willing individuals are to substitute consumption over time. 00 u (c (t)) c (t) εu (c (t)) This is not surprising. The intertemporal elasticity of substitution regulates the willingness of individuals to substitute consumption (or labor or any other attribute that yields utility) over time.12).13) is the elasticity of the marginal utility u0 (c(t)).14) (8. since the concavity of the utility function u (·). we have c ˙ (t) 1 = (r (t) − ρ) c (t) εu (c(t)) where εu (c (t)) ≡ − u00 (c (t)) c (t) u0 (c (t)) (8. u0 (c (t)) c (t) µ (t) Substituting this into (8. εu (c (t)) is also the inverse of the intertemporal elasticity of substitution. This elasticity for dates t and s > t is deﬁned as σ u (t. Next. let us diﬀerentiate this with respect to time and divide by µ (t). More importantly. thus the elasticity of marginal utility.14. which yields ˙ (t) u00 (c (t)) c (t) c µ ˙ (t) = . we have ¶ µ Z t (r (s) − ρ) ds µ (t) = µ (0) exp − 0 µ Z t ¶ 0 (r (s) − ρ) ds . = u (c (0)) exp − 0 175 . which plays a crucial role in most macro models.

Now recalling that the solution to the diﬀerential equation y ˙ (t) = b (t) y (t) is y (t) = y (0) exp we can integrate (8. 0 In that case. ´ ³ R t notice that the term exp − 0 r (s) ds is a present-value factor that converts a unit of income at time t to a unit of income at time 0. lim a (t) exp − t→∞ 0 which implies that the strict no-Ponzi condition. lim exp (− (ρ − n) t) a (t) u (c (0)) exp − t→∞ 0 ∙ µ Z t ¶¸ (r (s) − n) ds = 0. In the special case where r (s) = r. In particular. we can express the conversion factor between dates 0 and t as exp (−r ¯ (t) t) . we can deﬁne an average interest rate between dates 0 and t as 1 r ¯ (t) = t Z t r (s) ds. to obtain c (t) = c (0) exp µZ t t 0 ¶ b (s) ds . We can derive further results on the consumption behavior of households. this factor would be exactly equal to exp (−rt).451: Introduction to Economic Growth where the second line uses the ﬁrst optimality condition of the current-value Hamiltonian at time t = 0. we have ∙ µ Z t ¶¸ 0 (r (s) − ρ) ds = 0. But more generally.13). µZ 0 ¶ r (s) − ρ ds εu (c (s)) 176 .11) has to hold. Now substituting into the transversality condition.14. (8.

3 Equilibrium Prices Equilibrium prices are straightforward and are given by (8. r (t). In the special case where εu (c (s)) is constant.15) in the iso-elastic utility case also similarly generalizes.16) as the equilibrium version of the consumption growth equation.e. 177 .14.8). for example. i. θ Z and moreover. This implies that the market rate of return for consumers.451: Introduction to Economic Growth as the consumption function.5) and (8. we have c ˙ (t) 1 = (f 0 (k (t)) − δ − ρ) c (t) εu (c (t)) (8. this equation simpliﬁes to c (t) = c (0) exp µµ ¶ ¶ r ¯ (t) − ρ t . r (t) = f 0 (k (t)) − δ. the lifetime budget constraint simpliﬁes to Z Z ∞ 0 c (t) exp (− (¯ r (t) − n) t) dt = a (0) + ∞ 0 w (t) exp (− (¯ r (t) − n) t) dt. the path of consumption can be exactly solved out. (8. Thus once we determine c (0).13). the initial level of consumption..6). εu (c (s)) = θ. and substituting for c (t) into this lifetime budget constraint in this iso-elastic case. we obtain c (0) = ∞ 0 ¶ ¶ ∙ µ µ ¸ Z ∞ (1 − θ) r ¯ (t) ρ − + n t dt a (0) + exp − w (t) exp (− (¯ r (t) − n) t) dt θ θ 0 (8. 8.15) as the initial value of consumption. Substituting this into the consumer’s problem.2. is given by (8. Equation (8.

it is straightforward to see that these optimality conditions imply c ˙ (t) 1 = (f 0 (k (t)) − δ − ρ) . From Theorem 34. c. once again set up the current-value Hamiltonian. which in this case takes the form ˆ (k. the following are necessary conditions: ˆ c (k. µ) = u (c (t)) + µ (t) [f (k (t)) − (n + δ )k (t) − c (t)] . In particular. µ) = −µ ˙ (t) + (ρ − n) µ (t) = µ (t) (f 0 (k (t)) − δ − n) . so that it solves the problem max Z ∞ [k(t). To solve this problem. H t→∞ lim [exp (− (ρ − n) t) µ (t) k (t)] = 0. c.451: Introduction to Economic Growth 8. µ) = 0 = u0 (c (t)) − µ (t) . control variable c and current-value costate variable µ. H ˆ k (k. Going exactly through the same steps as before. subject to ˙ (t) = f (k (t)) − (n + δ )k (t) − c (t) k and k (0) > 0. suppose that the social planner gives exactly the same weights to people in diﬀerent generations.3 Optimal Growth Before characterizing the equilibrium further.14.c(t)]∞ t=0 0 exp (− (ρ − n) t) u (c (t)) dt. c. H with state variable k. c (t) εu (c (t)) 178 . it is useful to look at the optimal growth problem. deﬁned as the capital and consumption path chosen by a benevolent social planner trying to achieve a Pareto optimal outcome.

16).14. From (8. and as is the case there.4 Steady-State Equilibrium Now let us characterize the steady-state equilibrium (or equivalently the steady-state optimal allocation). this implies that irrespective of the exact utility function. In steady state. the equilibrium is Pareto optimal and coincides with the optimal growth path maximizing the utility of the representative household. and the natural Pareto allocation can be decentralized as a competitive equilibrium with exactly the initial endowments. with Assumptions 1.16). 2.17) which is the equivalent of the steady-state relationship in the discrete-time optimal growth model. This establishes that the competitive equilibrium is a Pareto optimum. (8. rather than the golden rule we saw in the 179 . This also corresponds to the modiﬁed golden rule. consumption per capita will be constant. we must have a capital-labor ratio k∗ such that f 0 (k ∗ ) = ρ + δ. lim k (t) exp − t→∞ 0 which is identical to (8. 8. This result is stated in the next proposition: Proposition 17 In the neoclassical growth model described above. 9 and 10. and the transversality condition ∙ µ Z t ¶¸ 0 (f (k (s)) − δ − n) ds = 0. it pins down the steady state capital-labor ratio only as a function of the production function.451: Introduction to Economic Growth which is identical to (8.11). thus c ˙ (t) = 0. the discount rate and the depreciation rate.

a steady state where the capital-labor ratio and thus output are constant necessarily satisﬁes the transversality condition. k∗ .451: Introduction to Economic Growth Solow model. the capital stock is chosen at a level that does not maximize steady-state consumption. with Assumptions 1. because earlier consumption is preferred to later consumption. c∗ . which means that the objective is not to maximize steady-state consumption. Given k∗ . but the steady-state capital-labor ratios determined diﬀerently. Rather than maximizing consumption. the steady-state consumption level is straightforward to determine as: c∗ = f (k∗ ) − (n + δ )k∗ .18) which is similar to the consumption level in the basic Solow model. (8.17). is given by (8. The steady-state consumption per capita. is uniquely determined by (8.14. Recall that transitional dynamics in the basic Solow model were given by a single diﬀerential equation with an initial condition. and is independent of the utility function. we can determine the transitional dynamics of this model. since the equilibrium is determined by two diﬀerential 180 .5 Transitional Dynamics Next. 9 and 10. given Assumption 10. the steady-state equilibrium capital-labor ratio. This is no longer the case. 2.18). This analysis therefore establishes: Proposition 18 In the neoclassical growth model described above. 8. This is because of discounting. but involves giving a higher weight to earlier consumption. Moreover.

5 and 6. but also a boundary condition at inﬁnity. what we want is saddle-path stability.14. c (t) εu (c (t)) Moreover.451: Introduction to Economic Growth equations.19) with initial value x (0). we have the following straightforward generalizations of Theorems 4 and 5: Theorem 37 Consider the following linear diﬀerential equation system x ˙ (t) = Ax (t) (8. Therefore. the consumption level (or equivalently the costate variable µ) is the control variable. we have an initial condition k (0) > 0. repeated here for convenience: ˙ (t) = f (k (t)) − (n + δ )k (t) − c (t) k and c ˙ (t) 1 = (f 0 (k (t)) − δ − ρ) . where x (t) ∈ Rn for all t and A is an n × n matrix. which involves the number of negative eigenvalues to be the same as the number of state variables. of the form ∙ µ Z t ¶¸ 0 (f (k (s)) − δ − n) ds = 0. Suppose that m ≤ n of the eigenvalues of A have negative real parts. lim k (t) exp − 0 t→∞ This combination of an initial condition and a transversality condition is quite typical for optimal control problems where we are trying to pin down the behavior of both state and control variables. It has to adjust in a way to satisfy the transversality condition at inﬁnity. In particular. This means that the notion of “stability” has to be diﬀerent from that of those in Theorems 4. rather than requiring all eigenvalues of the linear system or the linearized system to be negative. In particular. Then there exists an m-dimensional 181 . and its initial value c (0) (or equivalently µ (0)) is free.

Deﬁne A =∇F (x∗ ) . Theorem 38 Consider the following nonlinear autonomous diﬀerential equation x ˙ (t) = F [x (t)] (8. Then there exists an open neighborhood of x∗ . F (x∗ ) = 0.14. The ﬁrst one is simply by analyzing the above system diagrammatically. B (x∗ ) ⊂ Rn and an mdimensional manifold M ⊂ B (x∗ ) such that starting from any x (0) ∈ M .19) has a unique solution with x (t) → x∗ where x∗ is the steady state (zero) of the system given by Ax∗ = 0.20) has a unique solution with x (t) → x∗ . However. the diﬀerential equation (8.451: Introduction to Economic Growth manifold M of Rn such that starting from any x (0) ∈ M .20) where F : Rn → Rn and suppose that F is continuously diﬀerentiable. these two theorems state that only a lower-dimensional subset of the original space leads to stable solutions. the diﬀerential equation (8. Let x∗ be a zero of this system.e. since c (0) will adjust in order to place us on exactly such a lower-dimensional subset of the original space. This is done in the next picture: 182 . with initial value x (0).. There are two ways of seeing this. and suppose that m ≤ n of the eigenvalues of A have negative real parts and the rest have positive real parts. in this context this is exactly what we require. Put diﬀerently. i.

The reason why the c ˙ = 0 locus is just a vertical line simply follows from the fact that only the unique level of k ∗ given by (8. kgold . steady-state consumption is low. then the steady-state consumption is again low. If the capital stock is too low. which maximizes the state-state consumption per capita. The shape of the ﬁrst one can be understood by analogy to the diagram where we saw the golden rule.14. There exists a unique level. the rest of the diagram can be completed by looking at the direction of motion according to the diﬀerential equations. it is clear that there exists a unique stable arm.17) can keep per capita consumption constant. and if the capital stock is too high. The vertical line. on the The inverse U-shaped curve is the locus of points where k other hand.451: Introduction to Economic Growth ˙ = 0. the lower-dimensional manifold 183 . is the locus of points where c ˙ = 0. Given this direction of movements. Once these two loci are drawn.

thus violating the transversality condition. consumption would reach zero. if k (0) < k∗ . in ﬁnite time. The next important observation is that the initial consumption level c (0) has to adjust to be on this stable arm. with Assumptions 1. and eventually reach zero consumption or zero capital stock as shown in the ﬁgure. This establishes: Proposition 19 In the neoclassical growth model described above. and then (k. in ﬁnite time. c (t) εu (c (t)) 184 .451: Introduction to Economic Growth tending to the steady state. then k (t) ↑ k∗ and c (t) ↑ c∗ . 2. An alternative way of establishing the same result is by linearizing the set of diﬀerential equations. then k (t) ↓ k∗ and c (t) ↓ c∗ . there exists a unique equilibrium path starting from any k (0) > 0 and converging to the unique steady-state (k∗ . c) will monotonically travel along this arm towards the steady state. Recall the two diﬀerential equations determining the equilibrium path: ˙ (t) = f (k (t)) − (n + δ )k (t) − c (t) k and c ˙ (t) 1 = (f 0 (k (t)) − δ − ρ) . 9 and 10. All points away from this stable arm diverge. the capital stock would reach 0 with positive consumption. This establishes that the transitional dynamics in the neoclassical growth model will take the following simple form: c (0) will “jump” to the stable arm. If it were below it. c∗ ) with k∗ given by (8. Moreover. whereas if k (0) > k∗ . thus capital would accumulate continuously. and looking at their eigenvalues.17).14. To see this note that if it were above it. violating feasibility.

from (8. However.14. the local analysis also leads to the same conclusion.451: Introduction to Economic Growth Linearizing these equations around the steady state (k ∗ . c∗ ).17). one negative and one positive. exactly as the stable arm in the above ﬁgure. A (t) L (t)] . there are two real eigenvalues. where A (t) = exp (gt) A (0) . the local analysis can only establish local stability.21) . det ⎝ ∗ 00 ∗ c f (k ) 0−ξ εu (c∗ ) It is straightforward to verify that. we have (suppressing time dependence) ˙ = constant + (f 0 (k∗ ) − n − δ ) (k − k∗ ) − c k c∗ f 00 (k∗ ) (k − k∗ ) . c ˙ = constant + εu (c∗ ) Moreover.6 Technological Change and the Canonical Neoclassical Model The above analysis was for the neoclassical growth model without any technological change. f 0 (k∗ ) − δ = ρ. Therefore. Let us now extend the production function to: Y (t) = F [K (t) . so the eigenvalues of this two-equation system are given by the values of ξ that solve the following quadratic form: ⎞ ⎛ ρ − n − ξ −1 ⎠ = 0. since c∗ f 00 (k∗ ) /εu (c∗ ) < 0. 8. This implies that there exists a one dimensional stable manifold converging to the steady state. whereas the above analysis established global stability. 185 (8.

8). technological change.1 A (t) L (t) ³ ´ ˆ (t) .22) where now ˆ (t) ≡ k is the capital to eﬀective labor ratio. This is a consequence of Theorem 7 above. We continue to adopt all the other assumptions.451: Introduction to Economic Growth Notice that the production function (8. or Harrodneutral. from (8. ≡ f k K (t) . implies that r (t) has to be constant. Only purely labor-augmenting technological change is consistent with balanced growth. taking into account that eﬀective labor is increasing because of labor-augmenting technological change. balanced growth requires that consumption and output grow at a constant rate. We deﬁne balanced growth as growth consistent with the Kaldor facts of constant capital-output ratio and capital share in national income. Assumption 10 will be strengthened further in order to ensure ﬁnite discounted utility in the presence of sustained economic growth. has to be constant. Now let us deﬁne y ˆ (t) ≡ Y (t) A (t) L (t) ∙ ¸ K (t) = F . which was proved in the context of the constant savings rate model.14. R (t). In addition to the assumption on technology. which. The Euler 186 . in particular Assumptions 1. The constant returns to scale feature again enables us to work with normalized variables. we also need to impose a further assumption on preferences in order to ensure balanced growth. 2 and 9. A (t) L (t) (8. but equally applies in this context. These two observations together also imply that the rental rate of return on capital. In addition.21) imposes purely labor-augmenting.

1−θ (8. When θ = 0. these represent linear (risk-neutral) preferences.23) as the canonical model. Therefore. More speciﬁcally. Given this restriction.14. we might as well start with a utility function that has this feature throughout.24) .451: Introduction to Economic Growth equation implies that 1 c ˙ (t) = (r (t) − ρ) . then c εu (c (t)) → εu . and inﬁnitely unwilling to substitute consumption over time. whereas when θ = 1.23) I refer to this model. these preferences become inﬁnitely risk-averse. given by ⎧ ⎨ c(t)1−θ −1 if θ 6= 1 and θ ≥ 0 1−θ u (c (t)) = . balanced growth is only consistent with utility functions that have asymptotically constant elasticity of marginal utility of consumption.. since it is the model used in almost all applications with steady growth (unless non-balanced growth is the purpose as will be discussed in some of the structural change models below). i. if the elasticity of marginal utility of consumption is asymptotically constant. the unique utility function with this feature is the CRRA preferences. we now assume that the economy admits a representative consumer with CRRA preferences Z ∞ 0 exp (−(ρ − n)t) c (t)1−θ − 1 dt. εu . with labor-augmenting technological change and CRRA preference as given by (8.e. ⎩ ln c(t) if θ = 1 where the elasticity of marginal utility of consumption. we have log preferences. the Euler equation in this case takes the simpler form: 1 c ˙ (t) = (r (t) − ρ) . As noted above. c (t) θ 187 (8. c (t) εu (c (t)) ˙ (t) /c (t) → gc is only possible if If r (t) → r∗ in BGP (balanced growth path). As θ → ∞. Clearly. is given by the constant θ.

25) In addition. c (t) will grow. lim k 0 t→∞ (8. = θ Moreover.8). so ³ ´ ˆ (t) − δ r (t) = f 0 k Since in steady state c ˆ (t) must remain constant. r (t) is still given by (8. let us deﬁne c ˆ (t) ≡ C (t) A (t) L (t) c (t) ≡ .451: Introduction to Economic Growth Let us ﬁrst characterize the steady-state equilibrium in this model with technological progress. we have c ˆ (t) c ˙ (t) ≡ −g c ˆ (t) c (t) 1 (r (t) − ρ − θg) . Since with technological progress there will be growth in per capita income. in analogy with y ˆ (t). Instead. we have · ³ ´ ˆ ˆ ˆ (t) . therefore r (t) = ρ + θg 188 . in turn. for the accumulation of capital stock. can be expressed as ½ µ Z th ³ ´ i ¶¾ 0 ˆ (t) exp − ˆ (s) − g − δ − n ds f k = 0.14. k (t) = f k (t) − c ˆ (t) − (n + g + δ ) k · The transversality condition. In particular. A (t) We will see that this normalized consumption level will remain constant along the BGP.

451: Introduction to Economic Growth or ˆ∗ uniquely. we have the following immediate generalization of Proposition 18: 189 . recall that in steady state we have r = ρ + θg and the growth rate of output is g + n. and they will also be related to issues of “dynamic eﬃciency” as we will see below.. in a way which pins down the steady-state value of the normalized capital ratio k similar to the model without technological progress.27) while per capita consumption grows at the rate g . 0 t→∞ which can only be the case if the integral within the exponent goes to zero. The level of normalized consumption is then given by ³ ´ ˆ∗ − (n + g + δ ) k ˆ∗ . if ρ − (1 − θ) g − n > 0. We will encounter conditions like this all throughout. Substituting (8.26) (8. we have to make sure that the transversality condition is in fact satisﬁed. Alternatively. Assumption 11 is equivalent to requiring that r > g + n.25).26) into (8.14. The only additional condition in this case is that because there is growth.e. c ˆ =f k ∗ ³ ´ ˆ∗ = ρ + δ + θg. Therefore. For now. or alternatively if the following assumption is satisﬁed: Assumption 11 ρ − n > (1 − θ) g. we have ½ µ Z t ¶¾ ˆ lim k (t) exp − [ρ − (1 − θ) g − n] ds = 0. f0 k (8. i. Note that this assumption strengthens Assumption 10 when θ < 1.

2. Morek ˆ (t) converging to the unique steady-state k ˆ∗ with k ˆ (0) < k ˆ∗ . 190 . so the willingness of individuals to substitute consumption today for consumption tomorrow determines how much they will accumulate and thus the equilibrium capital to eﬀective labor ratio. 9 and 11 hold. Then there exists a unique equilibrium path of normalized capital and consumption.26). c ˆ∗ given by (8. if k ˆ (t) ↑ c ˆ∗ . A similar analysis to before also lead to an immediate generalization of Proposition 19. then k ˆ (t) ↑ k ˆ∗ and c ˆ (0) > k ˆ∗ . Suppose that Assumptions 1. 2. c ˆ∗ . which is stated here.451: Introduction to Economic Growth Proposition 20 Consider the neoclassical growth model with labor augmenting technological progress at the rate g and preferences given by (8. then k ˆ (t) ↓ k ˆ∗ and over.26) depends on the elasticity preferences is no longer the case. ³ ´ ³ ´ ˆ (t) .23).14. whereas if k c ˆ (t) ↓ c ˆ∗ . given by (8. The proof is left as at home work exercise. Interestingly.23). 9 and 11 hold. Proposition 21 Consider the neoclassical growth model with labor augmenting technological progress at the rate g and preferences given by (8. but the next ﬁgure gives the sketch already. The reason for this is that there is now growth.26). since now k of marginal utility (or the inverse of the intertemporal elasticity of substitution). the results that the steady-state capital-labor ratio was independent of ˆ∗ given by (8. Then there exists a unique balanced growth path equilibrium with a normalized ˆ∗ . Suppose that Assumptions 1. θ. and output per capita and consumption capital to eﬀective labor ratio of k per capita grow at the rate g .

451: Introduction to Economic Growth It is also useful to brieﬂy look at an example with Cobb-Douglas technology. In this case. Assume that the production function is given by F (K. = αk c ˆ θ · and the accumulation equation can be written as ˆ k ˆα−1 − δ − g − n − =k ˆ k 191 c ˆ . f k ˆα−1 − δ .14. Example 3 Consider the model with CRRA utility and labor-augmenting technological progress at the rate g . so that ³ ´ ˆ =k ˆα . AL) = K α (AL)1−α . the Euler equation becomes: and thus r = αk · ´ c ˆ 1 ³ ˆα−1 − δ − ρ − θg . ˆ k .

29) The two diﬀerential equations (8. in particular. Therefore.28) (8. If we were to go back to the proximate causes of diﬀerences in income per capita or growth across countries. 8. you will be asked to complete this example for the special case in which θ → 1 (i. θ θ · · · (8.451: Introduction to Economic Growth ˆ k ˆ. 1/θ. ρ. these ˆ and x ≡ k ˆα−1 . 192 . by the growth rate of labor-augmenting technological progress. on the intertemporal elasticity of substitution. ˆ z c ˆ k thus 1 z ˙ = (αx − δ − ρ − θg ) − x + δ + g + n + z z θ 1 ρ = ((α − θ)x − (1 − θ)δ + θn) − + z. and naturally the form of the production function f (·).14.28) and (8. the population growth rate. this model would give us a way of understanding those diﬀerences only in terms of preference and technology parameters. on the other hand. The level of income. the discount rate.e. which implies that x/x Now deﬁne z ≡ c ˆ/k ˙ = (α − 1) k/ two equations can be written as x ˙ = − (1 − α) (x − δ − g − n − z ) x ˆ z ˙ c ˆ k = − ..29) together with the initial condition x (0) and the transversality condition completely determine the dynamics of the system.7 The Role of Policy In the above model. In Problem Set 4. the rate of growth of per capita consumption and growth are determined exogenously. depends on preferences. δ . the depreciation rate. log preferences). n.

and since from Assumption 1. in terms of normalized capital. it reduces k capital accumulation. This implies c ˆ (t) c ˙ (t) ≡ −g c ˆ (t) c (t) 1 (r (t) − ρ − θg ) . at least according to the institutions view. Let us do this in the simplest possible way here and suppose that returns on capital net of depreciation are taxed at the rate τ and the proceeds of this are redistributed back to the consumers. 193 A higher tax rate increases the right hand side. Therefore.451: Introduction to Economic Growth However. we will discuss how large these eﬀects can be and whether they could account for the diﬀerences in cross-country incomes. k (t) = f k (t) − c ˆ (t) − (n + g + δ ) k but the rental rate of return is now ´ ´ ³ ³ 0 ˆ r (t) = (1 − τ ) f k (t) − δ . the capital accumulation equation. f 0 (·) is · . which. In that case. still remains · ³ ´ ˆ ˆ ˆ (t) . In the next section. higher taxes on capital have the eﬀect of depressing decreasing. (1 − τ ) f 0 k = θ so that the steady-state capital to eﬀective labor ratio is given by ³ ´ ˆ∗ = δ + ρ + θg . f0 k 1−τ ˆ∗ . the model can be easily enriched to include policy variables.14. play an equally important role in accounting for diﬀerences in physical (and human) capital and technology across countries. = θ ³ ³ ´ ´ ´ 1³ ˆ (t) − δ − ρ − θg .

Although this is a plausible starting point. Chad Jones uses data from the Summers-Heston data set on the price of investment goods relative to consumption goods and shows that there are large diﬀerences in the relative price of capital goods (compared to consumption goods).14. corruption or other policy diﬀerences. once you look at the data. but for now let us stick with the traditional approach and think of diﬀerential policies aﬀecting the relative price of capital. they were imposed on the amount of investment.8. for example. Suppose that all countries admit a representative consumer with identical preferences. 194 . Chari.451: Introduction to Economic Growth For now. We will see this in the next section. Kehoe and McGrattan (1997). This has led a number of economists. let us follow Jones (1995) and Chari.8 8. he also shows that a high relative price of capital goods is associated with low growth over the postwar period. Kehoe and McGrattan (1997) or Parente and Prescott (1994) to argue that a major diﬀerence across countries is the extent of distortions arising from taxes. 8. but from the fact that consumption goods are cheaper.1 Quantitative Evaluations Policy Diﬀerences For a qualitative evaluation of the eﬀect of policy diﬀerences. Imagine that the main policy diﬀerence across countries is in terms of tax structure that aﬀects the the relative price of capital goods. the diﬀerences in the relative price come not from the fact that investment goods are much more expensive in some countries. We will discuss this later below. we can also note that similar results would be obtained if instead of taxes being imposed on returns from capital. which aﬀect the relative price of capital.

31) with Hj representing exogenously given stock of eﬀective labor (human capital). which implicitly assumes that τ j Ij is wasted. There is no population growth. (8. This is without any major consequence. for example because of policies or diﬀerences in institutions/property rights enforcement. so we do not have to worry about the distribution of income in the economy.32) is still Yj . so Cj interchangeably refers to total or per capita consumption. K The only diﬀerence across countries is in the budget constraint for the representative consumer. as noted in Theorem 9 above. All countries also have access to the same production technology given by the Cobb-Douglas production function 1−α Yj = Kj (AHj )α . which takes the form (1 + τ j ) Ij + Cj ≤ Yj .32) where τ j is the tax on investment. Note that the right hand side variable of (8. rather than simply redistributed to some other agents in the economy.14. CRRA preferences as in (8. 195 .30) where j ∈ J denotes country j . This tax varies across countries.30) have the nice feature that they can be exactly aggregated across individuals. (8.451: Introduction to Economic Growth given by Z ∞ 0 exp (−ρt) 1−θ Cj −1 dt. Notice that 1 + τ j is also the relative price of investment goods (relative to consumption goods): one unit of consumption goods can only be transformed into 1/ (1 + τ j ) units of investment goods. The accumulation equation is ˙ j = Ij − δKj . since. 1−θ (8.

14.31) the capital output ratio is simply K/Y = (K/AH )α ).33) So countries that tax investment.30) subject to (8.32) and the capital accumulation equation. Cj θ (1 + τ j ) Kj Consider the steady state. the steady state corre˙ j /Cj = 0. 196 . at a higher rate will be poorer. what matters is the savings rate.) This immediately implies that Kj = (1 − α)1/α AHj [(1 + τ j ) (ρ + δ)]1/α So countries with higher taxes on investment will have a lower capital stock in steady state. Because A is assumed to be constant. they will also have lower capital per worker. With the same steps as above. Equivalently.31). in the Solow growth model. the Euler equation of the representative consumer is ¶α µ µ ¶ ˙j C 1 (1 − α) AHj = −δ−ρ .451: Introduction to Economic Growth The competitive equilibrium can be characterized as the solution to the maximization of (8. The advantage of using the neoclassical growth model for quantitative evaluation relative to the Solow growth model is that the extent to which diﬀerent types of distortions (here captured by the tax rates on investment) will aﬀect income and capital accumulation is determined endogenously. so we would need other evidence to link taxes or distortions to savings (or to other determinants of income per capita such as technology). and comparing two countries with diﬀerent taxes (but the same human capital). we obtain the relative incomes as Y (τ ) = Y (τ 0 ) µ 1 + τ0 1+τ α ¶ 1− α (8. either directly or indirectly. In contrast. (Alternatively. we could have A growing at a constant rate and C ˙ j /Cj sponds to C equal to the growth rate of A. Now substituting this into (8. or a lower capital output ratio (using (8.

with Cobb-Douglas production function.14. are unlikely to account for anywhere near as large diﬀerences in income per capita as we observe in practice. The Summers-Heston data suggest that there is a large amount of variation in the relative price of investment goods. 8.33) implies that the income gap between two such countries should be approximately threefolds: Y (τ ) ≈ (8)1/2 ≈ 3. these authors start from a one-sector model. diﬀerences in capital-output ratios or capital-labor ratios caused by taxes or tax-type distortions. Such diﬀerences do not feature in this model. The discussion above showed that diﬀerences in income per capita across countries cannot be accounted for by diﬀerences in capital per worker alone. This is not surprising. Nevertheless.451: Introduction to Economic Growth How large is this eﬀect? Or can such policy diﬀerences have quantitatively large eﬀects generating income diﬀerences comparable to what we observe in practice? Recall that a plausible value for α is 2/3.2 Extensions Basically. equation (8. But there is a constraint in this exercise: the share of capital in GDP is 197 . to explain such large diﬀerences in income per capita across countries. many authors have tried to use this model to go further. Then. Therefore. and try to generate large responses to distortions. the simplest model does not provide a good starting point. countries with the highest relative price of investment goods have almost eight times as high a value as countries with the lowest relative price. is equal to α. Instead. we need sizable diﬀerences in the eﬃciency with which these factors are used. since this is the share of labor income in national product which. For example. using α = 2/3. even very large diﬀerences in taxes or distortions.8. Y (τ 0 ) Therefore.

and in the simple Cobb-Douglas production function as in (8. but accumulates in exactly the same way as physical capital. and using the numbers implied by Mankiw. (8. H where X denotes investment in human capital. With this reasoning. They therefore conclude that the augmented Solow model is capable of explaining income diﬀerences across countries quantitatively based on distortions on investment.33)). so distortions that aﬀect capital can only have a small eﬀect on income. in this setup this means that variations in capital will have only a small eﬀect on income. Kehoe and McGrattan.451: Introduction to Economic Growth 1/3. A share of 2/3 for the accumulable factors in GDP is too high.14. Romer and Weil’s regression analysis discussed above.31). One line of attack is taken by Chari. this also turns out to determine the elasticity of output to distortions (this elasticity is simply (1 − α) /α as shown by (8. However. they posit (1 + τ j ) (Ij + Xj ) + Cj ≤ Yj and ˙ j = Xj − δHj .33) implies income diﬀerences as large as 64 fold. they take α = 1/3 (or they take the share of accumulable factors in GDP to be 2/3). This is intuitive: if capital only has a small share. In particular. Romer and Weil’s analysis. this conclusion is subject to exactly the same caveats as Mankiw. and implies implausibly large eﬀects of education on income as pointed out above. who suggest that the correct value for α is 2/3. In this case. They think that human capital is not exogenous. 198 .

Therefore. they argue that technology diﬀerences arise because there are barriers to technology adoption. Essentially. Consequently. Moreover. In particular. Each ﬁrm can ¯ ﬁrms in this economy). Consistent with the evidence presented above. and interpret a more broad concept of capital is “technology”. there is a sense in which what is being done here is to add the degree of freedom. is very similar to the tax-type distortions aﬀecting physical capital (and human capital) decisions in the neoclassical model.14. However. inducing economies with worse distortions not to adopt superior technologies. and I will treat it that way here. with a neoclassical production function. Suppose that output is given by Yt = At Nt where At is technology/knowledge which will be accumulated endogenously.9 Variants of the Neoclassical Model Parente-Prescott argue that the simple neoclassical model is not suﬃcient to account for the large diﬀerences in income per capita across countries. the ParentePrescott formulation does this while keeping exogenous growth. Their approach for technology diﬀerences. 199 . however. this explanation circumvents the problems of the neoclassical models without being forced to increase the “share of capital” in national income. Here is a very simple version of their model. This limit on ﬁrm¯ workers (so there will be Nt /N at most employ N level employment is imposed because the production technology exhibits increasing returns to scale.451: Introduction to Economic Growth 8. it is really a variant of it. they suggest that we have to take diﬀerences in technology into account. even though Parente-Prescott argue that their model is diﬀerent from the neoclassical model. this explanation turns technology into an accumulable factor. and otherwise all workers would be employed in one ﬁrm.

Denote Zt ≡ 1/θ (1−θ)/θ At Tt 1/θ (1−θ)/θ as the “eﬀective knowledge stock”.14. the investment required to improve technology from At to At+1 is Xt = φ −1 Z At+1 At µ S Tt θ ¶ 1− θ dS Intuitively. In particular. we have the law of motion of this knowledge stock as Zt+1 ≡ At+1 / (1 + g) Tt . Then. So Zt+1 = µ0 Zt + φXt as the law of motion. a high level of φ corresponds to better “technology” for absorbing world knowledge. which progresses exogenously as follows Tt = T0 (1 + g )t In order to beneﬁt from world technology. the budget constraint of the representative consumer is now Ct + Xt ≤ Yt . where µ0 is a constant. Solving this integral. As before.451: Introduction to Economic Growth In acquiring their technology. This equation makes it clear that the modeling 200 . each incremental improvement between At and At+1 costs an amount that depends on the distance of this improvement to the frontier technology. and also a shift parameter φ. and a low level of φ corresponds to signiﬁcant distortions in the process of technology adoption. we obtain Xt = φ −1 (1/θ) Tt At+1 − At 1/θ 1/θ (1−θ)/θ Clearly. countries/ﬁrms beneﬁt from world knowledge. countries/ﬁrms need to undertake some investments.

Parente and Prescott take us back to a production function of the form Y = K 0. for example θ = 0. if everything was priced according to marginal product–but because of the increasing returns to scale this is not the case). As a result.3 L0. except that Xt has replaced investments and Zt has replaced the capital stock. many models of endogenous technological progress will have a similar ﬂavor of technology being accumulated by purposeful investments.. they obtain signiﬁcantly larger eﬀects of distortions on income than implied by the neoclassical production function with Y = K 0. and knowledge is just another input. However.451: Introduction to Economic Growth question is very similar to a neoclassical model. Although the explanation is plausible. but they will allow for 201 .3 but with K replaced by Z .14.7. as in the neoclassical model with a small share of capital in national product. If θ is small. we have knowledge being accumulated. the model does not generate further insights than the statement that distortions lead to lower input use. As we will see below. Denoting output per capita in country j by y j . corresponding to a large share of payments to technology in GDP.g.7 L0. and θ is assumed to be large so knowledge is taken to be very important in production (e. output per capita diﬀerences in two economies can be expressed as functions of their knowledge stocks. So therefore the basic diﬀerence between this model and the standard neoclassical model is that here instead of capital. this economy will behave similar to the neoclassical model with a capital share equal to 2/3. In other words. by introducing the knowledge stock and increasing returns to scale. there will only be quantitatively small diﬀerences in output per capita across countries. we have !θ Ã j yt Ztj = . if θ is large. 0 j0 yt Ztj Will diﬀerences in φ now lead to large output diﬀerences? The answer depends on θ. Then.7 .

14.451: Introduction to Economic Growth much richer interactions. 202 .

These models. are referred to as overlapping generations models. 203 .Chapter 9 Growth with Overlapping Generations The models analyzed so far were assumed to admit a representative household or consumer. because they provide a tractable alternative to the inﬁnite-horizon representative agent models. the assumption of a representative household is not appropriate. ﬁrst analyzed by Paul Samuelson and then later Peter Diamond. second. since diﬀerent generations are born at diﬀerent points in time. These models are useful in providing us a tractable framework for the analysis of capital accumulation and neoclassical growth. Moreover. and third. But one speciﬁc set of circumstances where we have to depart from this assumption is the one where we look at an economy in which new households (individuals) arrive over time. In many situations. This was already discussed above. however. they allow a discussion of national debt and Social Security type issues. For economic growth. these models are useful. because they have some very diﬀerent implications. ﬁrst. they had the nice feature that the competitive equilibrium coincided with the natural Pareto optimal allocation.

he enjoys the consumption of commodities with the same index as his index and the next indexed commodity. j ∈ N. i. constitute a competitive equilibrium. denoted by x ¯. that is.1 Problems of Inﬁnity Let us start considering the following economy analyzed by Shell (1971).e. Assume all agents behave competitively (alternatively. Agent i has preferences given by: ui = ci + ci+1 . Proof. Moreover. Let us choose the price of the ﬁrst commodity as the numeraire. Consider the following static economy with a countably inﬁnite number of agents. the endowment vector ω of the economy is as follows: each agent has one unit endowment of the commodity with the same index as his. At p ¯. The following is straightforward to see: Proposition 22 In the above-described economy. we can have more than one agent of each type in order to ensure this). 204 .451: Introduction to Economic Growth 9.. This implies that consuming his endowment is optimal for each individual. i ∈ N and a countably inﬁnite number of commodities. p0 = 1. each individual has income equal to a 1. x ¯. establishing that p ¯ and no trade. thus the budget constraint of the form ci + ci+1 ≤ 1. the price vector p ¯ such that p ¯j = 1 for all j ∈ N is a competitive equilibrium price vector and induces an equilibrium with no trade.14.

When there are inﬁnite number of commodities and inﬁnite number of individuals. To see this. in this economy the First Welfare Theorem. recall also that Theorem 11 did not make use of summations in the same way as Theorem 10. This is in fact true.14. even without any change in endowments: 205 . The same issue of Pareto suboptimality will arise in this overlapping generations economy. summing over the value of the consumption basket of all individuals may (will often) lead to inﬁnite sums. while competitive equilibrium may be suboptimal. So one might conjecture that in this model. This is the reason why Theorem 10 was stated under the assumption of ﬁnite number of commodities. instead. This is a feasible allocation. This establishes Proposition 23 In the above-described economy. Therefore. Theorem 10. and the following proposition shows how the allocation x ˜ can be decentralized as a competitive equilibrium. Pareto optima must be decentralizable. and individual i = 0 is strictly better oﬀ. the competitive equilibrium at (p ¯. which is convex. x ¯) is not Pareto optimal. it made use of convexity. the competitive equilibrium in Proposition 22 is not Pareto optimal. and consumer i > 0 consumes one unit of good i + 1. consider the following alternative allocation. and the proof of Theorem 10 does not work. referred to as x ˜. The reason why this admittedly artiﬁcial economy is relevant is that it is isomorphic to the overlapping generations economy which we will analyze next. All consumers with i > 0 are as well oﬀ in this allocation as in the competitive equilibrium at p ¯.451: Introduction to Economic Growth However. This ﬁniteness assumption was used in the proof in making sure that summations existed and were ﬁnite. does not hold. Consumer i = 0 consumes one unit of good j = 0 and one unit of good j = 1. However.

individual 0 has a budget set c0 + c1 ≤ 2. all individuals born at time t live for dates t and t + 1. time is discrete and runs to inﬁnity. For example. Proof. Consequently. 9. while all other agents i have one unit of good i + 1. For now let us assume 206 .1 Demographics. Consider the following reallocation of the endowment vector ω: the endowment of agent i ≥ 1 is given to agent i − 1. thus chooses c0 = c1 = 1.451: Introduction to Economic Growth Proposition 24 In the above-described economy. At the price vector p ¯ . establishing that x ˜ is a competitive equilibrium. Preferences and Technology In this economy.14. at the new endowment vector ω ˜ . individual i = 0 has one unit of good j = 0 and one unit of good j = 1. Each individual lives for two periods. thus each i > 0 is happy to consume one unit of the good ci+1 . and p ¯ is such that p ¯j = 1 for all j ∈ N.2. there exists a reallocation of the endowment vector ω to ω ˜ . x ˜) that is Pareto optimal where x ˜ is as described above. 9. and an associated competitive equilibrium (p ¯. All other agents have budget sets given by ci + ci+1 ≤ 1.2 Overlapping Generations and Overaccumulation We now discuss the following canonical two-period overlapping generation economy. which is within his budget set and gives as high utility as any other allocation within his budget set.

The production side of the economy is the same as before. Let us also assume that there is population growth. (9. (9.e.14. so that total population is L (t) = (1 + n)t L (0) .2) where f (k) is the standard per capita production function described above.3) . i. earning the equilibrium wage rate w (t). and supply one unit of labor inelastically. r(t) = R (t) = f 0 (k (t)) . we have that the rate of return to saving equals the rental rate of capital.451: Introduction to Economic Growth a general utility function for individuals born at date t. As a result. 1) is the discount factor. and c2 (t + 1) this the consumption during the second period of his life (at date t + 1). c1 (t) denotes the consumption of the individual born at time t during the ﬁrst period of his life (which is at date t). of the form U (t) = u (c1 (t)) + βu (c2 (t + 1)) .. Also assume that capital fully depreciates after being used. satisfying Assumptions 1 and 2. Individuals can only work in the ﬁrst period of their lives. and represented by the standard constant returns to scale aggregate production function. Also β ∈ (0. characterized by a set of competitive ﬁrms.1) where u (·) satisﬁes the conditions in Assumption 9. and the wage rate is w (t) = f (k (t)) − k (t) f 0 (k (t)) . 207 (9.

since with negative savings. this is a solution to the following maximization problem c1 (t). I have not imposed the constraints that s (t) ≥ 0.c2 (t+1).2. we also obtain the following implicit solution for savings s (t) = s (w (t) . It is clear that both constraints will hold as equalities given that u (·) is strictly increasing.4) Solving these equations for consumption and thus for savings.s(t) max u (c1 (t)) + βu (c2 (t + 1)) subject to c1 (t) + s (t) ≤ w (t) and c2 (t + 1) ≤ R (t + 1) s (t) . Then the ﬁrst-order condition for a maximum implies: u0 (c1 (t)) = βR (t + 1) u0 (c2 (t + 1)) . (9.451: Introduction to Economic Growth 9. R (t + 1)) . individuals would violate their secondperiod budget constraint (given non-negativity of consumption). (9.14.). 208 . and may be increasing or decreasing in its second argument. Denoting savings by s (t).5) The function s (·.2 Consumption Decisions Let us start with the individual consumption decisions. where we are using the convention that old individuals rent their savings of time t as capital to ﬁrms at time t + 1. The second constraint incorporates the notion that individuals will only spend money on their own end of life consumption (there is no consumption term for descendants etc. so receive the gross rate of return R (t + 1). ·) is increasing in its ﬁrst argument.

Writing everything in terms of per worker units. the factor price sequence {R (t) . Therefore.5).e.14.6) as the fundamental law of motion of the overlapping generations economy.2) and (9. the law of motion of the capital stock is given by K (t + 1) = N (t) s (w (t) .3). which can lead to a unique stable equilibrium. i. R (t + 1)) . we obtain the k (t + 1) = s (f (k (t)) − k (t) f 0 (k (t)) . and multiple steady states are possible.3). A steady state is given by a solution to this equation such that k (t + 1) = k (t) = k ∗ .. The next ﬁgure shows some potential plots of the equation (9. w (t)}∞ t=0 is given by (9.2.4) and (9. Consequently. given the full depreciation assumption. to multiple equilibria. f 0 (k∗ )) 1+n (9. simply by the savings of the newly born at time t. ·) can take essentially any form. or to an equilibrium with zero capital stock. R (t + 1)) 1+n or substituting for R (t + 1) and w (t) from (9.7) Since the savings function s (·. the diﬀerence equation (9.451: Introduction to Economic Growth 9. while individual consumption and saving decisions are given by (9. 209 . or in words. N (t).6) can lead to quite complicated dynamics. f 0 (k (t + 1))) 1+n (9.6).3 Equilibrium An equilibrium in this economy is an allocation in which ﬁrms maximize and consumers optimize.2) and (9. this implies k (t + 1) = s (w (t) . k∗ = s (f (k∗ ) − k∗ f 0 (k∗ ) .

so that f (k ) = kα Everything else is the same as above.451: Introduction to Economic Growth 9. (9. let us now specialize the above setup by assuming CRRA utility functions. Furthermore. in particular.: c1 (t)1−θ − 1 +β U (t) = 1−θ Ã c2 (t + 1)1−θ − 1 1−θ ! . assume that technology is Cobb-Douglas.14. c1 (t) 210 .8) where θ > 0.2.4 More Speciﬁc Utility Functions To get more insights. This simpliﬁes the ﬁrst-order condition for consumer optimization and implies c2 (t + 1) = (βR (t + 1))1/θ . 1). β ∈ (0.

14. which was c/c ˙ = (r − ρ)/θ. The case of θ = 1. ∂w (t) ψ (t + 1) µ ¶ s (t) ∂s (t) 1−θ ≡ = .10) (9. and sr = 0 if θ = 1.451: Introduction to Economic Growth You can recognize this expression as the discrete-time counterpart of the Euler equation from the Ramsey model. the ﬁrst-order condition can be written as s (t)−θ βR (t + 1)1−θ = (w (t) − s (t))−θ .11) . The relationship between the rate of return on savings and the level of savings reﬂects the counteracting inﬂuences of income and substitution eﬀects you are familiar with from basic micro. ψ (t + 1) (9.9) implies that the saving rate can be written as s (t) = where ψ (t + 1) ≡ [1 + β −1/θ R (t + 1)−(1−θ)/θ ] > 1. Moreover.6) implies k (t + 1) = w(t) s(t) = . Alternatively. i. With log preferences. (1 + n) (1 + n)ψ (t + 1) 211 (9. log preferences. and thus changes in the interest rate (and therefore changes in the capital-labor ratio of the economy) have no eﬀect on the savings rate.9) Note that 0 < sw < 1. income and substitution eﬀects exactly cancel each other. sr > 0 if θ < 1.. (βR (t + 1))−1/θ ∂R (t + 1) θ ψ (t + 1) w (t) . Equation (9. is of special importance and is often used in many applied models.e. but sr < 0 if θ > 1. ensuring that savings are always less than earnings. Now equation (9. The relationship between the savings and factor prices is given by sw ≡ sr ∂s (t) 1 = .

k (t + 1) = (1 + n) [1 + β −1/θ f 0 (k(t + 1))−(1−θ)/θ ] f (k∗ ) − k∗ f 0 (k∗ ) .14.12) The steady state then involves a solution to the following implicit equation: k∗ = Now using the Cobb-Douglas formula. which always has a unique solution. in which case.12): k (t + 1) = (1 + n) [1 + β −1/θ (αk(t + 1)α−1 )−(1−θ)/θ ] (1 − α) k (t)α . α (9. (1 + n) [1 + β −1/θ f 0 k∗ )−(1−θ)/θ ] f (k (t)) − k (t) f 0 (k (t)) (9. substitute for the Cobb-Douglas production function in (9.13) can be rewritten as h i 1−α (1 + n) 1 + β −1/θ (R∗ )(θ−1)/θ = R∗ . this steady-state equilibrium is globally stable for all k (0) > 0.451: Introduction to Economic Growth or more explicitly. we have that the steady state is the solution to the equation h ¡ ¢(θ−1)/θ i = (1 − α)(k∗ )α−1 . (1 + n) 1 + β −1/θ · α(k ∗ )α−1 (9. To do this. The next ﬁgure shows the dynamics diagrammatically in this particular (well-behaved) case. deﬁne R∗ ≡ α(k∗ )α−1 as the marginal product of capital in steady-state.14) The steady-state value of R∗ and thus k ∗ can now be determined from equation (9.14). (9. the following proposition can be proved (proof left for Problem Set 5): Proposition 25 In the overlapping-generations model with two-period lived agents CobbDouglas technology and CRRA preferences. which look very similar to the dynamics of the basic Solow model: 212 .15). equation (9. there exists a unique steady-state equilibrium with the capital-labor ratio k∗ given by (9.13) For simplicity. We can next investigate the stability of this steady state.13) and as long as θ ≥ 1.15) Now using (9.

14.1: 213 .451: Introduction to Economic Growth Figure 9.

N (t)) = K (t + 1) + N (t) c1 (t) + N (t − 1) c2 (t) .2.4) noting that R (t + 1) = f 0 (k (t + 1)). In particular. However. 1+n . which is identical to (9. Substituting from (9. This is not surprising. the social planner’s and the competitive economy’s allocations across individuals will diﬀer. since the social planner would allocate consumption of a given individual in exactly the same way as the individual himself would do.14. and compare the overlapping-generations equilibrium to the choice of a social planner wishing to maximize a weighted average of all generations’ utilities. the social planner in question maximizes ∞ X t=0 βt S U (t) where β S is the discount factor of the social planner.451: Introduction to Economic Growth 9. which can alternatively be divided by N (t) and written in per capita terms as f (k (t)) = (1 + n) k (t + 1) + c1 (t) + This maximization problem immediately implies u0 (c1 (t)) = βf 0 (k (t + 1)) u0 (c2 (t + 1)) .1). this implies: ∞ X t=0 βt S (u (c1 (t)) + βu (c2 (t + 1))) subject to the resource constraint F (K (t) . since the social planner is giving diﬀerent weights to diﬀerent generations 214 c2 (t) .5 Pareto Optimality Let us now return to the general problem.

7).451: Introduction to Economic Growth as captured by the parameter β S . note that in steady state we have −1 ∗ f (k∗ ) − (1 + n)k∗ = c∗ c2 1 + (1 + n) ≡ c∗ . it can be shown that the socially planned economy will converge to a steady state with capital-labor ratio k S such that ¡ ¢ β S f 0 kS = 1 + n. More interesting is the question of whether the competitive equilibrium is Pareto optimal. k S is typically diﬀerent from the steady-state value of the competitive economy. The example from Shell in the previous section suggests that it may not be. that is. Clearly. we cannot use the First Welfare Theorem (Theorem 10) because of the inﬁnite number of commodities. In particular. β . thus by reducing savings. exactly as in the Shell’s example. consumption can increase for every generation. k∗ . In fact. which is not surprising given the diﬀerent preferences that are being maximized. which is similar to the modiﬁed golden rule we saw in the context of the Ramsey growth model. the economy is to the right of the golden rule. the competitive equilibrium is not in general Pareto optimal.7). In particular. k∗ . can be so high that it is in fact greater than kgold . More speciﬁcally. where the ﬁrst line follows by the accounting identity. The simplest way of seeing this is that the steady state level of capital stock.14. given by (9. it does not depend on preferences (the utility function u (·)) and does not even depend on the individual rate of time preference. In particular. Therefore ∂c∗ = f 0 (k∗ ) − (1 + n) ∗ ∂k 215 . given by (9. and the second deﬁnes c∗ as the total steady-state consumption.

Now if k∗ > kgold . the transversality condition (which follows from individual optimization) required that r > g + n. Recall that in the inﬁnite-horizon Ramsey economy. Then we have ∆cT = (1 + n) ∆k > 0 ∆ct = − (f 0 (k ∗ − ∆k) − (1 + n)) ∆k for all t > T Since k∗ > kgold . Another way of expressing dynamic ineﬃciency is that r∗ < n. i. therefore. Consider the following variation where the capital stock for next period is reduced by a small amount.451: Introduction to Economic Growth and kgold is deﬁned as f 0 (kgold ) = 1 + n. the steady-state interest rate r∗ = R∗ − 1 is less than the rate of population growth. so reducing savings can increase (total) consumption for everybody. The increase in consumption for each generation can be allocated equally during the two periods of their lives. changed by −∆k.14. and we move immediately to a new steady state. dynamic ineﬃciency could never arise in this Ramsey economy. suppose we start from steady state at time T with k∗ > kgold . In particular. This variation clearly creates a Pareto improvement in which all generations are better oﬀ. the economy is referred to as dynamically ineﬃcient. If this is the case. that is. thus ∆ct > 0 for all t ≥ T . where ∆k > 0. Dynamic ineﬃciency arises because of the heterogeneity inherent in the overlapping generations model which removes the transversality condition. This establishes: 216 . f 0 (k∗ − ∆k) − (1 + n) < 0. then ∂c∗ /∂k∗ < 0.e. thus necessarily increasing their utility (by the assumption that u (·) is strictly increasing from Assumption 1). for small enough ∆k.

the capital stock. that is the rate of interest less than the rate of population growth. in which the young make contributions the Social Security and their contributions are paid back to them in their old age. Dynamic ineﬃciency. As the above derivation makes it clear. it is possible to reduce the capital stock starting from the competitive steady state and increase the consumption level of all generations. possible lack of Pareto eﬃciency in the competitive equilibrium is intimately linked with dynamic ineﬃciency. where transfers from the young directly go to the current old. 9.3 Role of Social Security in Capital Accumulation We now brieﬂy discuss how Social Security can be introduced as a way of dealing with overaccumulation in the overlapping-generations model. The alternative is an unfunded system or a pay-as-you-go Social Security system. and this is invested in the only productive asset of the economy. In Problem Set 5. the government at date t raises some amount d (t) from the young (by compulsory contributions to their Social Security accounts etc. Very brieﬂy. we will consider a fully-funded system. 9.451: Introduction to Economic Growth Theorem 39 In the overlapping-generations economy the competitive equilibrium is not necessarily Pareto optimal. and pays the workers when they 217 . you will be working through a numerical example that will illustrate under what conditions dynamic ineﬃciency can happen.3.14.1 Fully Funded Social Security In a fully funded system. More speciﬁcally. whenever r∗ < n and the economy is dynamically ineﬃcient.). is not a theoretical curiosum.

for a given choice of d (t) by the government. Suppose that s (t) ≥ 0 for all t. the privately-optimal saving sequence {s (t)}t=0 is such that s (t) > 0 for all t. then the set of competitive equilibria without Social Security are the set of competitive equilibria with Social Security. with the constraint that s (t) ≥ 0 and without. More speciﬁcally. This discussion immediately establishes: Proposition 26 Consider a fully funded social security system in the above-described environment whereby the government collects d (t) from young individuals at date t. 218 . It is clear that as long as s (t) is free.s(t) max u (c1 (t)) + βu (c2 (t + 1)) subject to c1 (t) + s (t) + d (t) ≤ w (t) and c2 (t + 1) ≤ R (t + 1) (s (t) + d (t)) . When s (t) ≥ 0 is imposed as a constraint. Therefore this economy can be analyzed under two alternative assumptions. since they have the income from Social Security. the competitive equilibrium applies. If given the feasible sequence {d (t)}∞ t=0 of Social Security payments. then the competitive equilibrium applies if given the sequence ∞ {d (t)}∞ t=0 . whatever the sequence of Social Security payments {d (t)}∞ t=0 (as long as it is feasible).c2 (t+1).14. we now have the individual maximization problem as c1 (t). 1. Notice that now the total amount invested in capital accumulation is s (t) + d (t) = (1 + n) k (t + 1).451: Introduction to Economic Growth are old an amount R (t + 1) d (t). It is also no longer the case that individuals will always choose s (t) > 0. the utility-maximizing sequence of savings {s (t)}∞ t=0 is such that s (t) > 0 for all t.

Only s (t) goes into capital accumulation. the set of competitive equilibria without Social Security are the set of competitive equilibria with Social Security. Without the constraint s (t) ≥ 0. Now we have that the government collects d (t) from the young at time t and distributes this to the current old with per capita transfer b (t) = (1 + n) d (t) (which takes into account that there are more young than old because of population growth).3.451: Introduction to Economic Growth 2. 9. What this implies is that the rate of return on Social Security payments is 1 + n rather than R (t + 1).2 Unfunded Social Security The situation is very diﬀerent with unfunded Social Security. Therefore.14. the individual maximization problem becomes c1 (t).c2 (t+1). intuitively we expect unfunded Social Security to 219 . Therefore. for a given feasible sequence of Social Security payment levels {d (t)}∞ t=0 . given any feasible sequence {d (t)}∞ t=0 of Social Security payments. This is very intuitive: the d (t) taken out by the government is fully oﬀset by a decrease in s (t) as long as individuals were performing enough savings (or always when there are no constraints to force positive savings privately).s(t) max u (c1 (t)) + βu (c2 (t + 1)) subject to c1 (t) + s (t) + d (t) ≤ w (t) and c2 (t + 1) ≤ R (t + 1) (s (t)) + (1 + n) d (t + 1) . because unfunded Social Security is a pure transfer system.

This leads to the following proposition (proof left for Problem Set 5). 220 . Proposition 27 Consider the above-described overlapping generations economy and suppose that there is dynamic ineﬃciency into decentralized competitive equilibrium. In many ways. Then there exists a feasible sequence of unfunded Social Security payments {d (t)}∞ t=0 which will constitute a competitive equilibrium starting from any date t.451: Introduction to Economic Growth reduce capital accumulation. unfunded Social Security reduces the overaccumulation and improves the allocation of resources. this may be a good thing. Intuitively.14. this is equivalent to commodities being transferred from high indexed agents to low indexed agents in the Shell example above. and in economies with dynamic ineﬃciencies.

For now. for all practical purposes. so equilibrium growth is the same as optimal growth. In fact. However. except that we have to think of expectations.Chapter 10 Recitation Material: Stochastic Growth 10. it turns out that despite the stochastic shocks.k(t)}∞ t=0 max E0 β t u (c (t)) (10. the Brock-Mirman model is the starting point of the Real Business Cycle models you will study later. This was done in the context of optimal growth. if the economy admits a representative household. it is a solution to the following program: ∞ X t=0 {c(t).1 The Brock-Mirman Model The classic analysis of economic growth with stochastic shocks was undertaken by Brock and Mirman in their 1972 paper. is identical to the non-stochastic model. the First and Second Welfare Theorems still hold. it suﬃces to note that this model. In particular.1) 221 .

14.451: Introduction to Economic Growth subject to k (t + 1) = A (t) f (k (t)) + (1 − δ ) k (t) − c (t) and k (t) ≥ 0, (10.2)

with given k (0). Here E0 is the expectations operator conditional on information available at time t = 0. The budget constraint with the production function substituted in, equation (10.2), requires some care in interpreting. A (t) is now introduced as a stochastic productivity term. The expectations are taken because the time path of the sequence {A (t)}∞ t=0 is not known in advance. This implies that strategies have to have the proper measurability conditions. In particular, in general we can do this by assuming that information at time t is represented by a partition Ft , so that E0 [x] = E [x | Ft ], and variables chosen at time t have to be measurable with respect to Ft . This simply means that they can not be conditioned on realizations of future-dated stochastic variables. The above model can be enriched by assuming that there are stochastic preference shocks, for example by augmenting the utility function u (c (t)) by a shock b (t), so that u (c (t) | b (t)) is also a random function dependent on the realization of b (t). In addition, analysis of growth under uncertainty makes the standard assumptions on the production function as in Assumptions 1 and 2 above, and the standard assumption on preferences as in Assumption 8. Given this setup, the problem can again be written as a dynamic programming problem, but now it is a stochastic dynamic programming problem, in particular, it takes the form V (k) = max {u (c) + β EV [Af (k) + (1 − δ ) k − c]}

c∈Γ(k)

(10.3)

where the expectation is included because there is uncertainty about future values of the stochastic variable A. The rest of the analysis is very similar to the non-stochastic case, except that the Euler equations also include expectations. For example, assuming that A (t) 222

14.451: Introduction to Economic Growth is known at time t, the key Euler equation becomes: u0 (c (t)) = β Et [(A (t + 1) f 0 (k (t + 1)) + (1 − δ)) u0 (c (t + 1))] .

10.2

Application: Risk, Diversiﬁcation and Growth

I now present the model from Acemoglu-Zilibotti (JPE 1997) aimed at capturing the interaction between diversiﬁcation of risks and capital accumulation, and emphasizing the endogenous generation of risks in the growth process. This model will give an example of stochastic growth and also illustrate how the productivity of capital can change endogenously over the development process, and diﬀer across countries. Finally, this model will also introduce some tools that are useful for analyses of dynamic stochastic economies.

10.2.1

The Environment

Consider the following model. There is a continuum of equally likely states represented by the unit interval. Agents have to invest their savings in intermediate sectors, which will than payoﬀ in the form of capital in the next period. Intermediate sector j ∈ [0, 1] pays a positive return only in state j and nothing in any other state. This formulation implies that investing in a sector is equivalent to buying a Basic Arrow Security that only pays in one state of nature. More formally, an investment of F j in sector j generates capital of the amount RF j if state j occurs and F j ≥ Mj , and nothing otherwise. There is also is a safe project, which transforms one unit of savings into r < R of capital. The requirement F j ≥ Mj implies that all intermediate sectors have linear technologies but some require a certain minimum size, Mj , before being productive. The distribution of 223

14.451: Introduction to Economic Growth minimum size requirements is given by:

½ Mj = max 0,

Sectors j ≤ γ have no minimum size requirement and for the rest of the sectors, the minimum size requirement increases linearly). The next ﬁgure shows the minimum size requirements diagrammatically, and will be used for determining the equilibrium as well once demand for assets is introduced:

¾ D (j − γ ) . (1 − γ )

There are two important features here 1. risky investments have a higher expected return than the safe asset (i.e. R > r); 2. diﬀerent projects are imperfectly correlated so that there is safety in variety.

_

A

convenient implication of this formulation is that if a portfolio consists of an equiproportional investment F in all projects j ∈ J ⊆ [0, 1], and the measure of the set

_

J is p, then the portfolio pays the return RF with probability p, and nothing with probability 1 − p. 224

14.451: Introduction to Economic Growth These features imply that if the aggregate production set were convex (i.e. D = 0), all agents would invest an equal amount in all intermediate goods sectors and diversify all risks. However, in the presence of nonconvexities, as captured by the minimum size requirements, there is a trade-oﬀ between insurance and high productivity. The preferences of consumers over ﬁnal goods is deﬁned as: Et U (ct , ct+1 ) = log(ct ) + β Z

1

log(cj t+1 )dj,

(10.4)

0

which again ensure a constant savings rate. Note that integration over [0,1] is over the states of nature. The individual life cycle and decisions are summarized in the next ﬁgure:

**Output of the ﬁnal good sector is given by:
**

−α . Yt = AKtα L1 t

(10.5)

and let us normalize labor to 1. 225

14.451: Introduction to Economic Growth The aggregate capital stock depends on the realization of the state of nature. If the R j j )dh where Fh,t is the amount of savings state of nature is j , then Ktj+1 = Ωt (rφh,t + RFh,t

invested by agent h ∈ Ωt in sector j , φh,t is the amount invested in the safe asset, and Ωt is the set of young agents at time t. Since both labor and capital trade in competitive markets, equilibrium factor prices in state j are given as: Wtj+1 ¡ ¢α = (1 − α)A Ktj+1 ≡ (1 − α)A ¡ ¢α−1 = αA Ktj+1 ≡ αA µZ

Ωt

µZ

(rφh,t +

j RFh,t )dh

Ωt

¶α

,

(10.6)

ρj t+1

(rφh,t +

j RFh,t )dh

¶α−1

.

(10.7)

10.2.2

Equilibrium

Now consider the portfolio decisions of households. Each household takes the set of traded securities as given, and maximizes its utility by allocating its savings across diﬀerent assets. Securities are labeled by the indices of the project to which they are attached. Therefore, one unit of security j entitles its holder to R units of t + 1 capital in state of nature j . Denote the unit price of security j (in terms of savings of time t) by Pj,t . Assume that the intermediates are supplied by ﬁnancial intermediaries. Since 1 unit of savings invested in a project that’s open yields one unit of capital, competition among ﬁnancial intermediaries ensures that in equilibrium Pj,t = 1–that is, all projects will be oﬀered to households at marginal cost. Therefore, denoting the set of open projects at time t by Jt , optimal consumption, savings and portfolio decisions can be characterized by: log(ct ) + β max j st ,φt ,{Ft }0≤j≤1 226 Z

1

log(cj t+1 )dj,

(10.8)

0

14.451: Introduction to Economic Growth subject to: φt + Z

1

Ftj dj = st ,

(10.9) (10.10) (10.11) (10.12)

¡ j j¢ = ρ + RF rφ cj t , t t+1 t+1 / Jt , Ftj = 0, ∀j ∈ ct + st ≤ wt ,

0

It is important that these agents not only take wt , ρj t+1 , but also the set of risky assets Jt as given. A static equilibrium given wage earnings of young agents, Wt , (or given Kt ) is a solution to the maximization problem (10.8) subject to (10.9)-(10.12), such that Ftj ≥ Mj for all open sectors. A dynamic equilibrium is a sequence of static equilibria linked to each other through (10.6) Because preferences are logarithmic, the following saving rule is obtained irrespective of the risk-return trade-oﬀ:

∗ s∗ t ≡ s (wt ) =

β wt . 1+β

(10.13)

Given this result, a household’s optimization problem can be broken into two parts: ﬁrst, the amount of savings is determined, and then an optimal portfolio is chosen. Next observe that in equilibrium we will have 1. Ftj ∗ = Ftj ∗ ∀j, j 0 that are open (i.e. ∀j, j 0 ∈ Jt ). Since each individual is facing the same price for all of the traded symmetric Arrow securities, he would want to purchase an equal amount of each–i.e., a balanced portfolio. 2. The set of open projects will be Jt = [0, nt ] for some nt ∈ [0, 1]. This states that when only a subset of projects can be opened in equilibrium, “small projects” are opened 227

0

14.451: Introduction to Economic Growth before “large projects”. As a result, if a sector j ∗ is open, all sectors j ≤ j ∗ must also be open. Given this result, the maximization problem simpliﬁes to: h i h i (qG ) (qB ) max nt log ρt+1 (RFt + rφt ) + (1 − nt ) log ρt+1 (rφt ) ,

φt,Ft

(10.14)

subject to: φt + nt Ft = s∗ t, (10.15)

(q )

B ∗ where nt and ρj t+1 s are taken as parametric by the agent, and st is given by (10.13). ρt+1 =

**α (rφt )α−1 is the marginal product of capital in the “bad” state, when the realized state
**

G = α (RFt + rφt )α−1 applies in the “good is j > nt and no risky investment pays oﬀ. ρt+1

(q )

state”, i.e. when the realized state is j ≤ nt . Maximization of (10.14) gives: φ∗ t = (1 − nt )R ∗ s, R − rnt t (10.16)

Ftj,∗ Then let: n∗ t (Kt ) = ⎧ ⎨ ⎩

⎧ ⎨ F ∗ ≡ R−r s∗ , ∀j ≤ n t t R−rnt t = . ⎩ 0 ∀j > nt

1/2

(10.17)

Γ α +γR (R+rγ )−{(R+rγ )2 −4r[(R−r)(1−γ ) D Kt ]} 2r

if Kt ≤ if Kt >

1

β . Then there exists a unique equilibrium such that s∗ where Γ ≡ A(1 − α) 1+ t = β j∗ ∗ α)AKt , and φ∗ t , Ft are given by (10.16) and (10.17) with nt = nt (Kt ).

¡ D ¢1/α

Γ

¡ D ¢1/α

Γ

(10.18)

β (1 1+β

−

This equilibrium can be expressed as the intersection of the aggregate demand of each risky asset, F ∗ (nt ), with the thick curve that traces minimum size requirements in the ﬁgure. 228

14.451: Introduction to Economic Growth

10.2.3

Dynamics

**Next, it is straightforward to characterize the full stochastic equilibrium process, the equilibrium law of motion of Kt is: Kt+1 = ⎧ ⎨ ⎩
**

r(1−n∗ t) RΓKtα R−rn∗ t

RΓKtα

prob. 1 − n∗ t prob. n∗ t

(10.19)

∗ where n∗ t = n (Kt ) is given by equation (10.18).

The capital stock follows a Markov process in which the level of capital next period depends on whether the economy is lucky in the current period (which happens when the risky investments pay-oﬀ, probability n∗ t ). Moreover, the probability of this event changes over time. As the economy develops, it can aﬀord to open more sectors, and the probability of transferring a large capital stock to the next period, n∗ t , increases. Also from (10.19), the expected productivity of an economy depends on its level of development and diversiﬁcation. To see this, deﬁne expected “total factor productivity” (conditional on the proportion of sector open) by r(1 − n∗ ) R + n∗ R ∗ R − rn

σ e (n∗ (Kt )) = (1 − n∗ )

(10.20)

Simple diﬀerentiation establishes that as n∗ t increases, this measure also increases. To formalize the dynamics of development, deﬁne the following concepts; (i) QSSB : “quasi steady state” of an economy which always has unlucky draws. An economy would converge to this quasi steady state if it follows the optimal investments characterized above but the sectors invested never pay-oﬀ due to bad luck . 229

14.451: Introduction to Economic Growth (ii) QSSG : “quasi steady state” of an economy which always receives good news.

The capital stocks of these two quasi steady states are:

K QSSB

1 " ¡ ¢ # 1− α r 1 − n∗ (K QSSB ) R Γ = R − rn∗ (K QSSB )

and

K QSSG = (RΓ) 1−α .

1

(10.21)

If uncertainty could be completely removed, that is n(K QSSG ) = 1, then there would never be bad news, and the good quasi steady state would be a real steady state; a point, if reached, from which the economy would never depart. From equations (10.18) and (10.21), the condition for this steady state to exist is that the saving level corresponding to K QSSG be suﬃcient to ensure a balanced portfolio of investments, of at least D, in all the intermediate sectors. Thus, if:

D < Γ 1−α R 1−α ,

1

α

(10.22)

a steady state will exist and we denote it by K SS . The following ﬁgure is useful in understanding the dynamics.

230

their output will fall in case they receive bad shocks. Regions I and II are separated by K QSSB . and the probability of bad news is very high when the economy has a level of capital stock just above K QSSB . When they are above this level. Finally. the Inada conditions of the production function guarantee positive growth even conditional on bad news (both curves lie above the 45◦ line). Note that even when it grows. all economies will grow towards it.451: Introduction to Economic Growth At very low levels of capital. the economy will eventually enter region III where 231 . and will typically experience some set-backs. Then.14. the economy is still exposed to large undiversiﬁed risks. As good news is received. provided (10. the capital stock will grow and the probability of a further lucky draw will increase. there is a range (region II ) in which growth only occurs conditional on good draws (the bad draws curve is below the 45◦ line).22) is satisﬁed. When they are below this level.

4 Eﬃciency Since all agents are price takers.18) and St∗ = s∗ t denote total savings. 232 . φF B (Kt ) < φ∗ (Kt ) and each agent receives the following portfolio of assets: ⎧ ⎪ ⎪ F j. nF B (Kt ) = n∗ (Kt ) and Ftj. Then. Jt . the social planner explicitly chooses Ftj and the number of open sectors.14. The diﬀerence between the social planner’s allocation and a decentralized equilibrium is © ª that. This turns out not to be the case. It is straightforward to see that the subset of projects in which the planner will invest will be of the form J F B = [0.F B = St∗ . nF B ]–similar to the decentralized equilibrium.2. nF B (Kt ) > n∗ (Kt ).F B = Mj ⎪ ⎪ ⎪ ⎩ F j. if nF B (Kt ) ≥ jt ≥ jt (10. subject to feasibility.F B = Mj ∗ > Mj if ⎪ ⎨ t Ftj.{Ft }0≤j≤nt Z nt 0 log(RFtj + rφt )dj + (1 − nt ) log(rφt ). ∀j . consider the portfolio allocation that a social planner maximizing the welfare of the current generation of savers would choose taking the amount of savings as given.23) This maximization problem leads to the following result: ∗ Let n∗ (Kt ) be given by (10.φt . ∃jt ∗ .F B = 0 t ∗ jt < jt ∗ ∗ < nF B (St ) s.t.24) if jt > nF B (Kt ) ∗ And ∀St ≥ D. it may be conjectured that the decentralized equilibrium here is eﬃcient. and there will be deterministic convergence to K SS . (10.451: Introduction to Economic Growth all diversiﬁable risks will be removed (since all sectors are open and an equal amount is invested in all sectors).∀St < D. the planner will solve max j nt . To illustrate this. 10. Therefore.

all commodities. all existing projects become more attractive relative to the safe asset because the amount of undiversiﬁed risks they carry are reduced. the externality is not internalized.14. and as a result. risk-averse agents are more willing to buy the existing securities. This is shown in the next picture: Why is the decentralized equilibrium ineﬃcient? The answer is a pecuniary externality due to missing markets. whereas such a price schedule does not exist in this economy because of the non233 . even those that are not traded in equilibrium are priced. and will ﬁnance this by investing less in the sectors without the minimum size requirement. and this gives the technical intuition for the ineﬃciency. Since each agent ignores his impact on others’ diversiﬁcation opportunities. In an Arrow-Debreu equilibrium. the social planner will always open more sectors/projects than the decentralized equilibrium. As an additional sector opens.451: Introduction to Economic Growth In other words. It is important to also note that the decentralized equilibrium did not correspond to an Arrow-Debreu equilibrium.

will have fewer sectors open. This intermediary can collect all the savings and oﬀer to each saver a complex security (as diﬀerent from a Basic Arrow Security) that pays B B RFtj.2. As a result. Instead. Although from this discussion it may appear that the ineﬃciency we identiﬁed may 234 . The analysis also implies a systematic relationship between the variability of the current performance and the level of development (the level of the capital stock).6 Ineﬃciency with Alternative Market Structures Would the market failure in portfolio choices be overcome if some ﬁnancial institution could coordinate households’ investment decisions? Imagine that rather than all agents acting in isolation and ignoring their impact on each others’ decisions.14. they will “correctly” fear undiversiﬁed risks. these economies will invest in the low productivity safe assets. Holding t t this security would make each consumer better oﬀ compared to the equilibrium. funds are intermediated through a ﬁnancial coalition-intermediary. 10. the analysis here uses a more natural competitive equilibrium notion (common in general equilibrium analyses of monopolistic competition) where only commodities traded in equilibrium are priced. and achieve low productivity. This is a pattern we see in the data.451: Introduction to Economic Growth convexities of the production set. where Ftj. Richer countries will have less variable growth rates. and therefore. 10. that is those that had received bad shocks in the past.F B and φF are as in the optimal portfolio.2.F B + rφF in each state j .5 Implications An important implication of this analysis is that there will be systematic diﬀerences in productivity across countries depending on the realization of past shocks: economies with low capital.

Put diﬀerently. a coalition cannot commit to a path of action that will be against the interests of its members in the continuation game. Coalitions at all points maximize a weighted utility of their members. An agent cannot be part of two coalitions at the same time. and its importance will be discussed further below. the unique equilibrium allocation with ‘unfettered’ competition among intermediaries will be identical to the one we characterized as the equilibrium above. The second one is the most important assumption. some agents initiate the formation of a coalition of households which buys securities on behalf of its members. 2. In return. In particular. we will show that this is not the case. In order to model the endogenous formation of coalitions. ν ¯. let us now assume that savings can be intermediated by some households who decide to act as middlemen and run an investment fund . this assumption makes it easier for an eﬃcient allocation to be sustained as an equilibrium. Let me now introduce the following three assumptions for the coalition-formation game: 1. We view this as a very natural assumption along the lines of subgame perfection. Unless some rather strong assumptions are made about the set of contracts that a ﬁnancial intermediary can oﬀer.451: Introduction to Economic Growth not be robust to the formation of more complex ﬁnancial institutions. following Townsend (1983). Coalitions cannot exclude other agents (or coalitions) from investing in a particular project.14. participants to the ﬁnancial coalition can be charged an intermediation fee. Projects are still run by individual households. 235 . The ﬁrst assumption is introduced to simplify the objective function of coalitions. In fact. 3.

then Zh = ∅. Zh = (0. we denote the set of ﬁrst-stage announcements of all agents by Z (1) : Ω → R+ × Υ.e. 1] × R+ is the set of all second-period announcements. In the second stage.h ). the set of all subsets of Ω. 1] × R+ . each household h ∈ Ω can announce that he is willing to act as an intermediary for a speciﬁed set of households Ξh (where Ξh ∈ Υ. now. If agent h announces that he household h is an announcement Zh = (¯ will not act as an intermediary. the second-stage announcement for agent h is Zh = (j. which means that h will only intermediate (at most) his own savings. as in the game discussed in Section 3. In general.451: Introduction to Economic Growth Assumption three is also mainly expositional. In the third stage.14. Let Ξa h ⊆ Ξh denote this subset of households. and we deﬁne µ(). i. Formally. Pj. coalitions would never want to exclude others. Ξh ) ∈ R+ × Υ. Pj.h ) ∈ [0. We will also denote the set of minimum security prices announced in the second stage of the game by P = {P j }j ∈J .e. securities are sold to ﬁnancial intermediaries rather than directly to households. In the ﬁrst stage. Formally. in equilibrium Υ will be partitioned into disjoint coalitions. each agent h ∈ Ω can announce his plan to run at most one project and sell the corresponding Basic Arrow Security. the game has now three stages. each household takes the set of prior announcements. only a subset of agents belonging to Ξh will accept the oﬀer of the intermediary. and thus this assumption is only imposed to simplify the exposition. But. The intermediary h will invest the savings he collects (net of his commission ν ¯h ) in shares of both risky and safe projects so as to maximize the total utility of the agents belonging to Ξa h . {h}). 236 (2) (1) (1) (1) . as the Lebesque measure over Υ). and Z (2) : Ω → [0. A ﬁrst-stage strategy for ν h . Note that because of assumption 1. Z (1) and Z (2) . We will see below that as long as Assumption number two holds. h announces a pair (j. i. there is autarky. Among the possible non-null announcements. Finally.

Note that although the set Mh Z (1) could be empty. if a coalition Ξ invests FΞ j . a price function P ∗ (Z ∗ ) for all Basic Arrow Securities. and factor payments W ∗ and ρ∗ such that given the announcements of the previous stage(s) and the announcements of all other agents in the current stage.6) and (10. (i) P j.14. Z (3)∗ ) at each stage of the game. each intermediary makes the optimal investment decision. Here we emphasize perfection in order to reiterate the importance of Assumption 3 in our analysis. Lemma 3 In equilibrium. since any agent can costlessly make the autarky announcement in the ﬁrst stage. this will never be the case in equilibrium.451: Introduction to Economic Growth as given. h ∈ Ξi . after all agents announce which coalition-intermediary they will belong to. and induced holdings of the safe asset φh (Z ) and securities Fh (Z ) for all agents.7).∗ (Z ∗ ) = 1. Ξi ). Or equivalently. and chooses which coalition to join. it is now possible to establish the following proposition: 237 (i) . ∀j . With this remark. then Fh will be the share of agent h More precisely. ∀h ∈ Ω. (ii) ν ¯h = 0. Note that the deﬁnition of equilibrium used so far was also subgame perfect. in this coalition times FΞ (3) Deﬁnition 9 A (perfect) equilibrium is a set of announcements Z ∗ = (Z (1)∗ . every household chooses Zh that maximizes its utility as given by (5) and factor returns are determined by (10. a saving j∗ ∗ ∗ ∗ ∗ decision s∗ h (Z ). Fh j j in project j . We still use the notation φh . Zh is h’s choice of an o ¡ (1) ¢ n (1) intermediary from Mh Z ν i . The ﬁrst observation is that free entry will drive proﬁts (commissions) to zero in both the ﬁrst and second stages. Finally. This is established by the following lemma (proof omitted). j to denote the investment of an agent (through a coalition) in the safe and risky assets. the set of coalitions which ≡ i ∈ Ω | Zi = (¯ ¡ ¢ announced his name. Z (2)∗ .

the ineﬃciency cannot be prevented. n∗ t ] . ∀h ∈ Ω either Zh ∗ j ∈ [0.451: Introduction to Economic Growth Proposition 28 The set of (perfect) equilibria is non-empty and all allocations in this set have the following characteristics: 1. j∗ ∗ ∗ 3. nt ] ∃h ∈ Ω such that Zh (2)∗ (2)∗ = ∅ or Zh (2)∗ = (j. ∀j ∈ [0.18). In the third stage ∀Ξa h 6= ∅ will choose a portfolio which induces φt and Ft = Ft as given by equations (10.16) and (10. 2. 1). And. Let n∗ t be deﬁned in (10. Mh 6= ∅ (all agents are included in some coalition).14. 238 .17). The key feature is that each agent would be creating a positive externality by holding a non-balanced portfolio like the one necessary for eﬃciency. undermining eﬀorts to sustain the eﬃcient allocation. and they will typically ﬁnd a way of moving towards a balanced portfolio. This result implies that even with unrestricted coalitions. 1) where = (j. Then. ∀h ∈ Ω.

Part III Endogenous Growth 239 .

.

14. or growth came exogenously from the unmodeled process of labor-augmenting technological progress. either the economy settled into a steady state without any economic growth.451: Introduction to Economic Growth Until now. we investigated economic growth models without “growth” in the sense that. The rest of the lectures will look into issues of endogenous growth and how the process of development. whereby a society makes the transition from being a less-developed economy to a more developed one. takes place endogenously. 241 .

451: Introduction to Economic Growth 242 .14.

We will also see that in fact what matters is that the accumulation technology is linear. The ﬁrst one basically keeps the essence of the neoclassical approach.Chapter 11 First-Generation Models of Endogenous Growth The ﬁrst-generation models of endogenous growth made a big advance relative to the neoclassical growth model in generating sustained growth. But 243 . This is the so-called AK model. with competitive markets and no externalities. 11. not necessarily the production technology.1 AK Model Revisited Let us start with the simplest neoclassical model of sustained growth. Two approaches are noteworthy here. where the production technology is linear in capital. which we already encountered in the context of the Solow growth model. The second makes the ﬁrst attempt at endogenizing technology by introducing externalities and knowledge spillovers (ﬂows) across ﬁrms.

11.2) ¸ c1−θ − 1 exp (− (ρ − n) t) dt. More speciﬁcally.1) 0 where a is assets per person.14. We may as well assume these preferences from the beginning. We again impose the no-Ponzi game constraint: ½ ∙ Z t ¸¾ [r(s) − n] ds ≥0 lim a(t) exp − t→∞ 0 (11. we are forced to use preferences that are asymptotically consistent with balanced growth. Preferences and Technology Since there will be growth and we are. at least at ﬁrst.4) (11. I have now suppressed time dependence to simplify notation. thus choose the standard CRRA preferences of the canonical model.3) The Euler equation for the representative household is the same as before and gives: c ˙ 1 = (r − ρ) c θ and the transversality condition is: ½ ∙ Z t ¸¾ [r(s) − n] ds = 0. r is the interest rate.451: Introduction to Economic Growth for now it makes sense to start with the simpler case of the AK economy. interested in balanced growth.1.1 Demographics.5) 244 . (11. 1−θ ∙ ∞ (11. lim a(t) exp − t→∞ 0 (11. w is the wage rate. let us assume that the economy admits an inﬁnitely-lived representative household with utility given by Z U= subject to the constraint a ˙ = (r − n)a + w − c. and n is the growth rate of population.

we have that per capita output is given by y = f (k ) = Ak. w. as is obvious from equation (11. (11. More important is the fact that the Inada conditions embedded in Assumption 2 are no longer satisﬁed. in which case there will be labor income coming from human capital. lim f 0 (k ) = A > 0. k→∞ This feature is essential for sustained growth.. R = r + δ . in this model we can think of k as a combination of physical and human capital. the wage rate. Alternatively. 245 . The conditions for proﬁt-maximization are very similar to before. (11. More speciﬁcally. except that Assumptions 1 and 2 are not satisﬁed. we also have that the net rate of return on the savings is constant and equal to: r = A − δ. and there are no diminishing returns (i. is zero.14.6) has a number of notable diﬀerences from our standard production function satisfying Assumptions 1 and 2. First. This is a somewhat extreme result.6) with A > 0 being a constant. Since.e. In particular. Equation (11. and require that the marginal product of capital be equal to the rental price of capital.6). output is only a function of capital. the marginal product of capital is constant and equal to A.451: Introduction to Economic Growth The production sector is similar to before. and again it can be relaxed as we will see below. it is no longer the case that f 00 (·) < 0).7) Since the marginal product of labor is zero. which will be accumulating in the same way as physical capital (in particular linearly).

and substitute these into equations (11. and also we want to ensure positive growth.11) Since there is growth in this economy. k. To see this. and this is entirely independent of the level of capital stock per person. Therefore we impose: Assumption 12 A>ρ+δ > 1−θ (A − δ − ρ) + n + δ. c(t) = c(0) exp θ µ (11.451: Introduction to Economic Growth 11.4). we again use a = k.14. integrate equation (11. to obtain: ˙ = (A − δ − n)k − c k c ˙ 1 = (A − δ − ρ). (11.e.1.10) t→∞ The important result immediately follows from equation (11.2 Equilibrium To characterize the equilibrium.9).. that lifetime utility is bounded away from inﬁnity). we have to ensure that the transversality condition is satisﬁed (i. r = A − δ. θ 246 . This gives ¶ 1 (A − δ − ρ)t . This will also imply that there are no transitional dynamics in this model.2). c θ £ ¤ lim k(t)e−(A−δ−n)t = 0. and w = 0.9) (11. the economy will immediately start growing at a constant rate. (11. which is deﬁned in exactly the same way as in the basic neoclassical model. There is a constant rate of consumption growth (as long as A − δ − ρ > 0). Starting from any k (0). and (11.8) (11.9) starting from some initial level of consumption c(0) [still to be determined].5).

3 Transitional Dynamics We now more explicitly show there are no transitional dynamics. which yields ¶ µ 1 ˙ (A − δ − ρ)t .451: Introduction to Economic Growth The ﬁrst part of this condition ensures that there will be positive consumption growth. 11. not only the growth rate of consumption. k = (A − δ − n)k − c(0) exp θ (11. Therefore.14. then.9). This type of equation can be solved easily.8). and equal the growth rate of consumption given in equation (11. while the second part is the analogous condition to ρ + θg > g + n in the neoclassical growth model with technological progress. which was imposed to ensure bounded utility (and thus was used in proving that the transversality condition was satisﬁed). In particular recall that if z ˙ = az + g (t) . (11. but the growth rates of capital and output are also constant at all points in time.12) which is a ﬁrst-order.1.13) 247 . non-autonomous linear diﬀerential equation in k. that is. the solution is z (t) = z0 exp (at) + exp (at) Z t 0 exp (−as) g(s)ds for some constant z0 chosen to satisfy the boundary conditions. equation (11. To do this. let us substitute for c(t) from equation (11.12) solves for: © ª k(t) = κ exp((A − δ − n) t) + [(A − δ)(θ − 1)/θ + ρ/θ − n]−1 [c(0) exp ((1/θ) ((A − δ − ρ)t))] .11) into equation (11.

From (11. However. consider an increase in the rate of discount. it may look like capital is not growing at a constant rate. For example. Note that [(A − δ )(θ − 1)/θ + ρ/θ − n] > 0.13).451: Introduction to Economic Growth where κ is a constant to be determined. but it is also endogenous in the sense of being aﬀected by underlying parameters.15) (11.13) that: k(t) = [(A − δ)(θ − 1)/θ + ρ/θ − n]−1 [c(0) exp ((1/θ) ((A − δ − ρ)t))] = k (0) exp ((1/θ) ((A − δ − ρ)t)) . Notice also that Assumption 12 ensures that [(A − δ )(θ − 1)/θ + ρ/θ − n] > 0. It also pins down the initial level of consumption exactly as c (0) = [(A − δ )(θ − 1)/θ + ρ/θ − n] k (0) . (11. 248 .14) t→∞ It is also interesting to note that in this simple AK model. growth is not only “endogenous” in the sense of being sustained. (11. This equation naturally implies that capital and output grow at the same rate as consumption. since it is the sum of two components growing at diﬀerent rates. Therefore we have from (11. Thus the transversality condition can only be satisﬁed if κ = 0. so the second term in this expression converges to zero as t → ∞. this is where the transversality condition becomes useful.13) into the transversality condition.10). Let us substitute from (11. which yields lim [κ + [(A − δ )(θ − 1)/θ + ρ/θ − n]−1 c(0) exp (− [(A − δ )(θ − 1)/θ + ρ/θ − n] t)] = 0. where the second line immediately follows from the fact that the boundary condition has to hold for capital at t = 0.14. But the ﬁrst term is a constant.

16). capital and output all grow at the same rate g ∗ ≡ (A − δ − ρ)/θ starting from any initial positive capital stock per worker k (0). This can be proved either using First Welfare Theorem type reasoning. there is a representative household. the competitive equilibrium will be Pareto optimal. Finally. or by directly constructing the optimal allocation. clearly it will reduce the growth rate. changes in A and θ aﬀect the levels and growth rates of consumption. Summarizing. we can calculate the saving rate in this economy. the savings rate which was taken as exogenous in the basic Solow model is now a function of parameters. and more speciﬁcally of exactly the same parameters that determine the per capita growth rate.16) ˙ where naturally k/k = (A − δ − ρ)/θ. Consequently.14. Here. but had no eﬀect on the growth rate.6).1). and the savings rate is endogenously determined by (11. capital and output. In this economy there exists a unique equilibrium path in which consumption. The result is stated in the next proposition and left for you to prove): 249 . which was determined by the exogenous labor-augmenting rate of technological progress. Recall that in the Ramsey model. we have: Proposition 29 Consider the above-described AK economy. this aﬀected the level of income per capita. and there are no externalities. and the production technology given by (11. In Problem Set 5.451: Introduction to Economic Growth ρ. Similarly. with a representative household with preferences given by (11. It is deﬁned as total investment (increase in capital plus replacement investment) divided by output: s= ˙ +n+δ ˙ + δK k/k A − ρ + θn + (θ − 1)δ K = = Y A θA (11. One important implication of the AK model is that since all markets are competitive. you will also see that policy can now aﬀect the growth rate of output permanently.

the savings rate is constant in equilibrium as in the basic Solow model. as also discussed above. In this economy the unique equilibrium path in which consumption. that which aﬀects the rate of return to accumulation.1). to underlying preferences and to technology. which will now become: g= (1 − τ ) (A − δ ) − ρ . suppose that there is an eﬀective tax rate of τ on the rate of return from capital income. in this model.2 The Extended AK Model The model studied in the previous section is attractive in many respects. it responds endogenously to policy. 11. which responds to policy. θ Moreover.14. Repeating the analysis above immediately implies that this will adversely aﬀect the growth rate of the economy. The simplest and arguably one of the most relevant classes of policies is.6). with a representative household with preferences given by (11. it can be calculated that the savings rate will now be s= (1 − τ ) A − ρ + θn − (1 − τ − θ) δ .1. In particular. Therefore. but in contrast to that model. 11. 250 . and the production technology given by (11. θA which is a decreasing function of τ if A − δ > 0.451: Introduction to Economic Growth Proposition 30 Consider the above-described AK economy. capital and output all grow at the same rate g∗ ≡ (A − δ − ρ)/θ is Pareto optimal.4 The Role of Policy It is straightforward to incorporate policy into this framework. It generates sustained growth.

the Cobb-Douglas assumption here is quite important in ensuring that the share of capital in national income is constant [can 251 . however. The main diﬀerence is in the production technology.1)-(11. it is a very close cousin of the neoclassical model. One unattractive feature of this model. The preference and demographics are the same as in the model of the previous section. it is not that the production technology is AK . we now envisage an economy with two sectors. to simplify the analysis I will shut down population growth. (11. Rather than a single good used for consumption and investment.4) as will be discussed below). which in fact features a constant share of capital in national income less than 1.14. As I pointed out above. It also blurs what is the key underlying characteristic driving growth in this model. as argued there. In this section. the endogenous growth equilibrium is Pareto optimal. This makes it unattractive as an application to real world situations. Essentially.451: Introduction to Economic Growth Moreover. I will brieﬂy illustrate this by developing a more workable version of the AK model with two sectors. In fact. in particular. In fact. Moreover. but the related feature that the accumulation technology is linear. which has a Cobb-Douglas technology. equations (11. so n = 0.5) apply as before (but with a slightly diﬀerent interpretation for the interest rate in (11. is that all of national income accrues to capital.17) where the subscript “C ” denotes that these are capital and labor used in the consumption sector. Sector 1 produces consumption goods with the following technology C (t) = B (KC (t))α LC (t)1−α . and the total amount of labor in the economy is equal to L and is supplied inelastically. it is a one sector model with only capital as the factor of production.

there will be a relative price between the two sectors which will adjust endogenously. K where I (t) denotes investment. and LC (t) ≤ L. for labor (since labor is only used in the consumption sector). In the data. In particular. consumption and investment goods. An equilibrium in this economy is deﬁned similarly to that in the neoclassical economy.17).18). however.14. (11. The capital accumulation equation is given by: ˙ (t) = I (t) − δK (t) .18) The distinctive feature of the technology for the investment goods sector. though the capital intensities of many sectors have been changing over time as the nature of consumption and investment goods has changed. is that it is linear in the capital stock and does not feature labor. Investment goods are produced with a diﬀerent technology than (11. (11. for capital. Moreover. we can simplify notation by letting φ (t) denotes the share 252 . but also features an allocation decision of capital between the two sectors. we have I (t) = AKI (t) . Since both market clearing conditions will hold as equalities (the marginal product of both factors is always positive).451: Introduction to Economic Growth you see why?]. since the two sectors are producing two diﬀerent goods. This is an extreme version of an assumption often made in two-sector models that the investment-good sector is more capital-intensive than the consumption-good sector. there seems to be some support for this. Market clearing implies: KC (t) + KI (t) ≤ K (t).

(11. From proﬁt maximization. the Euler equation for consumers. still holds. which has now experienced a change in its price of p ˙I (t) /pI (t). the individual will buy 1/pI (t) units of capital goods. In other words. Let the price of the investment good be denoted by pI (t) and that of the consumption good by pC (t).4). let us choose the consumption good as the numeraire. the individual will get back the one unit of capital. pI (t) where gK is the steady-state (BGP) growth rate of capital. he will have to buy consumption goods. By giving up one unit of consumption. so that pC (t) = 1 for all t.14. it is the interest rate that measures how many units of consumption good an individual will receive tomorrow by giving up one unit of consumption today. but the relevant interest rate has to be for consumption-denominated loans. denoted by rC (t). then we have pI (t) A = pC (t) αB µ L (1 − φ (t)) K (t) ¶1−α . Moreover. the proper calculation goes as follows. whose prices changed by p ˙C (t) /pC (t).20) . Therefore.19) Deﬁne a steady-state (a balanced growth path) as an equilibrium path in which φ (t) is constant and equal to say φ. the general formula of the rate of return denominated 253 (11. Since the relative price of consumption goods and investment goods is changing over time. In addition. Then diﬀerentiating (11. This will have an instantaneous return of rI (t). and ﬁnally.451: Introduction to Economic Growth of capital used in the investment sector KC (t) = (1 − φ (t)) K (t) and KI (t) = φ (t) K (t).19) implies that at the steady state: p ˙I (t) = − (1 − α) gK . (11. the rate of return to capital has to be the same when it is employed in the two sectors. As noted above.

4). the steady-state consumption-denominated rate of return is: rC = A − δ − (1 − α) gK .22) . Substituting this into (11.21) Finally. pI (t) and in steady state. Finally. C (t) θ (11. from (11. Moreover. rI (t) =A−δ pI (t) given the linear technology in (11. we have rC (t) = A − δ + p ˙I (t) . from the constancy of φ(t) in steady state. we have p ˙C (t) /pC (t) = 0. this implies a consumption growth rate of gC ≡ ˙ (t) 1 C = (A − δ − (1 − α) gK − ρ) . p ˙I (t) /pI (t) is given by (11.451: Introduction to Economic Growth in consumption goods in terms of the rate of return denominated in investment goods is rC (t) = ˙I (t) p ˙C (t) rI (t) p + − .20). Therefore. C (t) KC (t) which.21). pI (t) pI (t) pC (t) In our setting.18). we have ∗ gK = A−δ−ρ 1 − α (1 − θ) 254 (11.14. diﬀerentiate (11. implies the following steady-state relationship: gC = αgK .17) and use the fact that labor is always constant to obtain ˙ (t) ˙ C (t) C K =α .20). given our choice of numeraire. From (11.

diﬀerent from the neoclassical growth model. in the balanced growth path. Capital grows at a faster rate than consumption 255 . Since labor markets are competitive.23) What about wages? Now since labor is being used in the consumption good sector. while the capital stock grows at the constant rate (11.451: Introduction to Economic Growth and ∗ =α gC A−δ−ρ .14. a lower discount rate will increase the equilibrium growth rate of the economy One important implication of this model. and as in the basic AK model. Therefore. It is straightforward to conduct policy analysis in this model. Similarly. taxes on investment income will depress growth. we obtain ˙ (t) K w ˙ (t) p ˙C (t) = +α w (t) pC (t) K (t) ∗ = αgK . Moreover. with exactly the same arguments as in the previous section.23). it can be established that there are no transitional dynamics in this economy. consumption and labor income grow at the constant rate given by (11. there will be positive wages. which implies that wages also grow at the same rate as consumption. starting from any K (0) > 0. the wage rate at time t is given by w (t) = (1 − α) pC (t) B µ (1 − φ (t)) K (t) L ¶α .22). is that there is continuous capital deepening. 1 − α (1 − θ) (11. This establishes the following result: Proposition 31 In the above-described extended AK economy.

capital-output ratio has been constant. Here we have steady state and “balanced growth” without this feature. 11. The Kaldor facts. I now present this model. His initial solution (later updated and improved in his and others’ work during the 1990s) was to consider knowledge as a byproduct of production that accumulates by itself. 11.3. which may explain why capital-output ratio has been constant in the earlier part of the century.451: Introduction to Economic Growth and output. Whether this is a realistic feature or not is debatable.1 Preferences and Technology Consider an economy without any population growth (we will see why this is important) and a production function with labor-augmenting knowledge (technology) that satisﬁes the standard assumptions. instead of working with the aggregate production function.14. but realized that this would be diﬃcult in the context of a competitive economy. Part of the reason why it has been increasing recently but not before is because of relative price adjustments. For reasons that will become clear. discussed above. New capital goods are of higher quality. let us look at the production function 256 . For much of the 20th century. but it has been increasing steadily over the past 30 years. but not recently. include constant capital-output ratio as one of the requirements of balanced growth.3 Growth with Externalities The model that started much all the interest in endogenous growth is Romer (1997). and this needs to be incorporated in calculating the capital-output ratio. These calculations have only been performed in the recent past. Assumptions 1 and 2. Romer wanted to explicitly model the process of knowledge accumulation.

e. R (t) = ∂K (t) w (t) = The key assumption of Romer (1997) is that although ﬁrms take A (t) as given. but spillovers work through human capital (i. Let us normalize the measure of ﬁnal good producers to 1. Lucas has human capital externalities). A (t) Li (t)) . this stock of technology (knowledge) advances endogenously for the economy as a whole. (11. The idea of externalities is not uncommon to economists. 0 where L is the constant level of labor (supplied inelastically) in this economy. so that we have the following market clearing conditions: Z 1 Ki (t) = K (t) 0 and Z 1 Li (t) = L. and moreover. which implies that they will all hire the same capital to eﬀective labor ratio. A (t) L) ∂L ∂F (K (t) . and attributes spillovers to physical capital. Romer assumes that this takes place because of spillovers across ﬁrms. since it is technology common to all ﬁrms. Firms are competitive in all markets. while Romer has physical capital externalities.14.24) where Ki (t) and Li (t) are capital and labor rented by a ﬁrm i. Notice that A (t) is not indexed by i. A (t) L) . thus ∂F (K (t) . but both Romer and Lucas make an extreme assumption of suﬃciently strong externalities such that A (t) can grow 257 .451: Introduction to Economic Growth facing each one of the many inﬁnitesimal ﬁnal good producers (each indexed by i): Yi (t) = F (Ki (t) .. Lucas (1998) develops a similar model in which the structure is identical. In particular. factor prices will be given by their marginal products.

451: Introduction to Economic Growth continuously at the economy level. Alternatively.24) and using the fact that all ﬁrms are functioning at the same capital-eﬀective labor ratio. the knowledge stock of the economy could be a function of the cumulative output that the economy has produced up to now. we have Y (t) = F (1.e. (11. In any case. we obtain the production function of the representative ﬁrm as Y (t) = F (K (t) .. where it is assumed that the manufacturing sector. substituting for (11. BL) K (t) ˜ (L) .25) i.25) into (11. ·) is homogeneous of degree one. making the production process itself more productive.14. The reason why the externalities work through capital might be justiﬁed along the lines of the structural change model we will discuss below. BK (t) L) . there is no compelling evidence that such externalities are very large). managers) in the production process. which is more capitalintensive. is more important for generating externalities (whether this is so or not is not very clear. Romer assumes A (t) = BK (t) . greater investments in certain sectors increases the experience (of ﬁrms. = f 258 . and in any case. This can be motivated by “learning-by-doing” whereby. the knowledge stock of the economy is proportional to the capital stock of the economy. workers. thus giving it more of a ﬂavor of “learning-by-doing”. Using the fact that F (·. In particular.

K (t)]∞ t=0 that maximize the utility of the representative household and wage and rental rates [w (t) . as a path of consumption and capital stock for the economy.27) (11. output per capita can be written as: y (t) ≡ Y (t) L Y (t) K (t) = K (t) L ˜ (L) . are external to the ﬁrm. = k (t) f where again k (t) ≡ K (t) /L is the capital-labor ratio in the economy.3. R (t)]∞ t=0 that clear markets. as speciﬁed in (11. they do not price the role of the capital stock in increasing future productivity. R (t) = R = f which is constant. it is also constant. The important feature is that because the knowledge spillovers. This immediately implies that consumption in this economy. [C (t) .26) 11. we have in terms of the normalized production function. marginal products and factor prices can be expressed ˜ (L). grows at the 259 . (11. now f ˜0 (L) w (t) = K (t) f and ˜ (L) − Lf ˜0 (L) . factor prices are given by (11.27)–that is.26) and (11.451: Introduction to Economic Growth Alternatively.25). As in the standard growth model.2 Equilibrium An equilibrium is deﬁned similarly to the neoclassical growth model.14. Since the market rate of return is r (t) = R (t) − δ . given by the usual Euler equation. In particular.

To do this.29) and (11. 1−θ (11. Moreover. ∗ = gC It is also clear that capital grows exactly at the same rate as consumption. capital. the economy will not admit a steady state and the growth rate of the economy will increase over time (output reaching inﬁnity in ﬁnite time and violating the transversality condition). ˜ (L) − Lf ˜0 (L) is always increasing in L (by Assumption 1). in particular. if population is growing constantly.30) are satisﬁed. ˜ (L) − Lf ˜0 (L) < f ρ + δ. output and consumption growth are all given by gC as given by (11. there are no transitional dynamics in this model.30) It is also straightforward to verify that as in the AK model above. f (L) − Lf θ (11. output and consumption grow at the constant rate (11.28).28) Let us assume that ˜ (L) − Lf ˜0 (L) − δ − ρ > 0. This establishes: Proposition 32 In the above-described Romer model with physical capital externalities. but also that growth is not fast enough to violate the transversality condition.28).29) so that there is positive growth. ﬁrst. f (11. there exists a unique equilibrium path where starting with any level of capital stock K (0) > 0. You can also see now why population was assumed constant in this model.14. 260 . in that when population (labor force) L is higher. the growth rate of the since f economy will increase. ´ 1³˜ ˜0 (L) − δ − ρ . as long as conditions (11.451: Introduction to Economic Growth constant rate. so the rate of ∗ capital. note that there is a scale eﬀect here.

k The current-value Hamiltonian is h i c1−θ − 1 ˜ ˆ + µ f (L) k − c − δk . noting that the per capita accumulation equation for this economy can be written as ˙ =f ˜ (L) k − c − δk.451: Introduction to Economic Growth 11. h i ˜ ˆ Hk (k. These equations imply that the social planner’s allocation will also have a constant growth rate for consumption (and output) given by S = gC ´ 1³˜ f (L) − δ − ρ . µ) = 1−θ and has the necessary conditions: ˆ c (k. c. H (k. Since this eﬀect is external to the ﬁrms. let us again set up on the currentvalue Hamiltonian.28)–since f which is always greater than gC sentially. µ) = 0 = c−θ − µ H t→∞ lim [exp (−ρt) µ (t) k (t)] = 0. it is not surprising that the decentralized equilibrium characterized in Proposition 32 is not Pareto optimal.3 Pareto Optimal Allocations Given the presence of externalities. θ ∗ ˜ (L) > f ˜ (L) − Lf ˜0 (L).14. the social planner takes into account that by accumulating more capital. the decentralized economy fails to internalize this externality. Therefore we have: 261 . c.3. Esas given by (11. To characterize the allocation that maximizes the utility of the representative household. µ) = −µ ˙ + ρµ = µ f (L) − δ . she is improving the productivity in the future. c.

262 . the decentralized equilibrium is Pareto suboptimal and grows at a slower rate than the allocation that would maximize the utility of the representative household.451: Introduction to Economic Growth Proposition 33 In the above-described Romer model with physical capital externalities.14.

Agriculture becomes less important. fuels further growth. The process of development. Many economic. is also one of the transformation of the economy. or perhaps “cumulative causation” (where an economic process becomes selfsustaining once underway) going on. the process of economic development is not simply a linear sustained growth process. To do justice to these topics. Simultaneously. and the historical process of economic growth leading to the modern world. which is important both for understanding why some countries are much richer today than others. However. which are 263 . there is a process of coordination. Urbanization increases. social and economic institutions also change in the process. manufacturing becomes more important (and then later services become more important). in which the increase in demand for certain goods and services (especially coming from cities). we need to delve much deeper into issues of development economics and political economy.Chapter 12 Multiple Equilibria and the Process of Development The models discussed so far generated sustained economic growth. as emphasized by Simon Kuznets.

or multiple steady states. However. Here we start with a simple two-period model of an economy with monopolistic competition. The model is a version of Murphy. As the name of the paper suggests.451: Introduction to Economic Growth beyond the scope of the current course. we can start getting a sense of these processes by quickly looking at models of economic development emphasizing multiple equilibria.1 Preferences and Technology Consider the following two-period economy. θ plays a similar to before. the focus there will not be on multiple equilibria. and also looking at a very simple version of a model of structural change incorporating the features emphasized by Simon Kuznets. which will lead to multiple equilibria. 12.1. with 1/θ being the intertemporal elasticity of substitution. regulating how willing individuals are 264 . Shleifer and Vishny’s (1989) “Big Push” paper. since ﬁrms that discover new machines will become the monopolistic suppliers of these machines or of goods produced with these machines. Below in discussing models of endogenous technological change. the idea is to think of the development process as a move from one equilibrium to another.14. But. 12. monopolistic competition will play a crucial role. All agents have preferences given by U= 1−θ −1 C1 C 1−θ − 1 +β 2 1−θ 1−θ where C1 and C2 denote consumption at the two dates. a big push.1 Multiple Equilibria From Aggregate Demand Externalities Let us start with a very simple model of multiple equilibria arising from aggregate demand externalities. likely due to a coordinated move.

This production function has the standard love-for-variety feature ﬁrst introduced by Dixit and Stiglitz. and wt is the wage rate at time t. Although individuals can borrow and lend. Individuals can borrow and lend. In particular. the resource constraints have to hold. and the aggregate production at time t is given by: Yt = ∙Z 1 ε ¸ ε− 1 yt (i) ε−1 ε di 0 where yt (i) is the output level of intermediate i at date t. R is the gross interest rate. and investment is only possible in the ﬁrst date. so an individual’s budget constraint is c1 + c2 w2 + π 2 ≤ w1 + π 1 + . Its advantage is that it provides an extremely tractable model of substitution between diﬀerent 265 . The new feature in this model is that output is an aggregate of intermediates.451: Introduction to Economic Growth to substitute consumption between date 1 and date 2. where I1 denotes investment in the ﬁrst date. The resource constraint for the economy is C1 + I1 ≤ Y1 C2 ≤ Y2 . in the aggregate. Yt is total output at date t. R R where π t denotes the proﬁts accruing to the representative consumer. so R will be determined in equilibrium to ensure this. This functional form can be used either for aggregating intermediates or directly as a utility function. and β is the discount factor of the households. with their total measure normalized to 1. there is a continuum of diﬀerentiated intermediate goods.14.

266 .. ε > 1. there is a designated producer for each intermediate.1) where α > 1 and lt (i) denotes labor devoted to the production of intermediate good i at time t. For now. At date 1.1). thus the designated producer will have some degree of monopoly power. the fringe will not beneﬁt from this technological improvement. the designated producer can also invest in the new technology. because the elasticity of demand for each good is constant. which costs F per ﬁrm. naturally. All ﬁrms are assumed to be owned equally by all the consumers. Labor market clearing. We will make extensive use of these preferences in the rest of the course. In contrast. but a competitive fringe can also enter and produce each good as productively as the designated producer. The production function of each good is as follows: y1 (i) = l1 (i) and ⎧ ⎨ l (i) with old technology 2 y2 (i) = ⎩ αl (i) with new technology 2 Z (12. this producer’s productivity at date 2 will be higher by a factor α as indicated by equation (12.2) At date 1. both with competition and monopolistic competition. If this investment is undertaken. supplied inelastically. requires 1 0 lt (i) di ≤ L (12.14.451: Introduction to Economic Growth goods. They will maximize proﬁts taking the market prices (especially the market interest rate) as given. The economy has a total labor supply of L. i. note that ε is the elasticity of substitution between intermediate goods within a given period. and is assumed to be strictly greater than one.e.

At date 2. we have Y2 = L and when the technology is adopted by all the ﬁrms.14. i. l2 (i) = L for all i ∈ [0. the designated producers have no monopoly power because of the competitive fringe. we have Y2 = αL. wages for both periods and an interest rate linking consumption between the two periods. Since we are looking at the symmetric equilibrium (SSPE). so labor will be allocated equally. to simplify the discussion. the ﬁrst period labor market clearing is straightforward and we will have l1 (i) = L for all i ∈ [0. In the ﬁrst date.2 Equilibrium Since this is a two-period economy. We now turn to the pricing decisions. when the technology is not adopted.e. SSPE. In either case. 1] (recall that the measure of sectors and ﬁrms is normalized to 1). we will be looking for a subgame perfect equilibrium. since all goods are symmetric. This implies that Y1 = L. Moreover. thus they charge price equal to marginal 267 . we only consider the two extremes where all ﬁrms adopt and no ﬁrm adopts. An SSPE consists of an allocation of labor across ﬁrms. 1] . the equilibrium will depend on how many ﬁrms have adopted the new technology. First. again the marginal productivity of all sectors are the same.. let us focus on symmetric subgame perfect equilibria.1. investment decisions for ﬁrms. Consequently.451: Introduction to Economic Growth 12.

Let us ﬁrst ﬁnd the demand facing each producer. Now they can produce α units of output with one unit of labor. while the fringe of competitive ﬁrms still produces one unit of output with one unit of labor.4) ˆ as given in (12.3). the same situation repeats. Since total output is equal to Y1 = L. and make zero proﬁts. so consumption at both dates is equal to L. if the gross interest rate is R Next consider the situation in which the designated producers have invested in the advanced technology. This implies that the designated producers have some monopoly power. this also implies that the equilibrium wage rate is equal to w1 = 1. and we have w2 = 1 and no proﬁts. The extent of this monopoly power depends on the comparison of ε and α.1] y2 (i) ε−1 ε 0 0 268 . In this case there is also no investment.14.3) (12. which can only be satisﬁed with C1 = C2 . In the second date. recall that the standard Euler equation in this case is −θ −θ = RβC2 . thus the interest rate that makes individuals happy to consume this amount in both periods is ˆ = β −1 .451: Introduction to Economic Growth cost. which is given as a solution to the following program of proﬁt maximization for the ﬁnal goods sector: max ∙Z 1 ε ¸ ε− Z 1 1 di − p2 (i) y2 (i) di. [y2 (i)]i∈[0. R To see this more formally. if the technology is not adopted. which is w1 . C1 (12.

This expression is useful in laying the foundations for the aggregate demand externalities. In that case. the demand for good i depends on the total amount of production. ﬁrst imagine the situation in which there is no fringe of competitive producers. So why is there an externality here?] A nice feature of the demand curve implied by equation (12. (12.14. which we will discuss soon. This will be a very convenient feature in many of the models using this class of utility or production functions below. π2 (i) = p2 (i) − α substituting from (12. The ﬁrst-order condition to this program implies y2 (i)−1/ε Y2 or y2 (i) = (p2 (i))−ε Y2 . you should ask yourself why this actually causes an externality. Y2 . each designated producer will act as an unconstrained monopolist and maximize its proﬁts given by price minus marginal cost times quantity.e..451: Introduction to Economic Growth where p2 (i) is the price of intermediate i at date 2. max π 2 (i) = p2 (i) − p2 (i) α which has a ﬁrst-order condition (p2 (i)) −ε ³ w2 ´ Y2 − ε p2 (i) − (p2 (i))−ε−1 Y2 = 0.e. the demand elasticity is constant). To make more progress.5). i. even with perfectly competitive markets.5) 1/ε = p2 (i) . α 269 . [However. the ﬁrm maximization problem is ³ w2 ´ (p2 (i))−ε Y2 . the demand for my goods may depend on the supply of other goods in the economy. ³ w2 ´ y2 (i) ..5) is that it is iso-elastic (i.

the monopolist can only charge this price if ε 1 ≤ 1. each monopolist would make per unit proﬁts equal to w2 − w2 α−1 = w2 . and thus increase its proﬁts. ε−1α Under this assumption. ε−1 α This is the standard monopoly price formula of a markup related to demand elasticity over the marginal cost. w2 /α. the monopolist could increase its price without losing the market. let us impose Assumption 13 ε 1 > 1. In other words. If it were any lower. Let us assume that α is not so high as to make the monopolist unconstrained. If it were any higher. Since the competitive fringe can produce one unit using one unit of labor. However. Here the markup is constant because the demand elasticity is constant. α α 270 .451: Introduction to Economic Growth which implies p2 (i) = pM 2 ≡ ε w2 .14. the monopolist can only charge this price if the competitive fringe could not enter and make proﬁts stealing the entire market at this price. steal the whole market and make positive proﬁts. This implies that under Assumption 13. ε−1α Otherwise. the price would be too high and the competitive fringe would enter. the competitive fringe would enter. It is straightforward to see that this equilibrium limit price would be p∗ 2 = w2 . the monopolist will be forced to charge a limit price.

as in the case without the technological investments. now implies ˜ (αL)−θ . Nevertheless.6) The wage rate can be determined from income accounting. α which has an equilibrium at w2 = 1. because there was investment in the new technology at date 1.4). all of these proﬁts are redistributed to the agents.14. C1 = L − F . (12.e.7) αL L−F ˆ > R. i. in this economy the increased marginal product does not translate into higher wages. who are the owners of the ﬁrms. Again the interest rate has to adjust so that individuals are happy to consume these amounts. However. the interest rate in this case is higher than the one in which there is no investment. This is natural. α (12. and this has to be distributed between proﬁts and wages. it leads to proﬁts for ﬁrms. Instead. Consequently. (L − F )−θ = Rβ which solves for ˜ = β −1 R µ ¶θ (12.5) as: π2 = α − 1 1−ε w2 Y2 . since investment implies that individuals are being asked to forgo date 1 consumption for date 2 consumption. Note also that the greater is θ. Total production will be equal to Y2 = αL. The Euler equation. thus α − 1 1−ε w2 αL + w2 L = αL.. the higher 271 .451: Introduction to Economic Growth The proﬁts of ﬁrms are then obtained from substituting from (12. Therefore. Thus C2 = αL. so that they have a steep consumption proﬁle without wanting to borrow.

and consider the incentives of a single designated ﬁrm to undertake such an investment. The reason for the possibility of multiplicity is that the answer to this question will depend on whether other ﬁrms are undertaking the investment or not. In this case. In this case total output at date 2 is equal to L (since the ﬁrm considering investment ˆ . The question is whether ﬁrms will ﬁnd it proﬁtable to undertake the investment at date 1. proﬁts at date 2 are πI 2 = (α − 1) L. Moreover.451: Introduction to Economic Growth ˜ . Therefore.6) and the is inﬁnitesimal). = −F + β L−F 272 .14. there is less intertemporal substitution. α where the superscript N denotes that no other ﬁrm is undertaking the investment. = −F + β α Next consider the case in which all other ﬁrms are undertaking the investment. Let us ﬁrst take a situation in which no other ﬁrm is undertaking the investment. since with a greater θ. the proﬁt gain from investing at date 1 is ∆π I = −F + 1 (α − 1) L ˜ R ¶−θ µ αL (α − 1) L. where the superscript I designates that all other ﬁrms are undertaking the investment. the net discounted proﬁts at date 1 for the ﬁrm in question is ∆π N = −F + 1 α−1 L ˆ α R α−1 L. Also a higher F . Consequently. meaning is R a greater consumption sacriﬁce at date 1 implies a higher interest rate. from (12. proﬁts at date 2 are πN 2 = α−1 L. and the market interest rate is given by R fact that w2 = 1.

In this context. while the other one corresponds to industrialization. when nobody else invests. since condition (12. the existence of multiple equilibria requires the interest rate eﬀect not to be too strong. where one of the equilibria corresponds to backwardness.9) It is also straightforward to see that whenever both equilibria exist. the condition for the existence of multiple equilibria is that: β µ αL L−F ¶−θ α−1 L. Counteracting this eﬀect is the fact that the interest rate is also higher when all ﬁrms invest.14. similar to the ideas of many economists writing on development before them. was to generate multiple equilibria. the fact that π I > π N . the equilibrium with investment Pareto dominates the one without investment. and when all other ﬁrms invest.8) is certainly possible. in the extreme case where preferences are linear. For example. they produce more. This is clearly possible because of the aggregate demand externality. we have that ∆π I = −F + β (α − 1) L > ∆πN = −F + β α−1 L. More generally. (12.. when other ﬁrms invest. α so (12. This is only possible if we have ∆π N < 0 and ∆π I > 0.8) that is. this means that for the same parameter values both no investment in the new technology and all ﬁrms investing in the new technology should be equilibria. 273 .9) implies that all households are better oﬀ with the upward sloping consumption proﬁle giving them higher consumption at date 2. Therefore. α (α − 1) L > F > β (12. investment is proﬁtable. there is more aggregate demand. θ = 0. i. the idea of the paper by Murphy.e. investment is not proﬁtable. Shleifer and Vishny (1989). and therefore proﬁts from having invested in the new technology are higher.451: Introduction to Economic Growth As discussed above.

The reason why they take the form of externalities is that the ﬁrm does not realize the full increase in the social product created by its investment. for example. the model here is essentially a static one. In turn.451: Introduction to Economic Growth This leads to the following result: Theorem 40 Consider the above-described environment and suppose that Assumption 13 holds and condition (12. that provided by economists such as Nurske or Rosenstein-Rodan. these linkages take the form of aggregate demand externalities. Intuitively. One interpretation of this result is that societies that can somehow coordinate on the equilibrium with investment (either because private expectations are aligned or because of some type of government action) will industrialize and realize both economic growth and Pareto improvement. multiple equilibria arise because (when) there is substantial aggregate demand at date 2 so that investing in the new technology at date 1 is proﬁtable. so that they are more productive and produce more at date 2. In fact.14. thus turning the demand linkages into aggregate demand externalities. so it does not allow a literal interpretation of a society being ﬁrst in the no investment equilibrium 274 . as noted above. because the monopoly markup implies that at the margin. The presence of the markup means that the monopolist does not internalize this ﬁrst-order gain. The equilibrium with investment Pareto dominates the equilibrium without investment. there will be substantial aggregate demand at date 2 when all ﬁrms invest in the new technology. further increases in output create a ﬁrst-order gain for consumers.9) is satisﬁed. Then there exist two pure strategy SSPE. one in which all ﬁrms undertake the investment at date 1 and the other one in which no ﬁrm does. This intuition highlights the importance of aggregate demand linkages. Naturally. and this corresponds to the “big push” ideas suggested by qualitative accounts of the early development process.

it is suggestive of such a process. when the economy is subject to credit market problems. Nevertheless. in the form of subsidies to ﬁrms. a topic that brings us to the political economy of development and economic growth. partly because it is not clear which sectors need to be subsidized.14. for example. leading to multiple equilibria. I will illustrate these issues in the simplest possible way looking at the eﬀect of credit market problems on human capital investments.” changing the coordination to the high-investment equilibrium.451: Introduction to Economic Growth and then changing to the investment equilibrium and thus industrializing. Also. credit market problems will illustrate how the distribution of income (and the incidence of poverty) in a society might aﬀect economic growth and the process of economic development. Similar issues arise. Underdevelopment may be thought to correspond to a situation in which the coordination is on the bad equilibrium. Moreover. Investment by diﬀerent ﬁrms may require coordination. and perhaps more importantly because government interventions are often captured by interest groups. 12. although the model makes it sound as if simple government action.2 Human Capital Accumulation with Imperfect Capital Markets The previous section illustrated the potential of development traps because of aggregate demand externalities. and the development process starts with the “big push. in practice government intervention is not easy. in a more dynamic way. might realize such a big push. 275 .

2. In other words.1 A Simple Case With No Borrowing When credit markets are imperfect.451: Introduction to Economic Growth 12. a major determinant of human capital investments will be the distribution of income (as well as the degree of imperfection in the credit markets). where w is the wage income of the individual. and may also selfperpetuate. 276 . The budget constraint is i i ci t + et+1 ≤ wt . here education. but simply about what they bequeath to them. There is consumption only at the end of adulthood. this utility function features impure altruism (sometimes referred to “warm glow” preferences): parents do not care about the utility of their oﬀspring. There are a number of important features embedded in this utility function: 1. childhood and adulthood. and e is the educational spending on the oﬀspring of this individual. Even though it is a very similar utility function to that we worked with in the overlapping generations model. and gets an oﬀspring in his adulthood. now the utility function refers to the utility that an individual obtains from his consumption and the indirect utility he obtains from leaving something to his oﬀspring. Preferences are given by i (1 − δ ) log ci t + δ log et+1 where c is consumption at the end of the individual’s life.14. I start with a discussion of the simplest case with no borrowing (extreme credit market problems) to illustrate how the distribution of income will matter. Consider an economy with a continuum 1 of dynasties. Each individual lives for two periods.

10) 277 . Given this description. It is logarithmic. 1) is some minimum level of human capital that the individual where α ∈ (0. The labor market is competitive. ht+1 = ⎩ h ¯ if ei < 1 t ¯ ∈ (0.451: Introduction to Economic Growth 2. which. Each individual will choose the spending on education that maximizes its own utility. and wage income simply depends on human capital: i = Ahi wt t Human capital of the oﬀspring of individual i of generation t in turn is given by ⎧ ⎨ (ei )α if ei ≥ 1 t t i . The budget constraint of individual i of generation t is: i i ci t + et ≤ wt . a nonconvexity in the technology of human capital accumulation. as with the two-period overlapping generations model. 1) and h will attain even without any educational spending. the equilibrium is straightforward to characterize. (12. will lead to constant savings rates. This equation introduces a crucial feature necessary for models of credit market imperfections to generate multiple equilibria or multiple steady states. the individual starts beneﬁting from the additional spending and accumulates further human capital (though with diminishing returns since α < 1).14. Once spending exceeds a certain level (here set equal to 1). This immediately implies the following “savings rate”: i i ei t = δwt = δAht .

Therefore. ∗ (as long as hi 0 < h . If at time −1 i i ¯ 0.11) Now.14. Then from (12. It is interesting to contrast two economies subject to the credit market problems. The most important result is that this simple model features poverty traps due to the nonconvexities created by the credit market problems. For example. but with diﬀerent distributions of income. we have hi 1 = h < (δA) .11).451: Introduction to Economic Growth This rule has one unappealing feature (not crucial for any of the results).10) implies that et < 1. imagine an economy with two groups starting at income levels h1 and h2 > h1 such that (δA)−1 < h2 . we have h1 = (δAh0 ) > α 1.11). and repeating this argument. in which case educational spendings are in fact wasted (do not translate into higher human capital of the oﬀspring). then (12. −1 −1 i ¯ Given (12. we have hi 0 < (δA) . (12. we have ht < (δA) −1 for all t. which is that because parents derive utility from educational spending on their children. they will spend on education even when ei t < 1. a signiﬁcant fraction of the population will never accumulate 278 . so the oﬀspring will have h1 = h. level greater than h −1 i i Next consider a dynasty with hi 0 > (δA) . so this dynasty will accumulate human capital and reach the “steady state” given by h∗ = (δAh∗ )α or h∗ = (δA) 1−α > 1. To obtain stark results let us also assume that ¯ δA > 1 > δAh. let us look at the dynamics of human capital for a particular dynasty i. Now if inequality (poverty) is high so that h1 < (δA)−1 . the dynasty would have started with too much human capital α and would decumulate human capital). otherwise. a dynasty that starts with hi will never reach a human capital 0 < (δA) ¯.

The budget constraint is i ci t + bt ≤ m 279 . he can either work or acquire education. In contrast. it depends on whether greater inequality pushes more people to below or above the critical thresholds. This is a general feature: in models with nonconvexities. neither group will accumulate human capital. 12. so that we push group 2 to h2 > (δA)−1 would increase human capital accumulation. and the above example with two classes seems to support this conclusion. take the same economy with two classes. this is not a general result. if inequality is limited so that h1 > (δA)−1 .2. now starting with h1 < h2 < (δA)−1 . all agents will accumulate human capital. In his youth.14.451: Introduction to Economic Growth much human capital. and redistributing resources away from group 1 to group 2 (thus increasing inequality). For example. In this case. This model and the next one (with imperfect capital markets) are sometimes interpreted as implying that an unequal distribution of income will lead to lower output (and growth). Each individual still lives for two periods. where again c denotes consumption at the end of the life of the individual. The utility function of each individual is i (1 − δ ) log ci t + δ log bt .2 The Galor and Zeira Model Now let us allow borrowing in the model above. However. An important implication of this model is that the distribution of income and how credit markets work are important for human capital accumulation and the process of economic growth. there are no general results about whether greater inequality is good or bad for accumulation and economic growth. eventually reaching h∗ .

Let us now write the utility of this agent (with x < h) in the two scenarios. and also the 280 . i. It will now be the individuals themselves who will use the monetary bequests to invest in education. which ﬁxes the lending rate at some constant r. This creates a wedge between the borrowing and the lending rates. If x < h. Education is a binary outcome. However. The required education expenditure to become skilled is h. Also. and workers acquiring education do not earn the unskilled wage.14. and educated (skilled) workers earn wage ws while uneducated workers earn wu .451: Introduction to Economic Growth where m is the individual’s income. assume that there is a linear savings technology open to all agents. Also assume that ws − (1 + r) h > wu (2 + r) (12. the logarithmic formulation will once again ensure a constant savings rate equal to δ .12) implies that individual will invest in education. If x ≥ h. because of costs of monitoring necessary to induce agents to pay back the loans. Let us now consider an individual with wealth x.12) which implies that investment in human capital is proﬁtable when ﬁnanced at the lending rate r. during the ﬁrst period of their lives. wu . assumption (12. the borrowing rate is i > r. Imperfect capital markets are modeled by assuming that there is some amount of monitoring required for loans to be paid back. Note that utility of the parent now depends on monetary bequest to the oﬀspring rather than the level of education expenditure. In particular. then whether it is proﬁtable to invest in education are not will depend on the wealth of individual and the borrowing interest rate.

281 . The correspondence (12.14. when he invests in education. constrained-investing and constrained-non-investing agents. the equilibrium correspondence describing equilibrium dynamics is ⎧ ⎪ ⎪ b (x ) = δ (ws + (1 + r) (xt − h)) if xt ≥ h ⎪ ⎨ n t xt+1 = (12.13).451: Introduction to Economic Growth bequest that he will leave to his oﬀspring. D is a constant term.13). This is because dynamics in this economy are “Markovian”–described simply by the Markov process without any general equilibrium interactions.13) describes the behavior of the wealth of each individual. However. More speciﬁcally. These utility levels and bequests are given by Us (x) = log (ws + (1 + i) (x − h)) + D bs (x) = δ (ws + (1 + i) (x − h)) . Note an important feature here. Comparing these expressions we obtain that an individual likes to invest in education if and only if x≥f ≡ (2 + r) wu + (1 + i) h − ws i−r The dynamics of the system can then be obtained simply by using the bequests of unconstrained. the whole wealth distribution can also be studied from (12. when he chooses not to invest.13) bs (xt ) = δ (ws + (1 + i) (xt − h)) if h > xt ≥ f ⎪ ⎪ ⎪ ⎩ b (x ) = δ ((1 + r) (w + x ) + w ) if xt < f u t u t u Equilibrium dynamics can now be analyzed diagrammatically by looking at the graph of (12. And Uu (x) = log ((1 + r) (wu + x) + wu ) + D bu (x) = δ ((1 + r) (wu + x) + wu ) .

The distribution of income again has a potentially ﬁrst-order eﬀect on the income level of the economy. i.e. Such an intersection will exist when the borrowing interest rate. i to be smaller given r.13) with the 45 degree line. we may expect the wedge between the borrowing rate and the lending rate to be smaller.13) and the 45 degree line where (12. It is also clear that ﬁnancial development should matter for human capital investments.14. is large enough. All individuals with xt < g converge to the wealth level x ¯U . In an economy with better ﬁnancial institutions. As in the example without credit markets. the economy will have low productivity. 282 . there is a poverty trap which attracts agents with low initial wealth.. while all those with xt > g converge to the greater wealth level x ¯S . With a smaller i. If the majority of the individuals start with xt < g .13) is steeper). i. when the equilibrium correspondence is steeper than the 45 degree line. more agents will escape the poverty trap.451: Introduction to Economic Growth Now deﬁne g as the intersection of the equilibrium correspondence (12. and in fact the poverty trap may not exist (there may not be an intersection between (12. low human capital and low wealth.

is not indexed by time. (12.1 Demographics. Technologies in the two sectors are given by the following diminishing returns production functions X M (t) = M (t) F (n (t)) F (0) = 0.3.14. F 00 < 0.15) where n (t) is the fraction of labor employed in manufacturing as of time t. which is the name given to the feature that the budget share of food declines as individuals become richer. A. G0 > 0. 283 . The most standard reason for this is thought to be Engel’s law. Pretty much all societies have started as agricultural economies. and labor is supplied inelastically. F 0 > 0. with the share of output of manufacturing (and services) increasing.14) (12. This way of writing the two production functions already imposes market clearing in the labor market. is that of structural change. and have grown together with a transformation of the economy. consisting of two sectors: manufacturing and agriculture. G(0) = 0. hence it is constant. Population is constant and equal to L = 1. 12. an important element of the process of economic development. X A (t) = AG(1 − n (t)). G00 < 0.3 Learning-by-Doing. Preferences and Technology Consider the following continuous time economy. especially starting from the early stages of development. Here I will outline a model by Matsuyama (1992). Structural Change and NonBalanced Growth As mentioned above. which incorporates both this feature and the possibility of learning-by-doing as an important factor in economic growth.451: Introduction to Economic Growth 12. Both sectors produce using only labor. Notice that agricultural productivity.

because greater production in manufacturing allows learning-by-doing in this sector. The parameter ρ is the discount factor. w (t) = AG0 (1 − n (t)) and w (t) = p (t) M (t) F 0 (n (t)) where p (t) is the relative price of the manufactured good (with the price of the agricultural goods normalized to 1 as the numeraire). this implies. increasing future productivity. learning-by-doing eﬀects are external to individual ﬁrms. Therefore. γ and ρ > 0. Consequently. and cA (t) denoting the consumption of the agricultural good and cM (t) denoting the consumption of the manufacturing good at time t. market clearing implies: (12.451: Introduction to Economic Growth Manufacturing productivity.18) with β . M (t) reﬂects knowledge accumulation taking place as a noninternalized byproduct of production.17) (12. is time-varying. M where δ > 0 measures the extent of these learning-by-doing eﬀects. for example. Moreover. M (t). Assuming an interior solution.14. 0 (12. w (t). More speciﬁcally. we have: ˙ (t) = δX M (t) . The economy admits a representative consumer with preferences given by Z ∞ ¤ £ W = β log(cA (t) − γ ) + log(cM (t)) exp (−ρt) dt. Matsuyama assumes that this knowledge accumulation beneﬁts only from production in the manufacturing sector.16) AG0 (1 − n (t)) = p (t) M (t) F 0 (n (t)). and β designates the importance of agricultural goods versus manufacturing 284 . as in the Romer (1997) model discussed above. In particular. each ﬁrm will choose its labor demand in order to equate the value of the marginal product to the wage rate. As in the Romer model.

In particular. This is the simplest way of introducing Engel’s law.19) The ﬁrst inequality states that the economy’s agricultural sector is productive enough to provide the subsistence level of food to all consumers–otherwise individuals would receive negative inﬁnite utility. Let us also assume that AG(1) > γL > 0.3.2 Equilibrium An equilibrium is deﬁned in the standard way as a sequence of consumption levels in the two sectors and allocation of labor between the two sectors at all dates. The presence of γ > 0 makes preferences non-homothetic and implies that the income elasticity of demand for agricultural goods will be less than unity (while that for manufacturing goods will be greater than unity). such that consumers maximize their utility and ﬁrms maximize proﬁts given prices. The budget constraint of consumers in each period is cA (t) + p (t) cM (t) ≤ w (t) + π (t) where π (t) is the proﬁts per representative household. and goods and factor prices are such that all markets clear. 285 . (12. The parameter γ is the new one relative to models we have seen so far and represents the subsistence level of food consumption. 12.14. the individual will obtain negative inﬁnite utility (recall log (negative number) is undeﬁned).451: Introduction to Economic Growth goods in the utility function. imagine that if cA (t) does not exceed γ .

with v 0 (A) > 0.14.17) and (12.21) is decreasing in A. Since production has to be equal to consumption.19) it is clear that the equilibrium condition (12. An equilibrium has to satisfy (12.20) . we have cA (t) = γ + βp (t) cM (t) .21) (12. 1) .22) (12. Since the right-hand side of (12.20) yields φ(n (t)) = γ/A. where φ(n) ≡ G(1 − n) − βG0 (1 − n)F (n)/F 0 (n). The function φ (n) can be interpreted as the “excess demand” for manufacturing over agriculture. Moreover. φ(1) < 0 and φ0 (·) < 0. 286 (12.21) has a unique interior solution in which n (t) ∈ (0. A: n (t) = v(A). we further have: cA (t) = X A (t) = AG(1 − n (t)) and cM (t) = X M (t) = M (t) F (n (t)) Now combining these equations with (12. From Assumption (12.21).451: Introduction to Economic Growth Maximization of (12.18) implies that for each household. this solution can be written as a function of agricultural productivity. and thus for the entire economy. we have φ(0) = G(1).

the reasoning is simple: manufacturing requires a suﬃciently large size of employment to grow rapidly (either for creating aggregate demand externalities or for learning-by-doing). δF (v(A)). This is an interesting observation. it generates another feature which is consistent with the empirical patterns in the data. and this can only be achieved if agriculture is productive enough that suﬃcient food can be produced by a relatively small fraction of the workforce. given the learning-by-doing in equation (12.16). However. Therefore. this implies that higher agricultural productivity also increases agricultural consumption. Since productivity and employment in agriculture are constant. the share of manufacturing output (and consumption of manufacturing goods) increases relative to those of agriculture. which emphasize how economies with high agricultural productivity were those that were able to make the transition to manufacturing. this discussion leads to the following simple result: 287 . F 0 (v (A)) which is also increasing in A. In those accounts and in this model. This observation is consistent with some historical accounts of the development process. This is not in line with the patterns we observe in the data. also positively related to A. given the learning-by-doing aspect.14. output in manufacturing grows at a constant rate.451: Introduction to Economic Growth This implies that the employment share of manufacturing is constant over time and positively related to A. aggregate food consumption and production stay constant at cA = X A = AG(1 − v(A)) = γ + AβG0 (1 − v(A)) F (v(A)) . and shows that the growth rate of output in manufacturing is positively related to productivity in agriculture. where the manufacturing share of employment also increases early on (and then declines while the share of services increases). In particular.

the combination of learning-by-doing and Engel’s law generate a unique equilibrium in which the share of employment of manufacturing and agriculture are constant. 288 . the expenditure on agricultural goods will not remain constant. because their output is becoming scarcer in the economy. in fact independent of Engel’s law. Although this proposition shows that the real consumption of agricultural goods is constant (and that of manufacturing goods is increasing). because relative prices will change in favor of agricultural goods. and manufacturing output and consumption grow faster than agricultural output and consumption. while the growth rate of manufacturing output is δF (v(A)).14. sectors that experience slower growth will also experience increases in their relative prices. This is a general phenomenon.451: Introduction to Economic Growth Proposition 34 In the above described model. The growth rate of real consumption of agriculture is zero.

not interacting with the rest of the countries in the world. and has proved a useful reduced-form model for many applications. creating interdependences across growing countries. I look at how international trade inﬂuences the process of economic growth. we have a ﬁrst look at some models of interdependences.1 Human Capital and Technology (Nelson-Phelps) The Nelson-Phelps model is the simplest model of technology diﬀusion across countries. In this chapter. First. Finally. This is clearly not the correct way to view the world. I begin with a model of technology transfer from an exogenously advancing world technology frontier.Chapter 13 Interdependence and Growth in the Open Economy The analysis so far treated each country as a closed island. I discuss a model of technology transfer and trade. 13. Then. In addition to its growth 289 .

The ﬁrst implication of (13. Nelson and Phelps postulate ˙ j (t) φ (hj ) (T (t) − Aj (t)) A = . For example. T (t) = T (0) exp (gt) . the faster is its rate of progress. ∂T (t) ∂φ (hj ) so that human capital becomes more valuable when frontier technology is more advanced.1) is in terms of technological progress. Aj (t) Aj (t) where hj is the human capital in country j . This equation states that the farther a country is from the world technology frontier. In the 290 (13.. But also φ0 (hj ) > 0 so that.1) . since there is more technology out there to be absorbed. which is assumed to be time invariant. But this is a human capital-intensive task. advancing at an exogenous rate g . note that although equation (13.e. T (t). Second. the faster will this convergence be. to ﬁll key positions in the implementation of these technologies and to train workers in the use of these new techniques. Countries can beneﬁt from this world technology by incorporating it into their production processes. the Nelson-Phelps model also suggests a new role of human capital.451: Introduction to Economic Growth applications.14.1) is that ˙ j (t) /Aj (t) ∂2A > 0. i. diﬀerent from those emphasized by the Mincer equations we used in order to understand the role of human capital in contributing to cross-country income diﬀerences. Imagine that there is a world technology frontier. the greater the human capital of a country is. a country needs highly skilled engineers to adapt world technologies to their conditions. it does have a unique stable stationary distribution as long as φ (hj ) > 0 for all countries.

1 The Basic Krugman Model Consider two sets of economies.2 Trade and Technology Diﬀusion A more subtle and in many ways more useful model of technology transfer is that of Krugman (1979). North and South. and suggests that human capital diﬀerences across countries can be more important in causing income diﬀerences than calculations based on private returns to schooling might suggest. and this stationary cross-country distribution is given by Aj (t) = φ (hj ) T (t) . which is also useful for our purposes because it combines interdependences due to technology transfer with those arising from international trade.2) Suppose now that output in each country is proportional to Aj (t).451: Introduction to Economic Growth stationary state. This eﬀect is in addition to the direct productive contribution of human capital to output. because they will absorb less of the frontier technology. M is the total number of goods that will be determined endogenously.2) then implies that countries with low human capital will be poor. 13.2. Equation (13. All individuals in all countries have the same Dixit-Stiglitz preferences with love for variety given by C= µZ M ε ¶ ε− 1 c (i) ε−1 ε di 0 where c (i) is the consumption of the ith good. g + φ (hj ) (13.14. and ε > 1 is the elasticity of substitution between these goods. all Aj (t)’s will grow at the same rate g. 291 . 13.

both new goods and old goods will command the same price. Their only advantage (and the only diﬀerence in technology) arises because they have access to a larger set of goods. There will now be income diﬀerences arising from “technology diﬀerences” across countries. while the Northern producers specializes in the production of new goods. Workers in the North have access to all goods. It is important to emphasize that when producing old goods. 292 .451: Introduction to Economic Growth There is free international trade between countries. but workers in the South only have access to “old goods”. Another possibility is that the South specializes in the production of old goods. Northern workers have no productive advantage. There can be two types of equilibria. and pN denotes the price of new goods produced in the North. and in this case. In the ﬁrst equilibrium. old goods have been invented in the past and their production technology has been transferred to the South.14. there are suﬃciently few new goods that both workers in the South and the North will produce some of the old goods. Goods fall into two categories: new goods are just invented in the North and can only be produced there. In this case prices and wages will satisfy pS = wS pN = wN > wS where pS is the price of the old goods produced in the South. so they can be produced both in the South and in the North. and incomes in the North and South will be the same (why?). One worker produces one unit of any good to which the country in which he is located has access to.

and this situation is also drawn diagrammatically in the next ﬁgure as the intersection of the relative demand curve for Northern labor with the relative supply curve at LN /LS . we need wN /wS > 1.14. Combining this with (13. Instead the ﬂat portion of the relative demand curve corresponds to the case where there is no full specialization. and hence some of the old goods are produced in the North (and wN /wS = 1). Note that wN /wS > 1 corresponds to an intersection when the relative demand services downward sloping. MN is the total number of new goods (produced in the North) and MS is the total number of old goods. and LS is the total labor force in the South.451: Introduction to Economic Growth When will we be in this full specialization regime? To answer this question.3) we obtain µ ¶− 1 ε wN = wS LN MS LS MN For this type of equilibrium to exist. note that from the ﬁrst-order condition of consumers the relative consumption of new and old goods have to satisfy cN = cS µ pN pS ¶−ε µ wN wS ¶−ε = (13. 293 .3) Full specialization implies that cN = LN LS and cS = MN MS where LN is total labor force in the North.

according to the Poisson process ˙ S = tMN M and recall that M = MS + MN In steady state.451: Introduction to Economic Growth So if there is a suﬃciently large technology gap between the North and South. formalizing an idea due to Vernon on the product cycle across countries.. we need the number of new and old goods to grow at the same rate. 294 .14.e. In particular. Northern wages and incomes will be higher. suppose that new goods are created according to the following Poisson process ˙ = iM M and these goods are imitated by the South slowly. i. What determines the number of new and old goods? Krugman developed a model to analyze this.

In this case. Then. consider the case in which LN t LS i > 1. consider a variation on this model without international trade. As the rate of imitation. there will be no specialization. It is straightforward to check that as i. and the South and the North will have the same level of income when there is international trade 295 .2. Standard arguments give incomes in the North and the South as 0 0 ε−1 wN = M ε−1 and wS = MS 1 1 So relative wages and incomes in steady state will now be µ ¶ 1 0 wN i ε−1 = 1+ 0 wS t The relative income diﬀerences are typically larger now. t. For example. Then M MS t = . the rate of creation of new technologies. 13.14. to illustrate this point. increases wages (and incomes) in the North relative to the South will increase. MN i Relative wages can be obtained as: wN = wS µ LN t LS i ¶− 1 ε In this economy.2 Understanding the Eﬀects of Trade Next. increases the North becomes relatively poor. relative utility and relative incomes per capita are simply proportional to relative wages. the number of goods produced and consumed in each country will diﬀer.451: Introduction to Economic Growth ˙ ˙ S /MS = M/M .

1 The Model Consider a world economy consisting of a continuum of “small” countries with mass 1. In contrast. One example is Acemoglu and Ventura (2002). An additional lesson from this model is that the stability of the world income distribution and ﬁndings of conditional convergence do not necessary rule out endogenous growth (recall that these patterns were used as evidence against endogenous growth models). 13. Here I outline a version of that model. Specialization and the World Income Distribution Perhaps the most major source of interaction between countries is through international trade. technology diﬀerences will typically matter more! (But not always! Why?) 13. and two ﬁnal products that are used for consumption and investment.14. In the absence of trade. There is free trade in intermediate goods and no 296 . There is a continuum of intermediate products indexed by z ∈ [0. without international trade. M ]. who develop a tractable framework for analyzing cross-country income diﬀerences that incorporates international trade. eﬀectively increasing their real incomes. Intuitively. trade between the North and the South enables the Southern consumers to consume goods that they did not have access to. A number of papers investigate how international trade aﬀects the process of economic growth and creates interdependences across countries. the North will be richer.3.3 Trade.451: Introduction to Economic Growth between the North and the South.

5) where pI and pC are the prices of the investment and consumption goods. savings and economic policies. φ)-country (this is the same as the CRRA preferences we have used so far with θ → 1). pI k (13. φj ). For example. ρj . with Z µ (j ) dG (j ) = M. ρ. where I have explicitly introduced the j to emphasize that these refer to country j . where µ is an indicator of how advanced the technology of the country is. The joint distribution of these characteristics is denoted by G(µ. All countries admit a representative consumer with utility function: Z∞ ln c(t) exp (−ρt) dt . but I will drop this notation below and talk of a representative country.4) where c(t) is consumption at date t in the (µ. ˙. income. Since there is no international trade in assets. y . ρ. and w is the wage rate. ρ.451: Introduction to Economic Growth trade in ﬁnal products or assets. ρ is its rate of time preference. and φ is a measure the eﬀect of policies and institutions on the incentives to invest. country j will be deﬁned by its characteristics (µj . k is capital stock. φ)-country. and also total wage income. since population in each country is normalized to 1. plus investment. Countries diﬀer in their technology. 0 (13. must equal to consumption. φ) and is assumed to be time invariant. pI k Specialization is introduced as follows: µ is assumed to be the number of intermediates produced by the (µ. 297 . The budget constraint of the representative consumer is ˙ + pC c = y ≡ rk + w.14. There is no depreciation of capital. pC c. r is the rental rate.

0 τ ⎞ 1− ⎛M ε Z BI (r. KI is capital used in the production of the investment good. I have written the unit cost functions for convenience. This is a convenient way of introducing endogenous growth following Rebelo (1991)–the accumulation equation is linear. (13.7) 0 where p(z ) is the price of the intermediate with index z .6) (13. the investment good would be produced as follows ·ε ⎞ ετ− ⎛M ε Z ε−1 1−τ ⎝ xI (z ) ε dz ⎠ I = φBKI 0 where B is a normalizing constant. p (z )) = w(1−γ )(1−τ ) rγ (1−τ ) ⎣⎝ p(z )1−ε dz ⎠ ⎦ . r. Each country also contains many competitive ﬁrms in the consumption and investment goods sectors with unit cost functions: ⎡⎛ τ ⎤ ⎞ 1− ε M Z ⎢ ⎥ BC (w. p (z )) = φ−1 r1−τ ⎝ p(z )1−ε dz ⎠ . The underlying production functions are quite similar. Labor is only used in the production of consumption goods. There are a number of noteworthy features introduced with these unit cost functions: 1.14.451: Introduction to Economic Growth A higher level of µ corresponds to the ability to produce a larger variety of intermediates. In all countries. For example. so we interpret µ as an indicator of how advanced the technology of the country is. and xI (z ) is the quantity of the z th intermediate good used in the production of the investment good. intermediates are produced by competitive ﬁrms using a technology that requires one unit of capital to produce one unit of any intermediate that belongs to that country. 298 . 2.

as in the twosector extended AK economy discussed above. Assume that ε > 1. 5.9) (13. Kehoe and McGrattan (1997)). The parameter τ is the share of intermediates in production and it will also turn out to be the ratio of exports to income (i. that is. pI (t) pC (t) c (t) and the transversality condition: pI (t) k (t) exp (−ρt) = 0. The parameter ε is the elasticity of substitution among the intermediates and also the price-elasticity of foreign demand for the country’s products.14. r+p ˙I p ˙C − .2 Equilibrium Consumer maximization of (13. to equal the rate of time preference plus the slope of the consumption path.e. The parameter φ corresponds to an inverse measure of the distortions aﬀecting investment (this corresponds to the tax distortions modeled as τ ’s in Jones (1995) and Chari. t→∞ pC (t) c (t) lim (13.. a measure of openness). The inverse of this elasticity is often interpreted as a measure of the degree of specialization. since by 299 . pI pC The only diﬀerence from the familiar version of the Euler equation is that. ruling out immiserizing growth. the country becoming poorer despite accumulating more.3.5) yields the following ﬁrst-order condition r (t) + p ˙I (t) p ˙C (t) c ˙ (t) − =ρ+ . 4. 13.451: Introduction to Economic Growth 3.8) Equation (13. now the rate of return to savings includes the relative change in the price of investment goods compared to consumption goods.4) subject to (13.8) is the standard Euler equation and requires the rate of return to capital.

an individual will receive income tomorrow which will be spent on consumption goods.e.451: Introduction to Economic Growth investing in one unit of investment good today. Choose the ideal price index for intermediates as the numeraire. the optimal rule is found to be to consume a ﬁxed fraction of wealth: ⎛ ⎛ t ⎞ ⎞ Z r (s) + p ˙I (s) ⎠ ⎠ w (v ) exp ⎝ ds dv . pI (s) 0 pC (t) c (t) = ρ ⎝pI (t) k (t) + Z∞ 0 (13. i.e. pI pC Equation (13.12) Since all countries export practically all of their production of intermediates and import the ideal basket of intermediates. M Z 0 1−ε (13.14) . whose price may have changed. φ)-country is equal to: p (t) = r (t) . 300 (13. i.9) is the transversality condition.13) (13. (13.. ρ. Integrating the budget constraint and using the Euler and transversality conditions.11) p(z ) dz = Z µp1−ε dG = 1. The conditions for price to equal marginal cost for the consumption and investment sectors imply: pC = w(1−γ )(1−τ ) rγ (1−τ ) . this choice of numeraire implies that p is also the terms of trade of the country. the price of exports relative to imports.10) Next consider ﬁrm maximization. Thus the term ˙C p ˙I p − is the adjustment for this change in relative prices. The price of any variety of intermediate produced in the (µ. pI = φ−1 r1−τ .14.

1 − (1 − γ ) (1 − τ ) 301 (13.15) implies that when the number of varieties. is always proportional to consumption expenditure. pC c. (13. and exports τ µp1−ε Y.10). this demand is (1 − γ ) (1 − τ ) times consumption expenditure. Equation (13. Intuitively.17) . Market clearing for labor is also straightforward. is larger. and a lower rental rate. because (13. and given the Cobb-Douglas assumption. and we drop market clearing for capital.14. p. Conversely. w. r. a greater µ implies that for a given level of aggregate capital stock. w (13. the optimal consumption rule. µ. since r = p. one of these is redundant. So the market clearing condition for labor is: 1 = (1 − γ ) (1 − τ ) pC c . p.16) implies labor income.451: Introduction to Economic Growth Finally. a given level of income y is associated with better terms of trade. Intuitively. y . where Y ≡ R (13. Labor demand comes only from the consumption goods sector. Trade balance requires y = µp1−ε Y. By Walras’ law.15) ydG is world income. w. can be simpliﬁed to: pC c = ρ pI k. so each will command a higher price in the world market. divided by the wage rate. there will be less capital allocated to each variety of intermediate. and higher rental rate of capital. for a given µ.16) Finally. a greater relative income y/Y translates into lower terms of trade. each country imports a fraction τ of its output. we need to impose market clearing for capital and labor as well as trade balance.

14. k/k ˙ = x∗ . (13. (13.19) and (13.e. the market clearing conditions also imply that for each country: Z rk + w = µr 1−ε (rk + w)dG. ρ.20) For a given cross-section of rental rates. ˙ /Y .451: Introduction to Economic Growth The state of the world economy is described by a distribution of capital stocks. φ)-country Deﬁne the world growth rate as x ≡ Y ˙ = y/y as yR ≡ y/Y . and the relative income of a (µ. the set of equations in (13.5). the set of equations in (13.22) . This distribution of capital stocks can be obtained from the law of motion of the capital stock of each country: ˙ k = φrτ − ρ.19) (1 − γ ) (1 − τ ) · ρ w = . For a given distribution of capital stocks..21) φ =µ ρ + x∗ 302 µ 1 ¶ ε− τ . combined with equilibrium conditions in (13. k (13. In addition. Then. i.20) determine the cross-section of rental rates.18) determine the evolution of the distribution of capital stocks.17)). the steady-state cross-section of rental rates are: ¶1/τ µ ρ + x∗ ∗ r = φ Moreover: ∗ yR (13. rk + w [γ + (1 − γ ) τ ] φrτ + (1 − γ ) (1 − τ ) ρ (13.18) (this law of motion simply follows from the budget constraints of the representative consumer. (13. setting the same growth rate for all countries. It can now be shown that the world economy has a unique and stable steady state in which all countries grow at the same rate.

consider the limiting case where τ = 0). 13. So why is there a stable world income distribution here? The reason is due to changes in relative prices.23). despite the fact that in the absence of international trade.23) Equation (13.22) describes the steady-state world income distribution and states that rich countries are those which are patient (low ρ). The terms of trade and the rental return on capital for each economy is given by (13. when a country accumulates more capital.23) implicitly deﬁnes the steadystate world growth rate.22). it is supplying more of the goods that it produces to the world economy. but have unequal levels of income. The implications of this model and this proposition are described next.21) and the relative position of each country in the world income distribution is given by (13. Equation (13.3 Implications The important implications of this analysis are: 1. terms of trade and rates of return on capital.g.451: Introduction to Economic Growth Z φ µ ρ + x∗ µ 1 ¶ ε− τ dG = 1.14. create incentives to invest (high φ). In the open economy..3. There is a stable world income distribution. This discussion establishes: Proposition 35 In the above-described world economy. there exists a unique steady state equilibrium in which all countries grow at the same rate x∗ deﬁned by (13. 2. (13. each country would grow at diﬀerent rates (e. and have access to better technologies (high µ). experiencing a 303 .

4 Growth with Factor Price Equalization The above model incorporated trade between countries together with terms of trade eﬀects. When τ = 0. which equalizes factor prices). there is no trade in ﬁnancial assets (only in goods. and thus each country takes factor prices as given. distortions and savings rate. An alternative would be to incorporate trade assuming that each country is a small open economy. and they become arbitrarily large as ε → ∞ or as τ = 0. The strength of these eﬀects depend on ε and τ . This implies that consumption growth in all 304 . The second is the closed economy case. This reduces the return to capital and discourages further accumulation.14. where small diﬀerences will translate into inﬁnitely large level diﬀerences (since they imply diﬀerences in growth rates). this means there is factor price equalization. If each country is within the cone of diversiﬁcation. this relative price eﬀect is absent. in which there are no decreasing returns coming from relative price changes. 4. the share of capital in GDP is independent of this. Diﬀerences in saving rates or distortions can have much larger eﬀects than those implied by the standard neoclassical model. The ﬁrst of these is the Heckscher-Ohlin limit.451: Introduction to Economic Growth decline in its terms of trade. In the meantime. Imagine the world rate of return to capital is equal to r∗ . and each country has identical preferences given by our standard CRRA formula. and each country grows at a diﬀerent rate determined by its technology. This is done in Ventura (1997). 13. with standard endogenous growth. 3. determined largely by the share of consumption investment goods in income.

e. the more patient countries will become much richer. ρj .451: Introduction to Economic Growth countries will be given by 1 c ˙j = (r∗ − ρ) . 305 . = cj θ In this case. those with smaller discount rates (greater patience) will ultimately become much richer than the rest. In general. or one country produces almost all of the output of the world economy..14. This process will end either when the world moves out of the cone of diversiﬁcation. more patient countries will have lower initial consumption but higher consumption growth. In fact. we tend to assume that all individuals have the same discount rates in order to ensure a stable income distribution within a country. the more patient country will ultimately become much richer than the rest of the world is more general than the open economy model outlined here. discount rate. as we allowed in the previous model. this feature that with given prices. i. now imagine countries diﬀer according to their patience. and therefore they will accumulate more capital and invest in their own country. In a closed economy with individuals that have diﬀerent discount rates. cj θ However. Then the above equation becomes ¢ c ˙j 1¡ ∗ r − ρj . Ultimately.

14.451: Introduction to Economic Growth 306 .

Part IV Endogenous Technological Change 307 .

.

unless explicitly prohibited. the fact that I am making use of a particular idea does not preclude other people from doing so. and technological change is a consequence of purposeful investments by individuals. As originally noted by Arrow (1962).. or growth took place as a byproduct of knowledge spillovers. created by production. in essence. and as assumed so far in all of the models we studied. as noted by Arrow. Either growth was exogenous. These models will be discussed in the next few chapters. we have investigated models of economic growth of exogenous or endogenous variety. but growth was never a result of the actual process of technological change.14. an externality. Once an idea about how to produce a new good or how to improve the productivity of a certain process is out there. but not a purposeful activity. or it was sustained because of linear technology of accumulation. why would a competitive ﬁrm invest upfront resources to improve the production technology if other ﬁrms will also beneﬁt from this improvement (and it will still end up making zero proﬁts)? Romer’s (1997) model we studied above tried to avoid this problem by making knowledge accumulation endogenous. Before going into details of the speciﬁc models. Moreover. a general principle of this class of models is useful to highlight. a non-excludable and non-rival good.451: Introduction to Economic Growth Until now. Much more attractive are models in which growth is a consequence of technological change. but they make contact with industrial organization models of technology. This observation creates a problem in constructing models of purposeful innovation. Endogenous technological change models are explicitly about making knowledge accu309 . innovation. These models not only allow us to talk about the endogenous rates of technological progress. knowledge is. It was a byproduct. many individuals and ﬁrms will have access to it. R&D policy etc. In fact. making knowledge not only non-excludable but also non-rival. and also enable us to discuss issues of directed technical change. anti-trust.

The monopoly proﬁts the inventor expects will. In particular. Such protection will enable the inventor to become a monopolist. but it will also imply that private and social incentives for innovation will not be typically aligned. a new product or a new production process will be protected under either a patent law or because nobody else will be able to replicate this invention without the speciﬁc know-how of the inventor. in turn. 310 .451: Introduction to Economic Growth mulation endogenous. stimulate research and induce ﬁrms to make the upfront investments to improve productivity and generate growth. They break the paradox pointed out by Arrow by introducing monopolistic competition and patent rights. we will now be looking at models of monopolistic competition. where a ﬁrm that invents a new machine. which also goes back to Schumpeter. will be central to the models that follow. This insight that monopoly rights are important for innovation.14.

and individuals have love-for-variety. it is the variety of machines that expand (because of invention of new varieties). The key is that the R&D is purposeful. so they derive greater utility when they have more goods available. research leads to the invention of new goods.” increasing the productivity of ﬁnal good ﬁrms. and it leads to an output that increases the productivity of existing factors. Two versions of essentially the same model could be used. which is the one I will use here. we will use the Dixit-Stiglitz constant elasticity structure.Chapter 14 Expanding Variety Models The simplest models of endogenous technological change are those in which the variety of inputs used by ﬁrms increases (expands) over time as a result of R&D undertaken by research ﬁrms. and also in the models of quality competition we will see below. and a greater variety of machines leads to greater “division of labor. so “real” income increases. In the second. undertaken for proﬁts. In all of these models. In the ﬁrst. 311 .

14.451: Introduction to Economic Growth

14.1

The Lab-Equipment Model of Growth with Product Varieties

We start with a particular version of the growth model with expanding varieties of inputs and an R&D technology such that only output is used in order to undertake research. This is sometimes referred to as the “lab equipment” model, since all that is required for research is additional investment in more equipment in labs etc.

14.1.1

Demographics, Preferences and Technology

**Imagine an inﬁnite-horizon economy in continuous time admitting a representative household with preferences Z∞
**

0

C (t)1−θ − 1 exp (−ρt) dt. 1−θ

(14.1)

Throughout I suppress time dependence when this causes no confusion. There is no population growth. The unique consumption good of the economy is produced with the following aggregate production function: 1 Y = 1−β ∙Z

N

k(v )

1−β

dv Lβ

0

¸

(14.2)

where L is the aggregate labor input, N denotes the diﬀerent number of varieties of capital inputs, and k (v ) is the total amount of capital (machine) of input type v. The term (1 − β ) in the denominator is included for notational simplicity. Notice that for given N , which ﬁnal good producers take as given, equation (14.2) exhibits constant returns to scale. Therefore, ﬁnal good producers are competitive and subject to constant returns to scale, justifying our use of the aggregate production function to represent their production possibilities set. 312

14.451: Introduction to Economic Growth We simplify the analysis by assuming that the capital inputs are just like intermediate goods and they immediately depreciate after being used (thus it may be easier to think of them as intermediate goods instead of capital, though the machine interpretation may be nice for certain purposes). The budget constraint of the economy is C +I +X ≤Y (14.3)

where I is investment and X is expenditure on R&D, which is for now assumed to come out of the total supply of the ﬁnal good. (Other models of R&D will be discussed below). Assume that the creation of new inputs takes place as follows: ˙ = ηX, N and the economy starts with some initial technology stock N (0) > 0. This implies that greater spending on R&D leads to the invention of new inputs. There is no uncertainty in this process, at least at the aggregate level. One may want to think that there is uncertainty at the individual level, but with many diﬀerent research labs undertaking such expenditure, at the aggregate level, equation (14.4) holds deterministically. The important point is that R&D expenditure expands the potential set of capital/machine varieties. A ﬁrm that invents a new capital variety is the sole supplier of that type of machine, and sets its price χ(v ) to maximize proﬁts. The demand for capital of type v is obtained by maximizing (14.2). Namely, simply considering the aggregate production function, the maximization problem for inputs is: ∙Z N ¸ Z N 1 1−β β k(v) dv L − χ (v) k (v )dv − wL. max [k(v )]lv∈[0,N ] ,L 1 − β 0 0 313 (14.4)

(14.5)

14.451: Introduction to Economic Growth Recall that machines depreciate fully after use, so χ (v) is also the user cost of machines, which is incorporated in the expression above. The ﬁrst-order condition with respect to k (v ) for any v ∈ [0, N ] yields the demand for machines from the ﬁnal good sector. These demands take the convenient isoelastic form: Lβ k(v) = χ(v) ∙ ¸1/β . (14.6)

Assume also that, once the blueprint of a particular input is invented, the research ﬁrm can create one unit of that machine at marginal cost equal to ψ units of the ﬁnal good. Now consider the monopolist owning a machine of type ν invented at time t. This monopolist chooses an investment plan and a sequence of capital stocks so as to maximize the present discounted value of proﬁts starting from time t, as given by ∙ Z s ¸ Z ∞ exp − r (ω ) dω [χ(ν, s)k(ν, s) − ψk(ν, s)] ds V (ν, t) =

t t

(14.7)

where r (t) is the market interest rate at time t. Alternatively, assuming that the value function is diﬀerentiable in time, this could be written as a dynamic programming equation of the form ˙ (ν, t) = χ(ν, t)k (ν, t) − ψk(ν, t). r (t) V (ν, t) − V (14.8)

14.1.2

Digression on Continuous Time Value Functions

To see why (14.8) follows from (14.7), you should think of the principle of optimality again (now in continuous time rather than discrete time). In particular, rewrite (14.7) at time t as: V (ν, t) = Z

∆t

0

∙ Z exp −

∆t

t

Z r (ω) dω (χ(ν, s) − ψ) k(ν, s)ds+ 314

¸

∞

∆t

∙ Z exp −

s

∆t

r (ω) dω [χ(ν, s)k(ν, s) − ψk(ν

¸

14.451: Introduction to Economic Growth which is just an identity for any ∆t. For suﬃciently small ∆t, this can be written as V (ν, t) ' ∆t (χ(ν, t) − ψ) k(ν, t) + exp (r (t) ∆t) V (ν, t + ∆t) 0 ' ∆t · (χ(ν, t) − ψ) k(ν, t) + exp (r (t) · ∆t) V (ν, t + ∆t) − exp (r (t) · 0) V (ν, t), whereexp (r (t) · 0) = 1. Now divide both sides by ∆t and take the limit ∆t → 0, which makes the approximation exact, giving (χ(ν, t) − ψ) k(ν, t) + lim exp (r (t) · ∆t) V (ν, t + ∆t) − exp (r (t) · 0) V (ν, t) = 0. ∆t→0 ∆t

When the value function is diﬀerentiable in time, this is equivalent to ¯ ∂ (exp (r (t) · ∆t) V (ν, t + ∆t)) ¯ ¯ (χ(ν, t) − ψ) k (ν, t) + = 0. ¯ ∂t ∆t=0 thus, applying the chain rule, ˙ (ν, t) = 0, (χ(ν, t) − ψ) k(ν, t) − r (t) V (ν, t) + V which is identical to (14.8).

14.1.3

Characterization of Equilibrium

Since (14.6) deﬁnes isoelastic demands, the solution to the maximization problem of the monopolist involves setting the same price in every period, χ(ν, t) = ψ , 1−β

that is, all monopolists charge a constant rental rate, equal to a mark-up over the marginal cost. Without loss of generality, normalize the marginal cost of machine production to ψ ≡ (1 − β ), so that χ(ν, t) = χ = 1 315

14.451: Introduction to Economic Growth Proﬁt-maximization also implies that each monopolist rents out the same quantity of machines in every period, equal to k (v, t) = L, and makes proﬁts π (v, t) = (χ(ν, t) − ψ) k (v, t) = βL, (14.10) (14.9)

implying that all monopolists sell exactly the same amount, charge the same price and make the same amount of proﬁts. Substituting (14.6) and the machine prices into (14.2), we obtain Y (t) = 1 N (t) L. 1−β (14.11)

This is the major equation of the expanding product or input variety models. It shows that even though the aggregate production function is constant returns to scale from the viewpoint of ﬁnal good ﬁrms which take N as given, for the overall economy, there are increasing returns to scale and increases in the variety of machines, N , increase the productivity of output. In particular, (14.11) makes it clear that if N increases at the constant rate, so will output per capita. Similarly, the labor decision of the ﬁnal good sector, from the ﬁrst-order condition of maximizing (14.5) with respect to L, implies the following equilibrium condition w (t) = β N (t) . 1−β (14.12)

Finally, there is free entry into research. This implies that at all points in time we must have ηV (ν, t) = 1, 316 (14.13)

14.451: Introduction to Economic Growth where V (ν, t) is given by (14.7). Recall that one unit of ﬁnal good spend on R&D leads to the invention of η units of new inputs, each making proﬁts given by (14.7). Naturally, this free entry condition may be violated if research is so unproﬁtable that nobody wants to enter, so it should really be written as a complementary slackness condition with ηV (ν, t) ≤ 1, X (v, t) ≥ 0 and (ηV (ν, t) − 1) X (v, t) = 0, but for the relevant parameter values there will be entry and economic growth (though just technological change), so we simplify the exposition by writing it in the form of (14.13).

14.1.4

Deﬁnition of Equilibrium

**An equilibrium in this economy is described as a sequence of consumption and R&D decisions,
**

∞ [C (t) , X (t)]∞ t=0 such that given the price path [r (t) , w (t)]t=0 , the representative household

is maximizing its utility given by (14.1), capital demands by the ﬁnal goods sector satisfy (14.9), the wage rate is given by (14.12), and the value of each monopolist, V (ν, t), satisﬁes (14.7) and (14.13).

14.1.5

Steady State

Let us start with the steady state. In the steady state, the value of an invention will be ˙ = 0, and also the interest rate will be constant, i.e., r (t) = r∗ (where I constant, thus V again use stars to denote BGP/steady-state values). Substituting this in either (14.7) or (14.8), we obtain V = π r∗ (14.14)

where π is the (constant) ﬂow of net proﬁts per period, given by (14.10) above. 317

14.451: Introduction to Economic Growth For there not to be further incentives to undertake R&D, we need one unit of ﬁnal good spent for R&D to generate exactly the same discounted value. Therefore, the no entry (free entry) condition (14.13) can be expressed as: ηβL =1 r∗ This equation pins down the steady-state interest rate, r∗ , as: r∗ = ηβL From consumer maximization, in particular from the standard Euler equation, we also have that the rate of growth of consumption, gc , is given by

∗ = gc

˙ 1 C = (r∗ − ρ) C θ

(14.15)

**and in steady state, the rate of growth of the economy is the same as the rate of growth of
**

∗ consumption, so we have that the whole economy grows at the rate g ∗ = gc .

Therefore, given the steady-state interest rate we can simply determine the long-run growth rate of the economy as: g∗ = 1 (ηβL − ρ) θ (14.16)

Since this is a growing economy, we need to ensure that the transversality condition is satisﬁed in equilibrium. As usual, this requires r∗ > g∗ (since there is no population growth), i.e., (1 − θ) ηβL < ρ, which we assume holds. Notice that there is a scale eﬀect here: the larger is L, the greater is the growth rate. The scale eﬀect comes from the increasing returns to scale nature of the technology of the model 318 (14.17)

14.451: Introduction to Economic Growth of endogenous technical change (this is a point related to the non-rival nature of knowledge, emphasized in Romer, 1990). I will return to the issue of the scale eﬀect further below. This discussion establishes: Proposition 36 In the above-described expanding input-variety model of endogenous technological change, there exists a unique steady state in which technology, output and consumption all grow at the same rate given by (14.16).

14.1.6

Transitional Dynamics

It is also straightforward to see that there are no transitional dynamics in this model. To see this, let us go back to the value function for each monopolist. Substituting for proﬁts, this gives ˙ (ν, t) = βL. r (t) V (ν, t) − V Free entry gives ηV (ν, t) = 1. ˙ (ν, t) = 0, which is only Diﬀerentiating this with respect to time immediately implies V consistent with r (t) = r∗ for all t, thus r (t) = ηβL for all t. This establishes: Proposition 37 In the above-described expanding input-variety model of endogenous technological change, with initial technology stock N (0) > 0, there is a unique equilibrium path in which technology, output and consumption always grow at the rate g as in (14.16). 319

14.451: Introduction to Economic Growth In other words, exactly as in the AK model, the economy always grows at a constant rate. At some level this is not surprising, since the derived equation for output, (14.11), is essentially a linear AK production function.

14.1.7

Pareto Optimal Allocations

The presence of monopolistic competition implies that the competitive equilibrium is no longer Pareto optimal. There is a version of the aggregate demand externalities we saw in the static context in previous lectures. It is straightforward to set up the problem of the social planner and derive the optimal growth rate. To do this, notice that the social planner will also use the same quantity of all types of machines in production, but because of the absence of a markup, this quantity will be diﬀerent. The social planner will also take into account the eﬀect of an increase in the variety of inputs on the overall productivity in the economy, which monopolists could not because they do not capture the full surplus from inventions. More explicitly, given N , the social planner will choose ∙Z N ¸ Z N 1 1−β β k(v) dv L − ψk (v )dv − wL, max [k(v )]lv∈[0,N ] ,L 1 − β 0 0 which only diﬀers from the private maximization problem because the marginal cost of machine creation, ψ, is used. Recalling that ψ ≡ 1 − β , this implies ks (v) = thus (1 − β )−(1−β )/β Y (t) = N (t) L 1−β = (1 − β )−1/β N (t) L. 320 L (1 − β )1/β ,

Now. µ) = ρµ (t) − µ ˙ (t) = µ (t) η (1 − β )−1/β βL H t→∞ Z∞ 0 C (t)1−θ − 1 exp (−ρt) dt 1−θ The necessary conditions are lim N (t) µ (t) e−ρt = 0. t) dv 0 = (1 − β ) −1/β N (t) L − (1 − β )−(1−β )/β N (t) L = (1 − β )−1/β βN (t) L. Let Y n (t) ≡ Y (t) − I (t) be net output. C. Let us set up the current-value Hamiltonian 1−θ h i −1 ˆ (N.4).451: Introduction to Economic Growth Recall that the aggregate budget constraint is C (t) + I (t) + X (t) ≤ Y (t) . the maximization problem of the social planner can be written as max subject to ˙ (t) = η (1 − β )−1/β βN (t) L − ηC (t) . H 1−θ ˆ C (N. 321 . C. after the costs of machines are subtracted (recall that it is net output that is distributed between R&D expenditure and consumption). given this and (14. We have that Z N (t) −1/β n Y (t) = (1 − β ) N (t) L − ψks (v. C. µ) = C (t) + µ (t) η (1 − β )−1/β βN (t) L − ηC (t) . and C (t) is the control variable. N (t) is the state variable.14. µ) = 0 = C (t)−θ = ηµ (t) H ˆ N (N. N In this problem.

we obtain the following growth rate for consumption in the social planner’s allocation: ´ ˙ C 1³ −1/β = η (1 − β ) βL − ρ . since the monopoly markup reducing the demand for machines is absent in the social planner’s allocation. (14. The most natural alternatives to consider in this model are two: 322 .8 Policy in the Endogenous Technology Model The divergence between the decentralized equilibrium and the socially planned allocation introduces the possibility that there might be Pareto-improving interventions. (14. the decentralized equilibrium is not Pareto optimal.16). C θ The comparison boils down to that of (1 − β )−1/β β to β.18) which can be directly compared to the growth rate in the decentralized equilibrium.451: Introduction to Economic Growth Combining these conditions. because it will be able to use the machines more intensively after innovation. 1). the social planner values innovation more.14. 14. This implies that the socially-planned economy will always grow faster than the decentralized economy. and always grows less than the allocation that would maximize utility of the representative household. Intuitively. This establishes: Proposition 38 In the above-described expanding input-variety model.1. and it is straightforward to see that the former is always greater since (1 − β )−1/β > 1 by virtue of the fact that β ∈ (0.

including taxes on investment income and subsidies of various forms will have growth eﬀects not just level eﬀects in this framework. we also have to think of the objectives of policymakers and this brings us again to political economy issues. it is noteworthy that as in the ﬁrst-generation endogenous growth models. so subsidies to capital inputs given to ﬁnal good producers would also be useful in increasing the growth rate. I will discuss some of the implications of diﬀerent types of competition policies and intellectual property rights policies further below.451: Introduction to Economic Growth 1. Subsidies to Research: by subsidizing research. rather than go into a detailed discussion of optimal policy. once we start thinking of policy in order to close the gap between the decentralized equilibrium in the Pareto optimal allocation. and this can be turned into a Pareto improvement if taxation is not distortionary and there can be appropriate redistribution of resources so that all parties beneﬁt. leaving you to draw your own conclusions about what the implications of this gap will be. For that reason. Subsidies to Capital Inputs: the problem also arises from the fact that the decentralized economy is not using as many units of the machines/capital inputs (because of the monopoly markup). 323 . the government can increase the growth rate of the economy. Naturally. I simply note the gap between the decentralized equilibrium and the Pareto optimal allocation. a variety of diﬀerent policy interventions. 2. Moreover.14.

to the endogenous growth model of Rebelo (1991). instead of the lab-equipment.451: Introduction to Economic Growth 14. there will not be endogenous growth. in equilibrium. In other words. In particular. in some way. In fact. we saw that. growth resulted from the use of ﬁnal output for R&D. The greater is N . now current researchers need to “stand on the shoulder of past giants”. output took a linear form in the stock of knowledge (new machines). unless there are knowledge spillovers from past R&D. there will be competition between the production sector and the R&D sector for workers.2 Growth with Knowledge Spillovers In the model of the previous section. An alternative is to have “scarce factors” used in R&D. The term N on the right-hand side captures spillovers from the stock of existing ideas. As a result. or scientists or regular workers. and the marginal cost of workers and research would be given by the wage rate and production sector.19) where LR is labor allocated to R&D. In the latter case. LR could be skilled workers as in Romer (1990). imposing the standing on the shoulders of giants as part of the technological possibilities frontier of the economy.14. we now have scientists as the key creators of R&D. the more productive is an R&D worker. This is similar. the free entry condition is now ηN (t) V (v. In this case. since the accumulation equation is linear in accumulable factors. the original formulation by Romer (1990) was exactly of this knowledge-spillovers form. thus a AN form instead of the Rebelo’s AK form. t) = w (t) 324 . In other words. A typical formulation in this case is ˙ = ηNLR N (14.

the equilibrium wage rate was derived as (recall equation (14.20). and V (v. This discussion immediately establishes: Proposition 39 In the above-described expanding input-variety model with knowledge spillovers. t) is again given by (14. and we can also compare the decentralized equilibrium to the Pareto optimal allocation.14. r∗ . In particular. there exists a unique balanced growth path equilibrium in which. the growth rate of technology and output are also given by (14. technology.451: Introduction to Economic Growth where N is on the left-hand side because it parameterizes the productivity of an R&D worker from (14.19). It is also useful to note that there is again a scale eﬀect here–greater L increases the interest rate and the growth rate in the economy. C θ (14. output and con325 . becomes ηN (t) β βL N (t) = ∗ r 1−β Hence the steady-state equilibrium interest rate is r∗ = (1 − β ) ηL. Now using the Euler equation of the representative household. there are again no transitional dynamics.20) The rest of the analysis is unchanged. while the ﬂow cost of undertaking research is hiring workers for R&D. we have ∗ gc ≡ ˙ C 1 = ((1 − β ) ηL − ρ) .7) above.12)): w (t) = β N (t) 1−β So the steady-state free-entry condition. Also. thus the wage rate w (t). with a constant steady-state (balanced growth path) interest rate. In the model I outlined in the previous section.

the fringe would prevent the monopolist from setting its ideal monopoly price.2. but they will not be able to produce at the same level of costs (because the inventor has more know-how).14. In particular in this case the monopolist would be forced to set a “limit price”. 1−β Imagine.. we can also relate the results to standard issues in industrial organization. since the monopolist could set its ideal. anti-trust. exactly equal to χ = γψ. 14. patents etc. this fringe is not a threat to the monopolist. the fringe could undercut and make proﬁts.20) starting from any initial level of technology stock N (0) > 0. If it were above this. For example. if γ < 1/ (1 − β ). they will have marginal cost of γψ with γ > 1. instead. such as competition policy. However. proﬁt maximizing. (14. For example. recall that the optimal markup that the monopolist charges is χ= ψ .21) This price formula follows immediately by noting that. in this model we can introduce a fringe of competitive ﬁrms which could limit the markup that each monopolist can charge. If γ > 1/ (1 − β ). if the price of the monopolist were higher than this.1 The Role of Competition Policy Since we now have a model with monopolistic competition. the monopolist could further increase its price without 326 .451: Introduction to Economic Growth sumption grow at the same rate given by (14. suppose that instead of a marginal cost ψ. markup and the fringe would not be able to enter without making losses. that a fringe of competitive ﬁrms can copy the innovation of any monopolist. In particular. since their marginal cost is equal to γψ .

with a lower markup. Another similar application is to that of patent policy. Thus. which is less than β . When the monopolist charges this limit price. also reduces long-run growth. This is because proﬁts are important in this model to encourage innovation by new research ﬁrms. the proﬁts per unit that the monopolist made in the absence of the competitive fringe.451: Introduction to Economic Growth losing any customers to the fringe and make more proﬁts. The exact tradeoﬀ between these two opposing eﬀects depends on the discount rate of the representative household. Of course. this growth rate would be ´ 1 ³ −1/β −(1−β )/β ηγ g = (γ − 1) (1 − β ) L−ρ . its proﬁts per unit would be proﬁts per unit = (γ − 1) ψ = (γ − 1) (1 − β ) .14. we assumed that patents are perpetual. For example. in the baseline model with the lab-equipment technology. In practice. it has a patent forever and it becomes the monopolist for that good 327 .21). incentives for research are also reduced. θ ∗ which is less than (14. If these proﬁts are cut. greater competition. welfare is not the same as growth. and some degree of competition reducing prices below the unconstrained monopolistic level might be useful for welfare depending on the discount rate of the representative household. What is the implication of this on the rate of economic growth? It is straightforward to work out that in this case the economy would grow at a slower rate. households are happier in the present. once a ﬁrm invents a new good. In the baseline model. patents are for limited durations. which reduces markups (and thus static distortions). in this model. Essentially. Therefore. somewhat counter-intuitively. there is a unique equilibrium price given by (14. but suﬀer slower consumption growth.16).

even in the absence of the competitive fringe. For example. growth would become faster and faster over time. Larger countries do not necessarily grow faster (though the larger market of the United States or European economies may have been an advantage during the early phases of the industrialization process). Also.3 Growth without Scale Eﬀects As we have seen. Again there is a tradeoﬀ here between the equilibrium growth rate of the economy and the static level of welfare.g.. 14. This is problematic for three reasons as argued in a series of papers by Chad Jones: 1. the ﬁrm will cease to make proﬁts on its innovation. In this case. restoring the growth rate of the economy to (14. then this might rule out the competitive fringe from competing. translates into a higher interest rate and a higher growth rate. but growing. the models used so far feature a scale eﬀect in the sense that a larger population. But more important than these trade-oﬀs between growth and level is the fact that these models are the most basic models. e. We will see this issue in Problem Set 6. Instead. competitive pressure from other ﬁrms might encourage faster innovation.16).14. so do not feature some of the potential beneﬁts of competition. The population in general is not constant. 2. If patents are enforced strictly. eventually leading to an inﬁnite output in ﬁnite time. these models would not feature a balanced growth path.451: Introduction to Economic Growth forever. If we have constant population growth as in the standard neoclassical growth model. L (t) = exp (nt) L (0). it can easily be shown that growth is maximized by having as long patents as possible. we can imagine that once the patent runs out. violating 328 . L.

3. we see the total amount of resources devoted to R&D increases steadily. (14. but this has not been associated with an increase in the growth rate.451: Introduction to Economic Growth the transversality condition. This good is produced as before. ˙ (t) = ηN (t)φ LR (t) N where φ < 1 and LR is labor allocated to R&D. So labor market clearing requires LE (t) + LR (t) = L. 1−β 329 . in particular. now there are limited knowledge spillovers.14. All agents have the standard CRRA preferences L Z ∞ 0 exp (−ρt) C 1−θ − 1 dt.e. ˙ (t) = nL (t)). more speciﬁcally.. These observations have motivated Jones (1995) to suggest the following modiﬁcation of the baseline model. Population at time t is L (t) and grows at the constant rate n (i. In the data. New goods are produced by allocating workers to the R&D process as in the knowledgespillovers model studied in the previous section.24) (14. 1−θ (14.2) and all the other assumptions are the same as before. with the production function.23) where LE (t) is the level of employment in the production sector.22) where C is consumption deﬁned over the ﬁnal good of the economy. (14. However. The fact that not all workers are in the production sector implies that the aggregate output of the economy (by an argument similar to before) is given by Y (t) = 1 N (t) LE (t) .

and as commented above.451: Introduction to Economic Growth and proﬁts of monopolists from selling their machines is π (t) = βLE (t) . = w (t) = ∗ r 1−β (1 − β ) LE (t) = 1.11).12). so consumption per capita gross at the rate ∗ = gN gc n ∗ . where a constant fraction of workers are allocated to R&D. let us focus on the BGP (steady state). with population growth this would lead to an exploding path. But now there (14. Since in BGP. this implies the total output grows at the rate gN + n.25) From equation (14. we have the following free-entry condition: ηN (t)φ βLE (t) β N (t) . r∗ where the wage is again substituted from (14. leading to inﬁnite utility.26) 330 . gc = 1−φ (14. The case where φ = 1 is the one analyzed in the previous section. we obtain ˙ E (t) ˙ (t) L N + = 0. the model is well behaved when φ < 1. This implies ηN (t)φ−1 Now diﬀerentiating this condition with respect to time. the interest rate and the growth rate are constant.14. However. the fraction of workers allocated to research is constant. L This implies that the BGP growth rate of technology is given by ˙ (t) n N ∗ = . The key assumption for the model is that φ < 1. ≡ gN N (t) 1−φ is population growth. (φ − 1) N (t) LE (t) ˙ E (t) /LE (t) = n. In particular. In this BGP allocation.

thus to maintain sustained growth more resources need to be allocated to R&D. it allocates more and more of the labor force to R&D.26) is determined only by population growth and technology and does not respond to taxes or other policies. and does so in the presence of population growth. The result is summarized in the next proposition: Proposition 40 In the above-described expanding input-variety model with limited knowledge spillovers as given by (14.451: Introduction to Economic Growth Consequently. More interestingly. and output grows at rate gN + n. (14.25). technology and consumption per capita grow at the rate (14. there exists a unique balanced growth path equilibrium in which. only features limited spillovers.23). starting from any initial level of technology stock N (0) > 0.23). this model generates sustained growth in income per capita as well. 331 . The reason for this is that the technology for creating new ideas. the per capita growth rate of the economy given in (14. This type of model is sometimes referred to as semi-endogenous growth. in order to achieve this growth rate. but where economic growth still responds to policies.14. because while there is sustained growth. Some papers in the literature have attempted to develop models of endogenous growth without scale eﬀects. though this normally requires a combination of restrictive assumptions.

14.451: Introduction to Economic Growth 332 .

This is captured in the models of vertical quality competition or quality improvement. In many ways.1 Baseline Model In the model of expanding machine variety. this seems to describe the growth process better. I have normalized it to 1. t)k (v. it replaces previous models. diﬀerent machines were complements in production. t) is the quality of machine v at time t and because now the number of varieties is constant. The major diﬀerence from the previous setup is that the production function is now 1 Y (t) = 1−β ∙Z 1 q(v. 333 .1) where q (v. here it takes place because existing inputs become more productive. while in the previous section growth took place because the variety of inputs expanded. However. such as the models in Aghion and Howitt. in practice when a better computer comes to the market. t) 1−β dv Lβ 0 ¸ (15. or Grossman and Helpman. Consequently.Chapter 15 Models of Quality Competition 15. Population and labor supply are again constant at L.

Without this assumption. incentives to undertake such innovations may diﬀer between the incumbent monopolist and entrants. One issue here. (14. which was absent in the expanding input variety model. as in the baseline endogenous technological change model. To invent a new machine. is whether the existing leader will undertake R&D and innovation. but unless λ is very large. But also. A major insight here comes from Arrow (1962). it will have to charge a limit price in order to exclude the previous leader. this was irrelevant. This is similar to the discussion of the limit price above (which led to equation (14. existing machines can be (and are) improved. When a better vintage of a particular machine is created. If a ﬁrm spends qz units of the ﬁnal good for R&D on a machine of quality q . since machines could not be improved upon. who noted the presence 334 . Here. Notice that the cost of undertaking R&D is proportional to the quality of the machine on which the ﬁrm is working.1). The rest of the setup is the same as before. then it has a ﬂow rate µz of inventing a new machine. it replaces (destroys) the existing vintage. so we will observe limited prices in equilibrium.451: Introduction to Economic Growth and it also has a nice Schumpeterian ﬂavor of creative destruction. assume that the marginal cost of production is ψq for a machine of quality q . the population and labor supply is ﬁxed at L. and this is the source of economic growth. In the expanding input variety model. in contrast. leading to an explosive path. ﬁrms undertake R&D on an existing machine (of type v). with quality λq . R&D would become more and more proﬁtable over time. and who undertook them was not important. The new machine will take over the market for this type of capital. The economy admits a representative household with preferences given by the standard CRRA form. Instead.14. Also. there is no population growth.21) there). This is natural. In particular. so there was only R&D for new machines. I assume that λ is not too large.

t) − x(v. since it creates a real sense of creative destruction or churning. t) = λ−1 Lq(v. 1 Q (t) L 1−β Z 1 q (v. a new entrant does not have this replacement calculation in mind.14. or there is only a limited number of new entrants. As a result. t)dv 0 The value of being the inventor is diﬀerent now. This generates proﬁts π (v. and sells k (v ) = L.1). Following the same analysis as before. with the same technology of innovation.3) Substituting (15. t) − V 335 . This is an attractive implication. t) λ (15. t)V (v. the incumbent would be replacing its own machine. t) = q (v. t)/χ(v. we obtain total output as Y (t) = where Q (t) = is the average total quality of machines.2) Let us normalize ψ = λ−1 . t)]1/β L. Of course in practice we see established big ﬁrms undertake innovation. it will always be the entrants–new ﬁrms– who do R&D in this model. the standard dynamic programming equation now becomes: ˙ (v. More formally. so the monopolist sets the price χ(v. the demand for machines are now k (v. and thus destroying the proﬁts that it is already making. (15. because this position will not last forever. t) = π (v.451: Introduction to Economic Growth of the replacement eﬀect .2) into (15. t) r (t) V (v. t) = [q (v. In contrast. This might be because the technology of innovation diﬀers between incumbents and new potential entrants. t). One of the questions in Problem Set 6 will get you to work through a model along these lines.

x∗ . (15. dropping time and sector dependence and using stars In steady state. Note that there is an immediate relationship between the innovation rate. there will be entry into or exit from research. ˙ (v. So.4) Otherwise.451: Introduction to Economic Growth where x(v. From then on. it receives zero proﬁts. and the last equality follows from free entry condition (15. we have V π∗ π∗ = r∗ + x∗ r∗ + g ∗ /(λ − 1) q (λ − 1)2 L = = µ−1 q. t). and let us denote it by x∗ . and those increase output by a factor λ − 1. t) = 0. λ [(λ − 1)r∗ + g ∗ ] = where the penultimate equality follows from substituting for proﬁts from (15. When this event occurs. t) is the rate at which new innovations occur in sector v at time t. growth occurs because there are new and better machines.4). V again to denote BGP values. This simply follows from the fact that on average. given by: g ∗ = (λ − 1) x∗ .3). the existing monopolist loses its monopoly position and is replaced by the monopolist of the higher-quality machine. g∗ . t) = q(v.14. 336 . since one more unit of the ﬁnal good provides a ﬂow rate µ of obtaining V . t) will be constant across diﬀerent types of goods and over time. In the balanced growth path x(v. Free entry into R&D implies that µV (v. and the BGP growth rate. and thus has zero value.

λ (15. 15. But in fact.5) Proposition 41 In the above-described quality-improvement model. we have that in steady state. the Euler equation (14.15) still applies. However. and does not fully internalize the beneﬁts accruing to ﬁnal good producers (and the economy) from further innovation. r∗ = θg ∗ + ρ. This tends to induce entrants 337 . λ((λ − 1)θ + 1) ∗ 1 g = (θ + 1/ (λ − 1)) This establishes: µ ¶ µ(λ − 1) L−ρ . there exists a unique balanced growth path equilibrium in which output and consumption grow at the same rate given by (15. The reason why there is too little innovation is the same as the model in the previous section: a monopolist does not sell as many units of the new machines as the social planner would like. a new innovation “steals” the proﬁts of the existing monopolist. so combining those. like that of the endogenous technology model with expanding input varieties. generally.2 Pareto Optimality This equilibrium. so µ(λ − 1)2 L =1 λ [(λ − 1) (θg∗ + ρ) + g ∗ ] therefore g∗ = or rearranging. The rate of innovation is g ∗ / (λ − 1). this can be because there is too little or too much innovation. Pareto optimal. is not.5). counteracting this there is the business stealing eﬀect coming from the Schumpeterian nature of the model.451: Introduction to Economic Growth Moreover. ¢ ¡ 1 µ(λ − 1)2 L − (λ − 1)λρ .14.

The analysis of Pareto optimality is straightforward here because of the parallel between the structure of this model to that with expanding input variety. Q 338 . it is immediate to see that a social planner would choose demands for machines as ks (v) = L = λ1/β L. t) ks (v. is equal to Y (t) = λ(1−β )/β Q (t) L. which is deﬁned as Y n (t) ≡ Y (t) − I (t) as in the expanding variety model. under the socially-planned economy. 1−β λ(1−β )/β Q (t) L − λ(1−β )/β Q (t) L (1 − β ) (15. In particular. even when R&D has small social returns.14. 1/β ψ given the assumption that in this case ψ = λ−1 . the social planner faces an aggregate technology frontier of the form ˙ (t) = µ (λ − 1) X (t) . and we have λ(1−β )/β Q (t) L − Y (t) = (1 − β ) n = = λ(1−β )/β β Q (t) L. because it enables them to become the monopoly producers of a new machine. t) dv 0 Finally. This implies that total output. It is once more useful to work in terms of net output. note that given the assumptions above. (1 − β ) Recall again that the aggregate budget constraint is C (t) + I (t) + X (t) ≤ Y (t) .451: Introduction to Economic Growth to do too much R&D. thus becoming the claimant of the natural monopoly power accruing to the leader in a particular line of machines.6) Z 1 ψq (v.

Q (t) is the state variable. 1−θ 1−β The necessary conditions are ˆ C (N.451: Introduction to Economic Growth since an R&D spending of Q (t) X (t) will lead to discoveries of better vintages at the ﬂow rate of µ. given this equation. and C (t) is the control variable. µ) = 0 = C (t)−θ = µ (λ − 1) ξ (t) H ˙ (t) = ξ (t) µ (λ − 1) λ ˆ N (N. 1−β L Combining these conditions. the maximization problem of the social planner can be written as max subject to ˙ (t) = µ (λ − 1) λ Q (1−β )/β Z∞ 0 C (t)1−θ − 1 exp (−ρt) dt 1−θ β 1−β Q (t) L − µ (λ − 1) C (t) . Now. Let us again set up the current-value Hamiltonian " # 1−θ (1−β )/β C ( t ) − 1 β λ ˆ (Q. we obtain the following growth rate for consumption in the social planner’s allocation: ˙ 1 C = C θ Ã ! λ(1−β )/β β µ (λ − 1) L−ρ . µ) = ρξ (t) − ξ H t→∞ (1−β )/β β lim N (t) ξ (t) e−ρt = 0. In this problem. C. C.6). each of these vintages increases average quality of machines by a proportional amount λ − 1. (15. and the budget constraint. C. 1−β 339 (15. µ) = H + ξ (t) µ (λ − 1) Q (t) L − µ (λ − 1) C (t) . where the constraint uses net output.7) .14.

14. we see that g ∗ > g S .451: Introduction to Economic Growth Comparing this to (15. In particular: Proposition 42 In the above-described quality—improvement model. For example. and may grow less or more rapidly than the allocation that would maximize the utility of the representative household. so that there is too much innovation in the decentralized economy relative to the social optimum. This illustrates the contracting inﬂuences of the standard underinvestment and the business stealing eﬀect discussed above. 340 . the decentralized equilibrium is not Pareto optimal. we can see that either could be greater.5). when β is small.

and think of endogenizing technology and technological diﬀerences within this more general framework. reasons for economists’ focus on Cobb-Douglas production function. once we abandon the Cobb-Douglas production function. balanced growth requires all technical change to be laboraugmenting. So it is important to consider the implications of more general production functions. The most important one is that a general production function. and a more generally think about various biases in the nature of technical change. Technical change is often not neutral towards diﬀerent factors of production. There are.Chapter 16 Directed Technical Change The framework analyzed so far assumed technical change to be neutral towards diﬀerent factors. with a nonCobb-Douglas production function. we need to develop a theory of why technical change is purely labor-augmenting. we limited ourselves to the Cobb-Douglas production function. Instead. Therefore. however. does not generate balanced growth. and the elasticity of substitution between diﬀerent factors is often found not to be equal to 1. in most applications. 341 . and in fact. associated with arbitrary technological progress.

who. This reasoning leads to the following major question: What explains these various biases and the direction of technical change? Let us consider a model in which proﬁt incentives determine what type of technologies are developed. is pursued for gain.. evidence suggests that technical change change during the 19th century may have been.” 342 . expected gain varies with expected sales of goods embodying the invention. argued: “invention is largely an economic activity which. When developing technologies complementing a particular factor (say skilled workers) is more proﬁtable. more of these technologies will be developed.451: Introduction to Economic Growth Do we have reason to think that biased technical change is important? The answer appears to be yes–there are many examples of systematic biases in technical change. The market size eﬀect: it is more proﬁtable to develop technologies that have a larger market. There is also a possible acceleration in skill-biased technical change during the past 25 years. For example. like other economic activities. What determines the relative proﬁtability of developing diﬀerent technologies? 1.14. for example.. 2. In contrast. at least in part. The price eﬀect: there will be stronger incentives to develop technologies when the goods produced by these technologies command higher prices. Whether the development of these technologies makes aggregate technology more skill-biased or not will depend on the elasticity of substitution between this factor and the rest. the consensus among labor and macroeconomists is that technical change throughout the 20th century has been skill-biased. skill-replacing.. Jacob Schmookler (1966). The importance of market size in innovation was much emphasized by the famous scholar of innovation.

1. 16. i. when σ < 1. take the standard the constant elasticity of substitution (CES) production function i σ h σ −1 σ−1 σ −1 σ σ + (1 − γ ) (AZ Z ) .1) This implies that when σ > 1. L (16.1 Basics and Deﬁnitions Deﬁnitions First consider what factor-augmenting and factor-biased technical change correspond to. In contrast. i.e. Assume that preferences are again given by the CRRA function Z∞ 0 C 1−θ − 1 exp (−ρt) dt. y = γ (AL L) where L is labor. AL is labor-augmenting (labor-complementary) and AZ is Z-complementary. Here σ ∈ (0.1. For this purpose.2) .2 Basic Model Now we are in a position to consider a simple model of directed technical change. when the two factors are gross complements. AZ is labor-biased and AL is Z-biased. AL is laborbiased and AZ is Z-biased.e. when the two factors are gross substitutes. and Z denotes another factor of production.. The relative marginal product of the two factors: MPZ 1−γ = MPL γ µ AZ AL 1 µ 1 ¶ σ− ¶− σ σ Z ..451: Introduction to Economic Growth 16.14. which could be capital or skilled labor. 1−θ 343 (16.1 16. ∞) is the elasticity of substitution between the two factors.

t) or χZ (j. YL and YZ . Here Y can either be interpreted as the ﬁnal good aggregated from the two intermediates. Assume that machines to both sectors are supplied by “technology monopolists”. X . t) dj Lβ . t) for the machine it supplies to the market. with elasticity of substitution ε.4) xZ (j. but exactly the same results apply with the knowledge-spillovers model. YL and YZ . YL (t) = 1−β 0 and 1 YZ (t) = 1−β µZ NZ (16. Intermediate good production functions are: µZ NL ¶ 1 1−β xL (j. and spending on R&D. These prices are potentially time-varying.14. spending on machines. This is a straightforward generalization of the endogenous technical change model of product variety discussed above. I . t) 1−β dj Z β . The fact that there is R&D spending signiﬁes that I will use the lab-equipment model to expose the basic ideas.3) In words. but we will see that they will be constant in equilibrium. Each monopolist sets a rental price χL (j. or Y could be an index of utility deﬁned over the two ﬁnal goods. Total output is again distributed between consumption. though we use the index j to denote either for notational simplicity).451: Introduction to Economic Growth The budget constraint: ε h ε−1 ε−1 i ε−1 C + I + X ≤ Y ≡ γYL ε + (1 − γ )YZ ε (16. 344 ¶ (16.5) . YL and YZ . the output aggregate is produced from two other (intermediate) goods. 0 Note here that the range of machines used with the two sectors are diﬀerent (there are two disjoint sets of machines. C .

(t)) Similarly pZ (t) xZ (j.4) and (16.7) ¸1/β Z.{xL (j.7).14. t) dj.451: Introduction to Economic Growth The marginal cost of production is the same for all machines and normalized to ψ ≡ 1 − β in terms of the ﬁnal good. (16. Substituting these into (16. all machine prices will be given by χL (j. t) = 1 for all j and t. t) = χZ (j. t) = χZ (j. In particular.8) Since the demand curve for machines facing the monopolist. and xZ (j. is iso-elastic.5). t) = [pZ (t)]1/β Z for all j . (16. we obtain YL (t) = and YZ (t) = 1−β 1 [pZ (t)] β NZ (t) Z 1−β 1−β 1 [pL (t)] β NL (t) L 1−β 345 . t) = χL (j. These imply that xL (j. (16. t) = [pL (t)]1/β L for all j . t) ∙ ∙ ¸1/β L. (16. t) xL (j.t)} 0 This gives machine demands as pL (t) xL (j. Price taking implies the following maximization problem for sector L ﬁrms at time t: Z NL max pL (t) YL (t) − wL (t) L − χL (j.6) L. the proﬁt-maximizing price will be a constant markup over marginal cost.

This highlights the two eﬀects on the direction of technical change that I mentioned above: 1. 2.3) that the relative price of good Z to good L will be given by µ ¶− 1 ε pZ 1 − γ YZ p ≡ = pL γ YL µ ¶− 1 ε 1−β NZ Z 1−γ p β = γ NL L 346 .451: Introduction to Economic Growth Proﬁts of technology monopolists at time t are then obtained as π L (t) = β [pL (t)]1/β L and π Z (t) = β [pZ (t)]1/β Z . Then in steady state. r r 1/β 1/β (16.9) Let VZ and VL be the net present discounted values of new innovations. rather than NL . The greater is VZ relative to VL . The price eﬀect: a greater incentive to invent technologies producing more expensive goods.10) The comparison of these two values is of crucial importance. the greater are the incentives to develop Z -complementary machines. The market size eﬀect encourages innovation for the more abundant factor.14. (16. The market size eﬀect: a larger market for the technology leads to more innovation. NZ . we have that (dropping time dependence): βp L βp Z VL = L and VZ = Z . It is straightforward from the ﬁnal good production function given in (16.

which will be. An increase in the relative factor supply.11) . L (16. in part. will increase VZ /VL as long as σ > 1 and it will reduce it if σ < 1. Suppose as in the analysis above that new machines in the two sectors are produced by investing in lab equipment: ˙ Z = η Z XZ .12) µ 1−γ γ ε µ ¶σ NZ NL 1 µ ¶− σ ¶ σ−1 Z σ . 347 (16. is the (derived) elasticity of substitution between the two factors. the elasticity of substitution regulates whether the price eﬀect dominates the market size eﬀect. Note also that we have σ ≥ 1 ⇐⇒ ε ≥ 1 So the two factors will be gross substitutes when the two goods in utility function (or the two intermediates in the production of the ﬁnal good) are gross substitutes.14. We have so far characterized the demand for new technologies.451: Introduction to Economic Growth Substituting for relative prices into the steady state (BGP) value functions. This gives the following steady-state “technology market clearing” condition: η L VL = η Z VZ. Next we have to determine the “supply” of all new technologies. regulated by the technological possibilities for generating new machine varieties. relative profitability is obtained as: VZ = VL where σ ≡ ε − (ε − 1) (1 − β ) . ˙ L = η L XL and N N where X denotes R&D expenditure. Z/L.13) (16. Therefore.

The study this issue. using the same type of analysis as before. ∗ NL γ L where the *’s denote that this expression refers to the steady-state value Before going further. 348 . Second.14) Starting from any NL (0) > 0 and NZ (0) > 0. the economy converges to this balanced growth NZ NL First. Because there are two state variables now. Z/L..15) which determines whether innovation for more abundant factors is more proﬁtable also determines whether a greater NZ /NL –i. the same combination of parameters. recall that relative factor prices are given by NZ wZ = p1/β = wL NL µ 1−γ γ ε µ ¶σ (16. the steady-state relative physical productivities can be solved for µ ¶ε µ ¶σ−1 ∗ NZ 1−γ Z σ =η .14). the economy features transitional dynamics.451: Introduction to Economic Growth Then. L (16. we can characterize the equilibrium in this economy. the relative factor reward. and consumption and output grow at the rate ³ £ ´ 1 ε σ −1 σ−1 ¤ σ−1 −1 ε g=θ β (1 − γ ) (η Z Z ) + γ (η L L) −ρ . there exists a unique balanced growth path equilibrium in which the relative technologies are given by (16. σ−1 . These are stated in the next proposition (and left for you to prove): Proposition 43 In the directed technical change model described here. but still has a unique balanced growth path.e. wZ /wL . More interesting than the aggregate growth rate of the economy in this case is how the direction of technical change aﬀects relative factor prices and how it responds to changes in relative supplies. is decreasing in the relative factor supply. σ 1 µ 1 ¶ σ− ¶− σ σ Z .14. path. a greater relative physical productivity of factor Z – increases wZ /wL .

This discussion.15) as implied by Proposition 44.. we see that the response of relative factor rewards to changes in relative supply is always more elastic in (16. This implies that irrespective of whether σ is greater than or less than one. which states that demand curves become more elastic when other factors adjust. an increase in Z/L will change NZ /NL in a direction that increases the relative reward to factor Z . in particular if σ > 2. Let us refer to a situation in which an increase in the relative supply of a factor changes technology so much that the relative price of the factor becoming more 349 . there is always weak endogenous (relative) bias.451: Introduction to Economic Growth When σ > 1. but with a new interpretation–that is. the relative demand curves become ﬂatter when “technology” adjusts. Relative factor rewards are wZ = η σ−1 wL µ 1−γ γ ¶ε µ ¶σ−2 Z .16) Comparing this equation to the relative demand for a given technology. meaning that an increase in Z/L always causes relatively Z -biased technical change. but when σ < 1. establishes: Proposition 44 In the above-described directed technical change model. together with the deﬁnition of weak endogenous bias. it has the opposite eﬀect. let us deﬁne weak endogenous (relative) bias as the phenomenon that an increase in the relative supply of a factor changes technology in a direction that beneﬁts the factor that is becoming more abundant.16) than in (16. i. To capture this notion. the relationship between relative factor supplies and relative factor rewards can be upward sloping. This is simply an application of the LeChatelier principle. wZ /wL . L (16. greater NZ /NL increases wZ /wL . The more important and surprising result here is that if σ is suﬃciently large.14.e.

and in particular of Propositions 44 and 45.3 Implications Let us now consider the implications of this simple model of directed technical change. the skill (college) premium has increased very sharply throughout the 1980s and 1990s. imagine that Z = H stands for skilled workers. In the United States labor market. following a brief period of decline during the 1970s in the face of the very large increase in the supply of college-educated workers. for example.451: Introduction to Economic Growth abundant increases as strong endogenous (relative) bias.1. to reach a level not experienced in the postwar era. college-educated workers. The following ﬁgure shows the general patterns by plotting the college premium and the relative supply of college graduate workers in the United States since WWII. the skill premium has shown no tendency to decline despite a very large increase in the supply of college educated workers. there is strong endogenous (relative) bias in the sense that an increase in Z/L raises the relative marginal product and the relative wage of the Z factor compared to the L factor. if σ > 2. 16. For this application. One of the most interesting applications is to changes in the skill premium. the analysis so far has established: Proposition 45 In the above-described directed technical change model.14. 350 . On the contrary. Therefore.

6 Rel.4 .2 . supply of college skills . in summary. or more generally throughout the entire 20th century? This question becomes more relevant once we remember that during the 19th century many of the technologies that were fueling economic growth. For example.451: Introduction to Economic Growth College wage premium . But why should the economy adopt and develop more skill-biased technologies throughout the past 20 years. Thus.4 .3 39 49 59 69 year 79 89 96 0 Relative Supply of College Skills and College Premium In the labor and macro literature. such as the factory system and the major spinning and weaving innovations.6 . Secular skill-biased technical change increasing the demand for skills throughout 20th century.5 . the computers or the a new IT technologies are argued to favor skilled workers relative to unskilled workers. 351 . supply of college skills Rel. the most popular explanation for these patterns is skillbiased technological change.8 College wage premium . were skill-replacing rather than skill-complementary. we have the following stylized facts: 1.14.

in particular. we have a natural explanation for all of the patterns mentioned above. 2. 1. In addition. Acceleration in the increase in the number of skilled workers over the past 25 years is predicted to induce an acceleration in skill-biased technical change. gives us a way to think about these issues. Possible acceleration in skill-biased technical change over the past 25 years. After a while 352 . In this case. or simply with the degree of skilled bias endogenized. With an upward sloping relative demand curve. It is reasonable to presume that the equilibrium skill bias of technologies. 3. Recall that if σ > 2. is a sluggish variable determined by the slow buildup and development of new technologies. The increase in the number of skilled workers that has taken place throughout 20th century is predicted to cause steady skill-biased technical change. Many skill-replacing technologies during the 19th century.14. Theorems 44 and 45. Large increase in the number of unskilled workers available to be employed in the factories during the 19th century could be expected to induce skill-replacing/laborbiased technical change. 3. The current model. this framework with endogenous technology also gives a nice interpretation for the dynamics of the college premium during the 1970s and 1980s.451: Introduction to Economic Growth 2. then the long-run relationship between the relative supply of skills and the skill premium is positive. a rapid increase in the supply of skills would ﬁrst reduce the skill premium as the economy would be moving along a constant technology (constant NH /NL ) curve in the ﬁgure. NH /NL .

and relates both to the large increase in the supply of skilled workers. To explain the larger increase in the 353 .14. This approach can therefore explain both the decline in the college premium during the 1970s and the subsequent large surge. though again it will be shallower than the short-run relative demand curve. and as technology starts adjusting the skill premium will increase. and the economy would move back to the upward sloping relative demand curve. the long-run relative demand curve will be downward sloping. with a very sharp increase in the college premium. But it will end up below its initial level.451: Introduction to Economic Growth the technology would start adjusting. Relative Wage Long-run Rel Wage Initial Rel Wage Short-run Response Exogenous Shift in Relative Supply Long-run relative demand for skills If on the other hand we have σ < 2. Then following the increase in the relative supply of skills there will be an initial decline in the skill premium (college premium).

The following proposition generalizes these results: 354 .2 Equilibrium Technology Bias: Some More General Results The above model derived the relative bias results by assuming a constant elasticity of substitution production function. the spirit of the results are much more general. In fact. in this case we need some exogenous skill-biased technical change.451: Introduction to Economic Growth 1980s. Relative Wage Initial Rel Wage Long-run Rel Wage Short-run Response Exogenous Shift in Relative Supply Long-run relative demand for skills 16.14. The next ﬁgure draws this case.

there are typically nonconvexities in the creation of new technologies as well. and that the costs of producing technologies AZ and AL . as in the standard “building on 355 . pose that factor supplies are given by Z. ¯ A∗ .18) so that there is always weak relative equilibrium bias. The proof is provided in Acemoglu (2005). AL ) ∈ R2 + . A∗ d ln wZ Z. the condition for strong equilibrium bias is σ > 2+ δ . σ−2−δ Z L Z L = . Moreover. d ln (Z/L) 1 + σδ so that there is strong relative equilibrium bias if σ − 2 − δ > 0. A∗ ∂ ln (A∗ /A∗ ) ∂ ln wZ Z. L) ∈ R2 + . Then we have that and equilibrium factor prices by wZ Z.14. In models with knowledge spillovers. concave and homothetic in its two arguments. supZ and L deﬁned by σ = − ∂ ∂ ln(wZ /wL ) ¯ AZ ∂ ln(AZ /AL ) AL ¡ ¢ ¯ L ¯ and denote equilibrium technologies by (A∗ . Assume that F is twice continuously diﬀerentiable. therefore. A∗ /wL Z. AL )) ln(Z/L) ¯ . Z L ¡ ¢ ¡ ¢ ¯ L. Proof. thus more restrictive than the previous model. ¯ A∗ . AL L). ¡ ¡ ¢ ¡ ¢¢ ¯ L.17) ∂ ln (Z/L) 1 + σδ and ¢ ¡ ¢¢ ¡ ¡ ¯ L. A∗ and wL Z. ¯ A∗ . ¯ L. (16. A∗ /wL Z. AL ) by CZ and CL . such that the production function is F (AZ Z. ¯ A∗ . Finally. ¯ A∗ . ∗ σ−1 ∂ ln (A∗ Z /AL ) = (16. Let σ be the (local) elasticity of substitution between ¯ ∂ ln(CZ (AZ .451: Introduction to Economic Growth Proposition 46 Consider the above economy with two-factors. Denote the ﬁrst derivatives of C (AZ . invention of skill-biased technologies today may make further invention of skill-biased technologies easier. and two factor-augmenting technologies. ¯ L.19) In this environment. This is because costs of creating new technologies are convex. Z L Z L ¡ ¢ ¯ L ¯ : for all Z. ¯ A∗ . (Z. (AZ . is also twice continuously diﬀerentiable. C (AZ . and let δ = . For example. ¯ L. strictly convex and homothetic in AZ and AL . A∗ . AL ). A∗ ).AL )/CL (AZ. Z L Z L Z L ≥0 ∂ ln (AZ /AL ) ∂ ln (Z/L) (16.

Here I outline a model which generates this results (though under somewhat more restrictive assumptions than the directed technical change results we have seen so far). 16. and S “scientists” who perform R&D.3.20) where C (t) is consumption at the time t and θ ≥ 0 is the elasticity of marginal utility. strong equilibrium bias requires suﬃcient substitutability between factors.1 Demographics. The distinction between unskilled workers and scientists is adopted to ensure that the production and R&D sectors do not compete for workers.. with the exact threshold depending on the structure of costs (or the technology possibilities frontier of the economy). Acemoglu (2002) shows that the condition for an upward-sloping relative demand curve (i.14. Thus in general. The economy again admits a representative consumer with the usual constant relative risk aversion (CRRA) preferences: Z∞ 0 C (t)1−θ − 1 exp (−ρt) dt 1−θ (16. Preferences and Technology Consider an economy consisting of L unskilled workers who work in the production sector. 16.451: Introduction to Economic Growth the shoulders of giants” speciﬁcation. strong relative equilibrium bias) is in fact σ > 2 − δ 0 .3 Endogenous Labor-Augmenting Technological Change One of the advantages of the models of directed technical change is that they allow us to investigate why technological change might be purely labor-augmenting as required for balanced growth. 356 .e. In that case. for some other parameter δ 0 > 0 measuring the extent of this nonconvexity.

We will see below that ε will also determine the short-run elasticity of substitution between capital and labor. r is the interest rate.21) where I denotes investment. The resource constraint of the economy implies that ε h ε−1 ε−1 i ε−1 ε ε .14.23) Let us also use this opportunity to develop a variant of the models studied above. w is the wage rate of labor. In particular. so the change in the capital stock (and in the representative consumer’s asset level) is given by ˙ = I.451: Introduction to Economic Growth The budget constraint of the representative consumer requires that consumption and investment expenditures are less than total income: C + I ≤ wL + rK + ω S S + Π. K (16. which in the context of this model implies that ε < 1. (16. For simplicity. let us assume that the labor-intensive and capital-intensive goods are produced competitively from constant elasticity of substitution (CES) production functions of laborintensive and capital-intensive intermediates.24) 357 . with elasticity of substitution ε. A host of evidence suggests that this short-run elasticity between capital and labor is less than one. and Π is total proﬁt income. with elasticity ν ≡ 1/(1 − β ): YL = ∙Z n 0 ¸1/β ∙Z yl (i) di and YK = β m yk (i) di β 0 ¸1/β . ωS is the wage rate for scientists.22) where Y is an output aggregate produced from a labor-intensive and a capital-intensive good. where 0 ≤ ε < ∞. K denotes the capital stock. wL + rK + ω S S + Π = Y = γYL + (1 − γ )YK (16. (16. respectively YL and YK . let us assume that there is no depreciation of capital.

R&D ﬁrms have access to the following technologies for invention: n ˙ m ˙ = bl φ (Sl ) Sl − δ and = bk φ (Sk ) Sk − δ. An increase in n–an expansion in the set of labor-intensive intermediates–corresponds to labor-augmenting technical change. Once an R&D ﬁrm invents a new intermediate. employed by R&D ﬁrms. n m 358 (16.26) 0 To close the model. and m that are produced using only capital. while an increase in m corresponds to capital-augmenting technical change. This formulation implies that there are two different sets of intermediate goods. it receives a perfectly enforced patent and becomes the perpetual monopolist of that intermediate. 1). Intermediate goods are supplied by monopolists who hold the relevant patent. we need to specify the innovation possibilities frontier–that is. Market clearing for labor and capital then requires: Z n l (i) di = L and 0 Z m k (i) di = K.14. and are produced linearly from their respective factors: yl (i) = l(i) and yk (i) = k(i).27) .451: Introduction to Economic Growth where y (i)’s denote the intermediate goods and β ∈ (0. who are. so that ν > 1 and diﬀerent intermediate goods are gross substitutes. n of those that are produced with labor. (16.25) where l(i) and k(i) are labor and capital used in the production of good i. (16. There is free-entry into the R&D sector. Let us assume that these blueprints are created by the R&D eﬀorts of scientists. the technological possibilities for transforming resources into blueprints for new varieties of capital-intensive and labor-intensive intermediates. in turn.

The fact that φ (·) is decreasing means that there are intra-temporal decreasing returns to R&D eﬀort. Sl and Sk denote. This feature will enable the analysis of whether equilibrium technical change will be labor. the behavior of Sl and Sk is discontinuous. bk and δ are strictly positive constants and φ (·) is a continuously diﬀerentiable and decreasing function such that φ (s) s is always increasing.or capitalaugmenting. the number of scientists working to discover new labor-intensive and capital-intensive intermediates. 2. while the same eﬀort devoted to the discovery of capital-using intermediates leads to a proportional increase at the rate bk . This might be. I also assume that the economy starts at t = 0 with n (0) > 0 and m (0) > 0. and φ(0) < ∞. for example.27) implies a number of important features: 1.28) . 3.14. φ (Sl ) Sl . Equation (16. Technical change is directed. in the sense that the society (researchers) can generate faster improvements in one type of intermediates than the other. respectively. with the market clearing condition Sl + Sk = S. when more scientists are allocated to the invention of labor-intensive intermediates. leads to a proportional increase in the supply of these intermediates at the rate bl . the productivity of each declines. This decreasing returns assumption is adopted to simplify the analysis of transitional dynamics–when φ (·) is constant. Research eﬀort devoted to the invention of labor-intensive intermediates. The parameters bl and bk potentially diﬀer since the discovery of one type of new intermediate may be “technically” more diﬃcult than 359 (16.451: Introduction to Economic Growth where bl . because scientists crowd each other out in competing for the invention of similar intermediates.

so each R&D ﬁrm takes the productivity of allocating one more scientist to each of the two sectors.27) scientists are “standing on the shoulders of giants” as in the model of knowledge spillovers analyzed above. the results are similar. ∗ Finally. its stock declines exponentially.. 4.e. a higher n increases the productivity of scientists working in the n-sector). so that when there is no research eﬀort devoted to a particular type of intermediates. bl φ (Sl ) or bk φ (Sk ). This. where we had a special form of n/n ˙ = bl φ (S ) S . bl φ (Sl∗ ) Sl∗ = δ and bk φ (Sk 360 . is the crucial assumption that enables the model to generate endogenous technological change that is purely labor augmenting.27) is that a higher stock of knowledge accumulated in one sector beneﬁts only that sector (i.27) is a direct generalization of the accumulation equation in the one-sector knowledge spillovers model analyzed above.451: Introduction to Economic Growth discovering the other type (the standard model with only labor-augmenting technical change can be thought as the special case with bk = 0). Each intermediate disappears at the rate δ . but there will exist multiple balanced growth paths (see below). equation (16. there is the issue of how innovations in one sector aﬀect the knowledge base of the other sector. i. internalizing these crowding-out eﬀects. An additional assumption implicit in (16. when we go to an economy with two sectors.e. deﬁne Sl∗ and Sk as the number of scientists required to keep the state of tech∗ ∗ ) Sk = δ . I also assume that the crowding eﬀect captured by the function φ (·) is not internalized by individual R&D ﬁrms..14. However. In fact. With δ = 0. as given when deciding which sector to enter. Notice that in (16. as we will see. The results are identical when R&D ﬁrms act “non-competitively” and form global research consortiums. Let us impose: nology in each sector constant.

This assumption implies that there is enough scientists in the society to enable technological progress in both sectors. ω S . [k (i)]i=0 . The consumption sequence [C (t)]∞ 0 also satisﬁes the lifetime budget constraint of the representative agent (the no Ponzi game constraint): ∙ Z t ¸ r (v) dv = 0.451: Introduction to Economic Growth ∗ Assumption 14 Sl∗ + Sk < S. and [l(i)]n i=0 .29) where recall that r is the rate of interest. Sl and Sk . which satisﬁes the familiar Euler equation: ˙ 1 C = (r − ρ).30) Consumer maximization gives the relative price of the capital-intensive good as: 1−γ pK = p≡ pL γ 361 µ YK YL ¶− 1 ε . m w. and all markets clear. [yl (i)]i=0 . m n m [l(i)]n i=0 . consumption and saving decisions. pL and pK . such that [yl (i)]n i=0 . C and I . r.3. [yk (i)]i=0 .14.31) .2 Consumer and Firm Decisions An equilibrium in this economy is given by time paths of factor. [yk (i)]i=0 . intermediate and good prices. Sl and Sk imply zero-proﬁts for all R&D ﬁrms. C and I maximize the utility of the m representative consumer given factor. m [pl (i)]n i=0 and [pk (i)]i=0 maximize proﬁts of intermediate goods monopolists. [pl (i)]n i=0 . and the allocation of scientists between the m two sectors. [pk (i)]i=0 . lim K (t) exp − 0 t→∞ (16. C θ (16. employment. I start with the optimal consumption path of the representative consumer. [k (i)]i=0 . 16. (16. intermediate and good prices.

35) into (16. equation (16.e. To determine the level of prices.33) Given these isoelastic demands.32) (16.14. I choose the price of the consumption aggregate.36) These equations reiterate that n and m correspond to labor. n m (16. i.34) Since.34). we obtain yl (i) = l(i) = L K and yk (i) = k(i) = .24) yield the following isoelastic demand curves for intermediates: ¶− 1 ¶− 1 µ µ pl (i) yl (i) ν pk (i) yk (i) ν = and = . proﬁt maximization by the monopolists implies that prices will be set as a constant markup over marginal cost (which is w for the labor-intensive intermediates and r for the capital-intensive intermediates): µ µ ¶−1 ¶−1 1 w 1 r pl (i) = 1 − w= r= . for all i. pL YL pK YK £ £ ¤ 1 ¤ 1 pK = γ ε pε−1 + (1 − γ )ε ε−1 and pL = γ ε + (1 − γ )ε p1−ε ε−1 . 362 .and capital-intensive goods as: YL = n 1−β β L and YK = m 1−β β K.and capital-augmenting technologies. in each period as numeraire. (16.451: Introduction to Economic Growth where pK is the price of YK and pL is the price of YL . Then from the market clearing equation (16. which implies that: γ pL + (1 − γ )ε p1 K Next..35) Substituting (16. Y .26). ¤ 1 £ ε 1−ε −ε 1−ε = 1.24) and integrating gives the total supply of labor.33) implies that yl (i) = yl . yk (i) = k for all i as well. and pk (i) = 1 − ν β ν β (16. all labor-intensive intermediates sell at the same price. and since all capital-intensive intermediates also sell at the same price. and similarly an increase in m raises the productivity of capital. from (16. Greater n enables the production of a greater level of YL for a given quantity of labor. (16. consumer maximization and the CES functions in (16.

Therefore. the relative price of the capital intensive good is ∙ ¸− 1 1−β 1 − γ ³m´ β K ε pK = . and for capital-intensive intermediates. using (16. (16.39).41) Equation (16.36). and competition between the two sectors and free-entry ensure that this wage is equal to the maximum of their contribution to the value of monopolists in the two sectors. so Π = 0 in (16. is: Vf (t) = Z ∞ (16. where Vl and Vk are given by (16.40) are the ﬂow proﬁts from the sale of labor.31) and (16.38) t ∙ Z s ¸ exp − (r(ω ) + δ ) dω π f (v )dv. δ is the depreciation (obsolescence) rate of existing intermediates. bk φ (Sk ) mVk } .37) Finally. (16. t (16. free-entry requires: ωS = max {bl φ (Sl ) nVl . Scientists are paid a wage ω S . so the marginal value of allocating one more scientist to the invention of labor-intensive intermediates is bl φ (Sl ) nVl .33).34). for f = l or k. and πl = 1 − β wL 1 − β rK and π k = β n β m (16. Recall that R&D ﬁrms do not internalize the crowding eﬀects. (16.41) implies zero expected proﬁts for all ﬁrms at all point in time.451: Introduction to Economic Growth Equations (16.21). (16. it is bk φ (Sk ) mVk .36) give the wage rate and the rental rate of capital as: w = βn 1−β β 1−β β pL and r = βm pK .39) where r(t) is the interest rate at date t. 363 .and capital-intensive intermediate goods.14.35) and (16. p≡ pL γ n L The value of a monopolist who invents a new f -intermediate.

14. consumption and the capital stock grow at the same ﬁnite constant rate.. To facilitate the analysis.35). good prices.451: Introduction to Economic Growth An equilibrium in this economy is therefore a set of factor prices.3. ˙ (t) /C (t) = limt→∞ Y ˙ (t) /Y (t) = limt→∞ K ˙ (t) /K (t) = g . A balanced growth path (BGP) is deﬁned as an AP where output. w. sequences of aggregate consumption and investment levels that satisfy (16.. possibly 0 (including the case where limt→∞ C (t) = 0 as a special case). there may exist asymptotic paths where consumption grows more than exponentially or grows at a diﬀerent rate than capital.. together with (16. with ε ≥ 1.36). 16.3 Asymptotic and Balanced Growth Paths Let us deﬁne an asymptotic path (AP) as an equilibrium path that the economy tends to as ˙ (t) /C (t) = t → ∞. that satisfy (16.29) and (16.30). this has to be a BGP. and sequences of Sl and Sk that satisfy (16. In an AP. output levels given by (16.e.e. but these artists interesting for us given our focus on ε < 1. so if the economy is going to tend to a non-cycling path. consumption grows more than exponentially (explodes). and does not include limit cycles.41). i.34). the rate of consumption growth tends to a constant. [pk (i)]i=0 . allows me to write output in a more compact way: h i ε ε−1 ε−1 ε−1 ε ε Y = γ (NL) + (1 − γ ) (MK ) 364 (16.36). i. or limt→∞ C i. r and ω S that m satisfy (16. [pl (i)]n i=0 . limt→∞ C This subsection will show that with ε < 1. we can have either limt→∞ C ˙ (t) /C (t) = gc . let us a dope the notation: N ≡n 1−β β and M ≡ m 1−β β .41). intermediate production levels given by (16.42) . only BGPs can be an AP. In contrast.e. and.37) and (16. ∞.

451: Introduction to Economic Growth In addition. since there can be capital-augmenting technical change.e..45) The relationship between the relative share of capital and the normalized capital stock depends on ε. the only asymptotic (non-cycling) paths will feature purely labor-augmenting technical change. they have limt→∞ M This result demonstrates that with ε < 1. but this is only to keep the state of technology in that sector at a constant level. with labor and capital as gross complements. all APs are BGPs and feature purely labor-augmenting tech˙ (t) /M (t) = 0. wL γ (16..38) and (16. Here the numerator contains the “eﬀective units of capital” as well. r = R(M.43) which is a direct generalization of the normalized capital stock deﬁned in the neoclassical growth model as capital stock divided by the eﬀective units of labor.14. sK will also increase if ε > 1. In response to an increase in k . as sK = rK 1 − γ ε−1 = pk = k ε . and will decrease if ε < 1. which is the elasticity of substitution between capital-intensive and laborintensive goods. nical change. Equation (16. I deﬁne a normalized capital stock. i.45) shows that ε is also the elasticity of substitution between capital and labor in this economy.44) Also. (16. i. k ) ≡ β (1 − γ ) M γk (16. we can write the interest rate as: 1 i ε− h 1 1 − ε− ε + (1 − γ ) . (16.37). This analysis leads to the following crucial result: Proposition 47 With ε < 1. There will be research eﬀort devoted to the invention of capital-intensive intermediates.43). 365 .32). NL (16.e. deﬁne the “relative share of capital”. using (16. sK . Then. k≡ MK .

and the capital stock. Furthermore. K and n. The denominator for Vl is diﬀerent from that of Vk because its BGP growth rate is lower than that of Vk : n. M/M = 0. β r + δ − (1 − 2β ) g/ (1 − β ) β r+δ−g (16. and characterize the properties of this equilibrium path.3. the BGP rate of interest has to be ˙ constant. grows along the balanced growth path. must remain constant. and therefore. allowing for the depreciation of technologies at the rate δ . 366 . output. Theorem 7. w. and the growth of w. p.39). Y . while m remains constant. Now I show that there in fact exists a unique BGP as long as δ > 0.14. pK .4 The Balanced Growth Path We saw above that with ε < 1.451: Introduction to Economic Growth This is essentially a generalization of the steady-state growth theorem. (16. to obtain the values of inventing labor. the wage rate. equation (16.46) Notice that these values also grow at a constant rate along the BGP because w. which showed that balanced growth is only consistent with purely labor-augmenting technological change. which is in the denominator of π l .and capital-intensive goods as: Vl = wL/n 1−β 1 − β rK/m and Vk = . only a BGP with purely labor-augmenting technical change can be an AP. for p to remain constant. K and n are growing. Therefore. In addition. Moreover. the relative price of capital-intensive goods.29). will all grow at a common rate. K . First note that from the Euler equation. We can then integrate equation (16. 16. g . in BGP. with M constant. since from Proposition 47. (16.31) implies that YL and YK should grow at the same rate. n has to grow at the rate βg/ (1 − β ) (or N has to grow at the rate g ).44) immediately implies that the price index for capital-intensive goods. This proposition shows the same is the case in this more general model.

Sk = Sk scientists will work on labor-augmenting technical change. r∗ + δ − (1 − 2β ) g∗ / (1 − β ) r∗ + δ − g∗ (16. The interest rate has to be higher when the growth rate is higher in order to convince consumers to delay consumption..and labor-augmenting technical ∗ ∗ change. It is clear from (16. ∗ ∗ bl φ (S − Sk ) wL ) r∗ K bk φ (Sk = .e. θ. bl φ (S − Sk ) nVl = bk φ (Sk ) mVk . M has to increase in order to keep the interest rate at r∗ ..44) that G0 > 0–that is. The Euler equation (16. so capital has to become more productive. ∗ bk φ (Sk ) ((1 − β ) (ρ + δ ) + ((1 − β ) (θ − 1) + β ) g ∗ ) (16. must satisfy: σ K = b∗ ≡ ∗ ) (1 − β ) (ρ + δ + (θ − 1) g ∗ ) bl φ (S − Sk . Next. This is because a greater k implies a lower price of capital-intensive goods. there is a strictly increasing relationship between M and k.e.49) 367 . The remaining change. The growth rate of the economy is therefore g∗ = ˙ 1−βn 1−β ∗ ∗ ) (S − Sk ) − δ] . or from equation (16. i. and the elasticity of marginal utility. σ K . let k∗ be the level of normalized capital such that at this normalized capital stock ˙ and at M/M = 0.47) Assumption 14 ensures that g∗ > 0. k )). This implies φ (Sk ) Sk = δ/bk .. at k = k∗ . Let k = G(M ) such that M and k are consistent with BGP (i.451: Introduction to Economic Growth Recall that in BGP. the relative share of capital. i. determines how strong this eﬀect needs to be. R&D ﬁrms are indiﬀerent between capital.e. p and m are constant.e.10)..48) This implies that. r∗ = R(M. = [bl φ (S − Sk β n β (16. so there is no net capital-augmenting technical ∗ as deﬁned above.29) then gives the BGP interest rate as r∗ = ρ + θg ∗ .14. i.

M ∗ is the level of capital-augmenting technology that is consistent with the equilibrium interest rate taking its BGP value when k = k∗ . the interest rate will be equal to r∗ and the relative share of capital will be b∗ . As a result.49) to hold. there must be both research to invent new labor-intensive and capital-intensive intermediates–if ˙ there were no research directed at capital-intensive intermediates. As a result. we have: k=k ≡ ∗ µ γb∗ 1−γ ε ¶ ε− 1 ⇐⇒ σ K = b∗ . Because of the depreciation of technologies. This proposition characterizes the unique BGP. factor shares remain constant in the long run. which features purely labor-augmenting technical change. k = k∗ .451: Introduction to Economic Growth with g ∗ given by (16. using equation (16. i. This implies that ﬁrms working to invent both types of goods have to make equal proﬁts. despite growth and capital deepening.45).50).e.. consumption and wages grow at the rate g∗ given by (16. There is just enough capital-augmenting technical change to keep the productivity of capital constant–that is. we would have M/M < 0. so we need conditions (16. r = r∗ = ρ + θg ∗ . ˙ ˙ In BGP. most research is devoted to the invention of labor-intensive intermediates. R&D ﬁrms are 368 . In other words.48) and (16. when the relative share of capital is equal to σ K = b∗ .14. Then there exists a unique BGP where k = k∗ as given by (16. which in turn requires that M = M ∗ so that r = r∗ . We can therefore state: Proposition 48 Suppose that ε < 1 and δ > 0. (16. let M ∗ be such that k∗ ≡ G(M ∗ ). and output.47). there is no net capital-augmenting technical change.. In this BGP.e.50) Finally. M/M = 0. when k = k∗ and M = M ∗ . i. M = M ∗ = G−1 (k∗ ). while N/N > 0. Intuitively.47).

and the capital stock grow at the same rate g∗ given by (16. consumption.49) and (16. Each BGP has a diﬀerent normalized capital stock. but there are now many balanced growth paths.e. we need M = M ∗ –i. These paths have the same growth rate.47) with δ = 0. but not continuous. but diﬀerent factor distributions of income. The intuition for the multiplicity of BGPs is simple: without depreciation. we will see also below that. and this can happen for a range of capital 369 . We have already seen that when ε < 1. so starting from diﬀerent initial conditions.50) with δ = 0. all that is required for a BGP is that labor-augmenting improvements should be more proﬁtable than capital-augmenting improvements. σ K . Then. where k∗ is given by (16.14. no net capital-augmenting technical change–is not surprising.e. δ = 0.. this BGP is dynamically stable.e. the BGP with purely laboraugmenting technical change is the only possible asymptotic equilibrium path. there exists a BGP for each M ≥ M ∗ ≡ G−1 (k∗ ). in δ at δ = 0. output. In addition. This reﬂects that the equilibrium correspondence is lower-hemi continuous. and the share of labor is constant.451: Introduction to Economic Growth just indiﬀerent between inventing capital-intensive and labor-intensive intermediates. g ∗ (given by (16.47) evaluated at δ = 0). Summarizing this result: Proposition 49 Suppose that ε < 1 and δ = 0. i. so in equilibrium they allocate their eﬀort between the two sectors precisely to keep the relative share of capital at b∗ .. In all BGPs. Given the CRRA preferences. the conclusion that for a BGP with constant interest rate and growth rate. What is important (perhaps surprising). under certain conditions. however. The results are similar in spirit when there is no technological depreciation. k = G (M ). the economy will tend towards this growth path. i. wages. and a diﬀerent relative share of capital. is that such a BGP exists despite the possibility of capital-augmenting technical change. Vk ≤ Vl .

we have the following result (which is proved in Acemoglu.6 Policy Implications Despite the similarity of this model to the neoclassical one. the feature that ε < 1 ensures this.14. In contrast. It can be veriﬁed that in the standard neoclassical growth model with exogenously laboraugmenting technological change. the implications are actually quite diﬀerent.3. 16.451: Introduction to Economic Growth (labor) shares. so that the budget constraint of the representative household becomes: C + I ≤ wL + (1 − τ ) rK + ω S S + Π + T. we would like to know whether the economy will tend to be balanced growth path with three labor augmenting technical change. here we have: 370 .5 Transitional Dynamics Finally. In particular. 16.3. Here. 2003): Proposition 50 Suppose δ > 0 and that ε < 1. Suppose that there is taxation of capital income. Therefore. this model provides a framework in which technological change can be capitalaugmenting in the short run or in the median run. but in the long run it will be endogenously labor-augmenting. an increase in τ will aﬀect the capital to eﬀective labor ratio and the share of capital in national income. then the BGP characterized above is locally saddle-path stable. Let us consider one example here. ensuring a balanced growth path equilibrium as in the standard neoclassical growth model.

14. Therefore. the composition of technology between capital-augmenting and labor-augmenting types adjusts endogenously in order to restore the rate of interest and the share of capital in national income back to its BGP level. 371 . then the BGP capital share in national income is constant and independent of τ . in this model. but may have much less eﬀect on long-run growth properties. The reason for this result is interesting: when taxes reduce the rate of return to capital. a variety of policies aﬀect the composition of technological change.451: Introduction to Economic Growth Proposition 51 Suppose δ > 0 and that ε < 1.

14.451: Introduction to Economic Growth 372 .

Recall that in previous models technological diﬀerences were often explained by assuming that technologies did not freely ﬂow from advanced countries to less advanced ones. etc. but even when ideas could ﬂow at no cost.Chapter 17 Recitation Material: Appropriate Technology Thinking of the composition of technology also opens the way for us to consider issues of appropriate technologies. Most technologies are developed in the North. Why should ideas not ﬂow and machines not be exported to poor countries? Perhaps distortions as in the previous model. productivity diﬀerences may stay. There is a mismatch between technologies developed in the North and LDCs’ weather conditions. where most technologies are developed. Why? Many technologies used by LDC’s are “inappropriate” because they are designed to make optimal use of the prevailing factors and conditions in DCs. For example. over 90% of the world R&D 373 . labor force skills.

1). there is a loss in eﬃciency. k0 0 for η ∈ (0.14. Moreover. 374 . For example. For example. Then productivity in a less developed country with the capital-labor ratio k < k 0 will be Y = A (k | k0 ) k1−α L = Ak 1−α+η (k 0 ) −η L So less developed countries will produce with worse technologies. suppose that ½ µ ¶η ¾ k A (k | k ) = A min 1.1 Diﬀerences in Capital-Labor Ratios (Atkinson-Stiglitz) Atkinson and Stiglitz suggested the following idea: new technologies are speciﬁc to a given capital-labor ratio. That is. and A (k | k0 ) is the productivity of technology designed to be used with capital-labor ratio k0 when used instead with capital-labor ratio k. they are less productive. which have greater capital-labor ratios. When used with diﬀerent capital labor ratios. 17.451: Introduction to Economic Growth expenditure takes place in OECD economies. Now suppose that new technologies are developed in richer economies. suppose that the production technology is Y = A (k | k0 ) K 1−α Lα = A (k | k0 ) k1−α L where k = K/L is the capital-labor ratio. when a technology designed for the capital labor ratio k 0 is used with a lower capital-labor ratio. this technological disadvantage will be larger when the gap in the capital intensity of production between these countries and in the technologically advanced economies is greater.

1 A Model Consider two groups of countries: the North and the South.14. Z∞ t C (τ )1−θ − 1 exp(−ρ(τ − t))dτ . Another possibility is mismatch between the skill requirements of the frontier technologies in the rich economies and the available skills in the LDCs. A recent paper by Basu and Weil (1998) presents a dynamic formulation based on the idea that A is determined not only by the current capital-labor ratio in the technologically advanced economies. 1−θ Macroeconomic equilibrium: 375 . but by the whole history of these capital-labor ratios. and this creates a mismatch between these technologies and the supply of human capital in the LDCs. 17.2. H n /Ln > H s /Ls All technological progress originates in the North. 17. Here I will outline a model where the skill requirements of new technologies are determined by directed technical change in the technologically advanced economies.451: Introduction to Economic Growth The problem with this formulation is that it is static. But the South can adopt technologies without any impediments or costs.2 The Role of Human Capital (Acemoglu-Zilibotti) The Atkinson-Stiglitz and Basu-Weil approach emphasizes diﬀerences in capital intensity between rich and poor economies.

There is a continuum of machines j ∈ [0. or an industry that will contribute to ﬁnal output. v ) 1−β 0 dv [(1 − i)l(i)] + ¸ β ∙Z NH kH (i. skilled workers are more productive in tasks/industries with high indices. Technological parameter. produces and rents machines at the rental rate χz (v). v ) 1−β dv [iZh(i)]β . Final goods sector is competitive and rents capital and labor services. as in the basic directed technical change model. 0 ¸ (17. Z .14.1) Here i denotes either a task that needs to be performed for production. Technology: y (i) = ∙Z NL kL (i. Skilled and unskilled labor have diﬀerent comparative advantages across sectors. The productivities of these two technologies are parameterized by NL and NH .2) There are 3 important features embedded in this technology: 1. the other one using unskilled labor.451: Introduction to Economic Growth C + I + X ≤ Y ≡ exp ∙Z 1 ln y (i)di . 2. and a continuum j ∈ [0. determines how productive skilled workers are relative to unskilled workers. In particular. NH ] (complementary to skilled workers). 0 ¸ (17. 3. one using skilled workers. NL ] (complementary to unskilled workers). Each task/industry output can be produced using two alternative technologies. A (technology) monopolist owns the patent for each type of machine. 376 .

Technical progress: increases in NL and NH (as in the baseline directed technical change model discussed above). Z NL Z NH χL (v ) kL (i. v )dv − χH (v ) kH (i.451: Introduction to Economic Growth Producers of the ﬁnal good i ∈ [0. v )dv. we obtain output in sector i as y (i) = δ −1 p(i)(1−β )/β [NL (1 − i)l(i) + NH iZ · h(i)] .3). NL and NH are the only state variables in the model. and In equilibrium: 377 ∀ i > J . χz (v) and replace kz (v ) in production functions: h i1/β (1 − β )p(i) ((1 − i)l(i))β /χL (v ) . Substituting machine prices into (17. ∃ a threshold J ∈ [0. p (i) y (i) − wL l (i) − wH h (i) − 0 0 Solve for equilibrium kz (v ). • ∀ i < J . The equilibrium will take a similar form both in the North and in the South.2). 1] such that skilled workers will be used only in sectors i > J . 1] are price takers. v ) = (1 − β )p(i) (iZh(i))β /χH (v ) kL (i. l(i) = 0. v ) = monopolists will again be a constant markup over marginal cost. They maximize proﬁts. h(i) = 0.14. the equilibrium is straightforward to characterize. More explicitly. kH (i. the optimal rental rates for the technology .3) (17.4) Given these isoelastic demand for machines. h i1/β . and then using the resulting expressions with (17. Now taking NL and NH . (17.

. This implies that for technological equilibrium: n PH = n PL µ ZH n Ln ¶−β . the innovation possibilities frontier as given by (16. (17. and h(i) = H/(1 − J ). PL NL L The equilibrium threshold will be given by µ ¶−1/2 NH ZH J = .12). steady state/balanced growth requires π H = π L . where PL and PH are price indices for goods produced intensively using the skilled or unskilled workers. Relative price of skill-intensive goods: ¶−β/2 µ PH NH ZH = . In particular. assuming no “state dependence” (i. we have to determine NL and NH . p(i) = PL · (1 − i)−β • ∀ i > J. Wage premium: wH =Z wL µ NH NL ¶1/2 µ ZH L ¶−1/2 Total output: . This is similar to the analysis of directed technical change we saw above. p(i) = PH · i−β and l(i) = L/J .e.5) Next. 1−J NL L £ ¤2 Y = exp(−β ) (NL L)1/2 + (NH ZH )1/2 .451: Introduction to Economic Growth • ∀ i < J.14. 378 . in terms of the model of directed technical change above.

451: Introduction to Economic Growth Hence. the skill premium in the North is n n wH /wL =Z independent of factor endowment in the North (this is the eﬀect of directed technical change. NL Jn Ln (17.6) It is interesting that in steady state. and maximize proﬁts. the equilibrium relative technologies have to satisfy: 1 − Jn ZH n NH = = . In equilibrium: Js > Jn In other words. Next. Technology levels NH and NL are determined in the North. assume that Southern producers take NL and NH from the North. but also the special case corresponding to σ = 2 in terms of the directed technical change model above). and Y s grows at the same rate g as in the North. 379 . certain tasks that are performed by skilled workers in the North will be performed by unskilled workers in the South–this is simply an implication of the greater skill abundance in the North. and the South is the follower. This captures the notion that the North is the technologically advanced economy. A “monopolist” in each Southern country copies each new machine and sells it to the producers in its country.14.

This will imply that capital-labor ratios may diﬀer between countries. Since J s > J n . so technologies developed in the North are more skill-biased then what is required in the South. and the South importing technologies developed in the North? Deﬁne A= y= Y L+ZH Y L+H : output per eﬀective unit of labor : output per capita.14. if cost of capital is higher in the South. 17.2. leading to endogenous productivity diﬀerences between these countries. Both output per eﬀective unit of labor and output per capita are greater in the North than the South.2. there are many more skilled workers in the North.3 Calibration Can this mechanism lead to sizable eﬀects? Can we generate output per worker diﬀerences which resemble those in the data? Can we improve on the neoclassical model? 380 . This is true a fortiori. even if both countries have the same cost of capital.451: Introduction to Economic Growth Let us allow for the price of capital to be larger in the South than in the North (see Jones. 1995).2 Implications What are the productivity implications of directed technical change in the North. 17. For example. these skill-biased technologies will be less useful in the South than in the North. Why? The world technologies are designed to make best use of factor abundance/scarcity in the North. Intuition: “TFP” is maximized in the North.

NL is chosen so as to normalize yAZ 381 . A (identical for all countries) chosen so as to normalize yNC AZ model.451: Introduction to Economic Growth Comparison between: (benchmark neoclassical model): c = A · (K c )1−β · (Lc + ZH c )β . NH /NL = ZH n /Ln North = U S U SA = 1.14. YNC U SA = 1. allowing for diﬀerences in the price of capital): c = NL · (K ) YAZ c 1−β · Ã (L ) c 1/2 + µ NH · ZH c NL ¶1/2 !2β .

26 0.31 0.14.03 (GDP 5th poorest country) 382 .0 0.8 0.625 0. compl.19 0.745 0.36 0.21 0.43 1.14 0.0 0.05 0.18 0. GDP per worker in non-OECD).651 0.07 0.49 Higher 1.0 0.45 1.540 0.723 0.540 0.728 0.40 0.21 0.39 0.840 0.934 0.17 0.11 0.13 0.17 0.45 1.843 0.931 0.10 0.49 Sec.49 y LDC = 0.39 Higher Primary Sec.21 0. att.8 0. 1.09 0.37 0.5 0. compl.28 0. Z LDC y ˆNC <2 NC LDC y ˆAZ <2 AZ 0.8 0.16 0. 1.918 0.540 0.39 0.666 0.09 0.07 0. 1.5 0.803 0.08 0.39 0.49 1.757 0. compl.689 1.17 Sec. 1.816 0.5 0.718 0.0 0.937 0.42 Higher Primary Sec. att.707 0.808 0. att.15 0.41 Sec.15 0.8 0. 1. y 5th− = 0.05 0.540 0.451: Introduction to Economic Growth Neoclassical model 5th− y ˆNC Our model 5th− y ˆAZ H/L Primary Sec.31 0.21 (avg.901 0.41 0.46 1.21 0.16 0.28 0.44 0.5 0.

or why some countries grow faster than while others stagnate. and how technology endogenously changes and is transferred from one country to another. how physical and human capital aﬀect economic growth and income levels. human capital and technology. The models we have seen are very useful for understanding how individuals accumulate capital. Exactly as in the empirical analysis of decomposing cross-country income diﬀerences into diﬀerences in physical capital. what we have focused on are the proximate causes of this process. the major question motivating much of the analysis of economic growth is to understand why some countries are rich while some others are poor. we have learned how to construct microfounded models which help us in thinking about the process of economic growth in a careful and rigorous way. At some level.Chapter 18 Epilogue: Political Economy of Growth This course so far has been about understanding the mechanics of economic growth. However. 383 .

18.14. the next step in the study of economic growth is to understand why diﬀerent countries adopt diﬀerent policies. the answer comes to preferences. Therefore. I include a brief discussion of some of the issues here and also provide a very simple model showing how institutions and policies can be incorporated into a simple growth-type model to analyze how distributional conﬂict inﬂuences the growth prospects of a society.451: Introduction to Economic Growth But after seeing all these models. associated with diﬀerences in the organization of society. The models we have seen give us a way of translating diﬀerences in preferences. institutional diﬀerences. This is the realm of the political economy of growth or political economy of development.1 Thinking of Institutions and Growth As discussed above. shape economic and political incentives and aﬀect the nature of equilibria via these 384 . Nevertheless. policies. can we answer the question of why is Nigeria poorer than the United States? The answer is yes and no. probably with more emphasis on the no. These topics fall beyond the scope of this course. policies and institutions. In this context. institutions contrast with other potential fundamental causes such as geographical diﬀerences or cultural factors. institutions (and sometime technology) into diﬀerences in growth rates and income levels. as a pointer for those who are interested in thinking about these topics. institutions (and related policy diﬀerences originating from institutional diﬀerences) have become popular recently in thinking of fundamental causes of diﬀerences in income per capita and growth performance of countries. Inevitably. While geographic characteristics of countries and regions may lead to diﬀerences in the technology available to individuals or make their investments in physical and human capital more diﬃcult.

7 percent of GDP per capita in Nigeria. and we have seen some of those in the early lectures of this course. A voluminous literature documents large cross-country diﬀerences in economic institutions. La Porta.02 percent of GDP per capita in 1999. Djankov et al. p. which are outside human control. All of these authors ﬁnd substantial diﬀerences in these measures of economic institutions. 3) oﬀers the following deﬁnition: “Institutions are the rules of the game in a society or. and work by Djankov. (2) that they are “the rules of the game” setting “constraints” on human behavior.16 percent in Kenya 0.1. Mauro’s (1995) study looks at measures of corruption. 1.” Three important features of institutions are apparent in this deﬁnition: (1) that they are “humanly devised.” which contrasts with other potential fundamental causes.91 percent in Ecuador and 4. more formally. and a strong correlation between these institutions and economic performance. 18. Lopez-De-Silanes and Shleifer compiles measures of entry barriers across countries. Knack and Keefer (1995). There are tremendous cross-country diﬀerences in the way that economic and political life is organized. and signiﬁcant correlation between these measures and various indicators of economic performance. are the humanly devised constraints that shape human interaction. ﬁnd that. For example.14. for instance. the same cost is 2. 1981).95 percent 385 . like geographic factors. while the total cost of opening a medium-size business in the United States is less than 0. look at measures of property rights enforcement compiled by international business organizations.1 The Impact of Institutions Douglass North (1990. while many studies look at variation in educational institutions and the corresponding diﬀerences in human capital. (3) that their major eﬀect will be through incentives (see also North.451: Introduction to Economic Growth channels.

Canada. Central America. especially those in Africa. After all.451: Introduction to Economic Growth in the Dominican Republic. for example. Europeans dominated and colonized much of the rest of the Globe. the United States diﬀers from Nigeria. In fact. Kenya and the Dominican Republic in its social. the Caribbean and South Asia. in one form or another. nor did they provide checks and balances against the government. European colonization of the rest of the world provides a potential laboratory to investigate these issues. Together with European dominance came the imposition of very diﬀerent institutions and social power structures in diﬀerent parts of the world. European powers set up “extractive states”. This colonization strategy and the associated institutions contrast with the institutions Europeans set up in other colonies. These institutions (again broadly construed) did not introduce much protection for private property. From the late 15th century. Johnson and Robinson. Nevertheless. including the rate of economic growth and the level of development. Consequently. (2001) document that in a large number of colonies. so that we approximate a situation in which a number of otherwise-identical societies end up with diﬀerent sets of institutions. These entry barriers are highly correlated with various economic outcomes. especially in colonies where they settled in large numbers. Australia and 386 . these diﬀerences may be the source of institutional differences themselves. one needs to isolate a source of exogenous diﬀerences in institutions. this type of correlation does not establish that the countries with worse institutions are poor because of their institutions. Acemoglu.14. The explicit aim of the European in these colonies was extraction of resources. cultural and economic fundamentals. To make further progress. so these may be the source of their poor economic performance. AJR. geographic. the United States. evidence based on correlation does not establish whether institutions are important determinants of economic outcomes. as already discussed in the earlier lectures.

for sustained growth property rights for a broad cross section seem to be crucial (AJR. In practice. Acemoglu. In these colonies the emphasis was on the enforcement of property rights for a broad cross section of the society. Based on this idea. but by aﬀecting the settlement patterns of Europeans. but the vast majority of the population enjoys no such rights and faces signiﬁcant barriers preventing their participation in many economic activities. A crucial determinant of whether Europeans chose the path of extractive institutions was whether they settled in large numbers. Consequently. These mortality rates should not inﬂuence output today directly. especially smallholders. the property rights of the elite are often secure.451: Introduction to Economic Growth New Zealand. Although investments by the elite can generate economic growth for limited periods. the path of institutional development should have been diﬀerent from areas where Europeans faced high mortality rates.14. merchants and entrepreneurs. the institutions were being developed for their own future beneﬁts. 2003). since even in the societies with the worst institutions. In colonies where Europeans settled. The term “broad cross section” is emphasized here. they may have had a ﬁrst-order eﬀect on institutional development. In colonies where Europeans did not settle. and other associated institutions. to oppress the native population and facilitate the extraction of resources in the short run. 2002a. Europeans faced widely diﬀerent mortality rates in colonies because of diﬀerences in the prevalence of malaria and yellow fever. during the time of colonization. AJR (2001) suggest that in places where the disease environments made it easy for Europeans to settle. They provide a possible candidate for a source of exogenous variation in institutions. their objective was to set up a highly centralized state apparatus. these potential settler mortality rates can be used as an instrument for broad institutional diﬀerences across countries in an instrumental-variables estimation strategy. The key requirement for an instrument is that it should have no direct eﬀect on the 387 .

consistent with the key idea in AJR (2001). but they had much more limited eﬀects on natives who. for example. Moreover. lead to as much as a 7-fold increase in Nigeria’s income. They also show that these institutional diﬀerences induced by mortality rates and European settlement patterns have a major (and robust) eﬀect on income per capita. the estimates imply that improving Nigeria’s institutions to the level of those in Chile could. over centuries. thus having a major eﬀect on settlement patterns.14. in the long run. there are also good reasons for why. mortality rates faced by Europeans were not the only determinant of Europeans’ colonization strategies. The data also show that there were major diﬀerences in the institutional development of the high-mortality and low-mortality colonies. these mortality rates should not have a direct eﬀect. AJR (2002) focus on another important aspect. various measures of broad institutions. had developed various types of immunities. as a ﬁrst approximation. They document that in more denselysettled areas. measures of protection against expropriation. Europeans were more likely to introduce extractive institutions because it was 388 . This evidence suggests that once we focus on potentially-exogenous sources of variation. Naturally. The exclusion restriction is also supported by the death rates of native populations. There are a number of channels through which potential settler mortality could inﬂuence current economic outcomes or may be correlated with other factors inﬂuencing these outcomes. For example. which appear to be similar between areas with very diﬀerent mortality rates for Europeans. are highly correlated with the death rates Europeans faced more than 100 years ago and with early European settlement patterns. the data points to a large eﬀect of broad institutional diﬀerences on economic development. Nevertheless.451: Introduction to Economic Growth outcome of interest (other than its eﬀect via the endogenous regressor). how densely diﬀerent regions were settled before colonization. Malaria and yellow fever were fatal to Europeans who had no immunity.

either by having them work in plantations and mines. the South Korean state was generally supportive of rapid development and is often credited with facilitating. or by maintaining the existing system and collecting taxes and tributes. But this is only part of a ﬁrst step in the journey towards an answer. relied on a capitalist organization of the economy. The geopolitical balance between the Soviet Union and the United States following the WWII led to separation along the 38th parallel. Although not democratic during its early phases. under the dictatorship of Kim Il Sung. and AJR (2002) show similar large eﬀects from this source of variation. The next question is even harder: if institutions have such a large eﬀect on economic riches. The North. Another example that illustrates the consequences of diﬀerence in institutions is the contrast between North and South Korea. This evidence suggests that to understand why some countries are poor we should understand why their institutions are dysfunctional. In the meantime. the economies of North and South Korea diverged. investment and rapid growth in Korea. South Korea. Under these two highly contrasting regimes.451: Introduction to Economic Growth more proﬁtable for them to exploit the indigenous population. why do some societies choose. with private ownership of the means of production. Overall. This suggests another source of variation in institutions that may have persisted to the present. or even encouraging. a variety of evidence paints a picture in which broad institutional diﬀerences across countries have had a major inﬂuence on their economic development. especially those under the umbrella of the chaebols. and legal protection for a range of producers. under communist institutions and policies. though far from a free-market economy.14. North Korea experienced minimal growth since 1950. adopted a very centralized command economy with little role for private property. end up with and maintain these dysfunctional 389 . the large family conglomerates that dominated the South Korean economy. While South Korea grew rapidly under capitalist institutions and policies.

and the political power of the diﬀerent groups will be the deciding factor. determine the constraints on and the incentives of the key actors.2 Modeling Institutional Diﬀerences As a ﬁrst step in modeling institutions. at the end of the day. economic institutions matter for economic growth because they shape the incentives of key economic actors in society. This leads to a conﬂict of interest among various groups and individuals over the choice of economic institutions. they inﬂuence investments in physical and human capital and technology. de jure (formal) and de facto political power (see Acemoglu and Robinson. The distribution of political power in society is also endogenous. To make more progress here. and herein lies part of the problem: diﬀerent institutions will not only be associated with diﬀerent degrees of eﬃciency and potential for economic growth. And because of their inﬂuence on the distribution of economic gains. Political institutions. similar to economic institutions. including history and chance. but also with diﬀerent distribution of the gains across diﬀerent individuals and social groups. (2) political power. not all individuals and groups typically prefer the same set of economic institutions.14.1. and the organization of production. De jure political power refers to power that originates from the political institutions in society. Economic institutions not only determine the aggregate economic growth potential of the economy. 2005). As already mentioned above. (3) political institutions. 390 .451: Introduction to Economic Growth institutions? 18. but also the distribution of resources in the society. let us distinguish between two components of political power. in particular. let us consider the relationship between three institutional characteristics: (1) economic institutions. economic institutions are collective choices of the society. How are economic institutions determined? Although various factors play a role here.

This type of de facto political power originates from both the ability of the group in question to solve its collective action problem and from the economic resources available to the group (which determines their capacity to use force against other groups). including changes in technologies and the international environment. This discussion highlights that we can think of political institutions and the distribution of economic resources in society as two state variables. reproducing the initial disparity. the distribution of political power in society is the key determinant of their evolution. In particular. have the potential to 391 . dictatorship or autocracy.14. even if they are not allocated power by political institutions. this will increase its de facto political power and enable it to push for economic and political institutions favorable to its interests. and they will generally opt to maintain the political institutions that give them political power. This creates a central mechanism of persistence: political institutions allocate de jure political power. An important notion is that of persistence . aﬀecting how political power will be distributed and how economic institutions will be chosen. for example. political institutions are collective choices. democracy vs. and the extent of constraints on politicians and political elites. Examples of political institutions include the form of government. the framework also emphasizes the potential for change. hire mercenaries. or undertake protests in order to impose their wishes on society.451: Introduction to Economic Growth but this time in the political sphere. for example. may possess political power. A group of individuals. Since. they can revolt. A second mechanism of persistence comes from the distribution of resources: when a particular group is rich relative to others. the distribution of resources and political institutions are relatively slowchanging and persistent. like economic institutions. and those who hold political power inﬂuence the evolution of political institutions. use arms. co-opt the military. Despite these tendencies for persistence. “shocks” to the balance of de facto political power.

and the role of the combination of de jure and de facto political power in shaping both economic and political institutions. Lack of property rights for landowners. 18.451: Introduction to Economic Growth generate major changes in political institutions. A simple way of summarizing some of these ideas in the form of a ﬂow diagram is as follows: ⎧ ⎫ ⎪ ⎪ ⎪ economic de jure ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ performancet ⎪ economic political political ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ =⇒ institutionst =⇒ institutionst =⇒ powert ⎪ & ⎪ ⎪ ⎬ ⎪ ⎪ ⎪ ⎪ distribution & ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎩ of resources ⎪ de facto ⎪ distribution t+1 ⎪ ⎪ ⎪ ⎪ ⎪ =⇒ of resourcest =⇒ political ⎪ political ⎪ ⎪ ⎪ ⎪ power ⎭ institutions t t+1 This diagram illustrates both the eﬀect of economic institutions on economic performance and the distribution of resources in a society. what we need is a framework for organizing our approach to the determination of economic institutions and policies.1. and allocate the 392 .industrialists was detrimental to economic growth during this epoch. consider the development of property rights in Europe during the Middle Ages. Since political institutions at the time placed political power in the hands of kings and various types of hereditary monarchies. The monarchs often used their powers to expropriate producers.14.3 Institutions in Action As a brief example. impose arbitrary taxation. such rights were largely decided by these monarchs. merchants and proto. renege on their debts. Therefore. and consequently in economic institutions and economic growth. and are chosen for their dynamic inﬂuences on economic allocations. taking into account that political institutions themselves are endogenous.

especially in England after the Civil War of 1642 and the Glorious Revolution of 1688. These changes in the distribution of political power led to major changes in economic institutions. strengthening the property rights of both land and capital owners and spurring a process of ﬁnancial and 393 . based both on internal and overseas. trade. or technology. Consequently. This de facto power overcame the Stuart monarchs in the Civil War and Glorious Revolution. enabled them to ﬁeld military forces capable of defeating the king. and leaving aside civil wars related to royal succession. and led to a change in political institutions that stripped the king of much of his previous power over policy. economic institutions during the Middle Ages provided little incentive to invest in land.451: Introduction to Economic Growth productive resources of society to their allies in return for economic beneﬁts or political support. solidifying their political power and ensuring the continuation of the political regime. The seventeenth century. These economic institutions also ensured that the monarchs controlled a large fraction of the economic resources in society. How did these major institutional changes take place? In England until the sixteenth century the king also possessed a substantial amount of de facto political power. especially Atlantic. no other social group could amass suﬃcient de facto political power to challenge the king. and in the Netherlands after the Dutch Revolt against the Hapsburgs. physical or human capital. But changes in the English land market and the expansion of Atlantic trade in the sixteenth and seventeenth centuries gradually increased the economic fortunes. however.14. witnessed major changes in the economic and political institutions that paved the way for the development of property rights and limits on monarchs’ power. and failed to foster economic growth. and consequently the de facto power of landowners and merchants opposed to the absolutist tendencies of the Kings. the growing prosperity of the merchants and the gentry. By the seventeenth century.

For example.451: Introduction to Economic Growth commercial expansion. Although these more secure property rights would foster economic growth. and also gives clues about the answers to. why do the groups with conﬂicting interests not agree on the set of economic institutions that maximize aggregate growth? Second. He could promise to respect property rights. James II had to be deposed for the changes to take place. two crucial questions. First. why did the gentry and merchants use their de facto political power to change political institutions rather than simply implement the policies they wanted? The issue of commitment is at the root of the answers to both questions. renege on his promise. The consequence was rapid economic growth. 394 . they were not appealing to the monarchs who would lose their rents from predation and expropriation as well as various other privileges associated with their monopoly of political power. This is why the institutional changes in England as a result of the Glorious Revolution were not simply conceded by the Stuart kings. and a very diﬀerent distribution of economic resources from that in the Middle Ages. culminating in the Industrial Revolution. as exempliﬁed by the numerous ﬁnancial defaults by medieval kings. why do groups with political power want to change political institutions in their favor? In the context of the example above.14. economic institutions that increased the security of property rights for land and capital owners during the Middle Ages would not have been credible as long as the monarch monopolized political power. This discussion poses. Credible secure property rights necessitated a reduction in the political power of the monarch. but then at some point. An agreement on the eﬃcient set of institutions is often not forthcoming because of the complementarity between economic and political institutions and because groups with political power cannot commit to not using their power to change the distribution of resources in their favor.

the gentry and merchants were interested in their proﬁts and therefore in the security of their property rights. In the example above. for example because the collective action problems that are solved to amass this power are likely to resurface in the future. as well as present.14. can become stronger in the future. but also to alter political institutions and the future allocation of de jure power. commitment to future allocations (or economic institutions) is in general not possible because decisions in the future are made by those who hold political power at the time. However. political institutions and changes in political institutions are important as ways of manipulating future political power. In a dynamic world. many revolutions are followed by conﬂict within the revolutionaries. Therefore. However. and thus indirectly shaping future. they would have liked to use their (de facto) political power to secure beneﬁts in the future as well as the present. Using political power to change political institutions then emerges as a useful strategy to make gains more durable. any change in policies and economic institutions that relies purely on de facto political power is likely to be reversed in the future. especially those controlling de jure power. this would not have been a problem. or other groups. In addition.451: Introduction to Economic Growth The reason why political power is often used to change political institutions is related. economic institutions and outcomes. the English gentry and merchants strove not just to change economic institutions in their favor following their victories against the Stuart monarchy. If the gentry and merchants would have been sure to maintain their de facto political power. Therefore. Consequently. de facto political power is often transient. Recognizing this. not only in the present but also in the future. 395 . individuals care not only about economic outcomes today but also in the future.

Revenue extraction: the group in power–the elite–will set high taxes on middle class producers in order to extract resources from them.451: Introduction to Economic Growth 18. however. These taxes are distortionary. Pareto eﬃciency is not a strong enough concept. This source of ineﬃciency results from the absence of non-distortionary taxes. since when distributional issues are important.14. The various sources of ineﬃciencies in policies are 1. An economy in which all of the resources are allocated to a single individual who has no investment opportunities. that the concept of ineﬃciency here is not that of Pareto ineﬃciency. thus growth is stiﬂed. Factor price manipulation: the group in power may want to tax middle class producers in order to reduce the prices of the factors they use in production. which in turn will translate into ineﬃcient (non-growth enhancing) institutions. which implies that the distribution of resources cannot be decoupled from eﬃcient production. may nevertheless be Pareto eﬃcient. I will present a simple model of this which will highlight various sources of ineﬃciencies in policies. Thus the concept of “ineﬃciency” here is being used in the sense of “non-growth enhancing” or “non-surplus maximizing”. the elite ensure lower factor prices and thus higher proﬁts for 396 . The basic setup is one in which an existing elite is in control of political power. This ineﬃciency arises because the elite and middle class producers compete for factors (here labor). 2. It should be noted at this point.2 A Simple Model of Non-Growth Enhancing Institutions Now I present a simple model of the determination of institutions in the context of investigating their impact on economic growth. By taxing middle class producers. and uses their monopoly of political power for their own interests even when this is costly for the society at large.

the revenue extraction is typically the least harmful. 3.451: Introduction to Economic Growth themselves. Following the literature on organizational economics. I refer to this as a holdup problem. greater middle class proﬁts reduce the elite’s political power and endanger their future rents. may improve the allocation of resources. and thus. The ineﬃciencies in policies translate into ineﬃcient institutions. taxes are typically higher and more distortionary. An interesting comparative static result is that greater state capacity shifts the balance towards the revenue extraction mechanism. Additional ineﬃciencies arise when there are “commitment problems” on the part of the elites. and economic institutions determine both the limits of various redistributive policies and other rules and regulations that aﬀect the economic transactions and productivity of producers. I 397 . In contrast. for example. Although all three ineﬃciencies in policies arise because of the desire of the elite to extract rents from the rest of the society. With holdup. are likely to be important. The elite will then want to tax the middle class in order to impoverish them and consolidate their political power. In the context of the simple model here. the analysis will reveal that of the three sources of ineﬃciency. since. in the sense that they may renege on policy promises once key investments are made. Holdup problems. in turn. in order to extract revenues. by allowing the elite to extract resources more eﬃciently from other groups. Institutions determine the framework for policy determination. Political consolidation: to the extent that the political power of the middle class depends on their economic resources. the elite need to ensure that the middle class undertakes eﬃcient investments.14. when the relevant investment decisions are long-term. the factor price manipulation and political consolidation mechanisms encourage the elite to directly impoverish the middle class. so that a range of policies will be decided after these investments are undertaken.

the elite may want to block the adoption of more eﬃcient technologies. regulation on the technology used by middle class producers. or at the very least. and may encourage the elite to use economic institutions to place credible limits on their own future policies (taxes).451: Introduction to Economic Growth associate economic institutions with two features: 1. political institutions govern the process of collective decision-making in society. limits on taxation and redistribution. Holdup problems. This again reiterates that when the factor price manipulation and political consolidation mechanisms are at work. the elite always wish to encourage the adoption of the most productive technologies by the middle class. when the source of ineﬃciencies in policies is factor price manipulation or political consolidation. In the baseline model. While economic institutions regulate ﬁscal policies and technology choices. and 2. which imply equilibrium taxes even higher than those preferred by the elite. create a possible exception. In particular. The same forces that lead to ineﬃcient policies imply that there will be reasons for the elite to choose ineﬃcient economic institutions. The model also sheds light on the conditions under which economic institutions discourage or block technology adoption. 398 . This suggests that economic institutions that restrict future policies may be more likely to arise in economies in which there are more longer-term investments and thus more room for holdup. If the source of ineﬃciencies in policies is revenue extraction. signiﬁcantly more ineﬃcient outcomes can emerge. they would choose not to invest in activities that would increase the productivity of middle class producers. However. they may not want to guarantee enforcement of property rights for middle class producers or they may prefer to block technology adoption by middle class producers.14.

revolutions or military action. Concentrating political power in the hands of the elite may have limited costs (may even be “eﬃcient”). I extend the framework here for analyzing changes in political institutions. middle class producers.451: Introduction to Economic Growth the elite have de jure political power. even though they have no formal say in a dictatorship or 399 . Certain groups may be able to disrupt the existing system. by solving their collective action problem and undertaking demonstrations. existing institutions. for example. become inappropriate to the new economic environment.14. which may have previously functioned relatively well. if the elite are suﬃciently productive (more productive than the middle class). To understand the ineﬃciencies in the institutional framework. However. The same forces that make the elite choose ineﬃcient policies also imply that the answer to this question is no. despite the ineﬃciencies that follow. In this case. a change in the productivity of the elite relative to the middle class could make a diﬀerent distribution of political power more beneﬁcial. Finally. however. The framework also enables me to discuss issues of appropriate and inappropriate institutions. which means that they have the formal right to make policy choices and inﬂuence economic decisions. Consequently. we need to investigate the induced preferences of diﬀerent groups over institutions. Yet there is no guarantee that there will be a change in institutions in response to the change in environment. unrest. as in the example of constitutions or elections determining the party in government. In this context. In the context of political institutions. Political institutions regulate the allocation of de jure political power. the institutional structure with elite control tends to persist. There is more to political power than this type of de jure power. Each group may therefore possess de facto political power even when excluded from de jure political power. this means asking whether the elite wish to change the institutional structure towards a more equal distribution of political power. protests.

14.451: Introduction to Economic Growth an oligarchic society, may sometimes have suﬃcient de facto political power to change the system or at least to demand some concessions from the elite. Under these circumstances, changes in political institutions may emerge as an equilibrium outcome. They are useful as a way of committing to future allocations, because, by aﬀecting the distribution of de jure political power in the future, they shape future policies and economic allocations. Such a commitment may be necessary when the current elite need to make concessions in response to a shift in the distribution of de facto political power and when their ability to make concessions within a given political system is limited. Consequently, changes in political institutions take place when the elite are forced to respond to temporary changes in de facto political power by changing the political system (and thus the distribution of de jure political power in the future). The analysis also shows that changes in political institutions are less likely when political stakes are higher, because, in this case, the elite will ﬁght and use repression to defend the existing regime. Rents from the natural resources or land tend to increase political stakes and thus contribute to institutional persistence. Interestingly, state capacity, which makes redistribution more eﬃcient, also increases political stakes and may create dynamic costs by increasing the longevity of the dictatorship of the elite.

18.2.1

Baseline Model

Consider an inﬁnite horizon economy populated by a continuum 1 + θe + θm of risk neutral agents, each with a discount factor equal to β < 1. There is a unique non-storable ﬁnal good denoted by y . The expected utility of agent j at time 0 is given by:

j U0 ∞ X t=0

= E0

β t cj t,

(18.1)

400

14.451: Introduction to Economic Growth where cj t ∈ R denotes the consumption of agent j at time t and Et is the expectations operator conditional on information available at time t. Agents are in three groups. The ﬁrst are workers, whose only action in the model is to supply their labor inelastically. There is a total mass 1 of workers. The second is the elite, denoted by e, who initially hold political power in this society. There is a total of θe elites. Finally, there are θm “middle class" agents, denoted by m. The sets of elite and middle class producers are denoted by S e and S m respectively. With a slight abuse of notation, I will use j to denote either individual or group. Each member of the elite and middle class has access to production opportunities, represented by the production function

j = yt

1 j 1−α j α (Aj )α (kt ) (lt ) , 1−α t

(18.2)

where k denotes capital and l labor. Capital is assumed to depreciate fully after use. The Cobb-Douglas form is adopted for simplicity. The key diﬀerence between the two groups is in their productivity. To start with, let us assume that the productivity of each elite agent is Ae in each period, and that of each middle class agent is Am . Productivity of the two groups diﬀers, for example, because they are engaged in diﬀerent economic activities (e.g., agriculture versus manufacturing, old versus new industries, etc.), or because they have diﬀerent human capital or talent. On the policy side, there are activity-speciﬁc tax rates on production, τ e and τ m , which are constrained to be nonnegative, i.e., τ e ≥ 0 and τ m ≥ 0. There are no other ﬁscal instruments (in particular, no lump-sum non-distortionary taxes). In addition there is a total income (rent) of R from natural resources. The proceeds of taxes and revenues from natural resources can be redistributed as nonnegative lump-sum transfers targeted towards 401

14.451: Introduction to Economic Growth each group, T w ≥ 0, T m ≥ 0 and T e ≥ 0. Let us also introduce a parameter φ ∈ [0, 1], which measures how much of the tax revenue can be redistributed. This parameter, therefore, measures “state capacity,” i.e., the ability of the states to penetrate and regulate the production relations in society (though it does so in a highly “reduced-form” way). When φ = 0, state capacity is limited all tax revenue gets lost, whereas when φ = 1 we can think of a society with substantial state capacity that is able to raise taxes and redistribute the proceeds as transfers. The government budget constraint is Ttw +θ

m

Ttm

+θ

e

Tte

≤φ

Z

j ∈S e ∪S m

j τj t yt dj + R.

(18.3)

j Let us also assume that there is a maximum scale for each ﬁrm, so that lt ≤ λ for all j

and t. This prevents the most productive agents in the economy from employing the entire labor force. Since only workers can be employed, the labor market clearing condition is Z

j lt dj ≤ 1,

(18.4)

j ∈S e ∪S m

j ≤ λ, (18.4) implies that if with equality corresponding to full employment. Since lt

θe + θm ≤

1 , λ

(ES)

there can never be full employment. Consequently, depending on whether Condition (ES) holds, there will be excess demand or excess supply of labor in this economy. Throughout, I assume that Assumption 15 θe ≤ 1 1 and θm ≤ , λ λ 402

14.451: Introduction to Economic Growth This assumption ensures that neither of the two groups will create excess demand for labor by itself. Assumption 15 is adopted only for convenience and simpliﬁes the notation (by reducing the number of cases that need to be studied).

18.2.2

Economic Equilibrium

m I ﬁrst characterize the economic equilibrium for a given sequence of taxes, {τ e t , τ t }t=0,1,...,∞

(the transfers do not aﬀect the economic equilibrium). An economic equilibrium is deﬁned as a sequence of wages {wt }t=0,1,...,∞ , and investment and employment levels for all producers, o n£ j j¤ m , lt j ∈S e ∪S m such that given {τ e kt t , τ t }t=0,1,...,∞ and {wt }t=0,1,...,∞ , all producers

t=0,1,...,∞

choose their investment and employment optimally and the labor market clears.

Each producer (ﬁrm) takes wages, denoted by wt , as given. Finally, given the absence of adjustment costs and full depreciation of capital, ﬁrms simply maximize current net proﬁts. Consequently, the optimization problem of each ﬁrm can be written as ¡ ¢ 1 − τj j j t j α j 1−α j α ) ( k ) − kt , max ( A lt − wt lt t j j 1 − α kt ,lt

where j ∈ S e ∪ S m . This maximization yields

j lt

⎧ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎩

j 1/α j j = (1 − τ j A lt , and kt t)

(18.5)

=0

if wt >

α (1 1−α α (1 1−α α (1 1−α

1/α j − τj A t)

∈ [0, λ] if wt = =λ if wt <

1/α j . − τj A t) 1/α j − τj A t)

(18.6)

A number of points are worth noting. First, in equation (18.6), the expression α(1 −

1/α j A / (1 − α) is the net marginal product of a worker employed by a producer of group τj t)

j . If the wage is above this amount, this producer would not employ any workers, and if it is 403

14.451: Introduction to Economic Growth below, he or she would prefer to hire as many workers as possible (i.e., up to the maximum, λ). Second, equation (18.5) highlights the source of potential ineﬃciency in this economy. Producers invest in physical capital but only receive a fraction (1 − τ j t ) of the revenues. Therefore, taxes discourage investments, creating potential ineﬃciencies. Combining (18.6) with (18.4), equilibrium wages are obtained as follows: (i) If Condition (ES) holds, there is excess supply of labor and wt = 0. (ii) If Condition (ES) does not hold, then there is “excess demand” for labor and the equilibrium wage is wt = min ¿ À α α e 1/α e m 1/α m . (1 − τ t ) A , (1 − τ t ) A 1−α 1−α (18.7)

The form of the equilibrium wage is intuitive. Labor demand comes from two groups, the elite and middle class producers, and when condition (ES) does not hold, their total labor demand exceeds available labor supply, so the market clearing wage will be the minimum of their net marginal product. One interesting feature, which will be used below, is that when Condition (ES) does not hold, the equilibrium wage is equal to the net productivity of one of the two groups of producers, so either the elite or the middle class will make zero proﬁts in equilibrium. Finally, equilibrium level of aggregate output is Z Z 1 1 j j e (1−α)/α e m (1−α)/α m (1 − τ t ) (1 − τ t ) A lt dj + A lt dj + R. Yt = 1−α 1 − α e m j ∈S j ∈S The equilibrium is summarized in the following proposition: Proposition 52 Suppose Assumption 15 holds. Then for a given sequence of taxes

m {τ e t , τ t }t=0,1,...,∞ , the equilibrium takes the following form: if Condition (ES) holds, then

(18.8)

404

14.451: Introduction to Economic Growth wt = 0, and if Condition (ES) does not hold, then wt is given by (18.7). Given the wage sequence, factor demands are given by (18.5) and (18.6), and aggregate output is given by (18.8).

18.2.3

Ineﬃcient Policies

Now I use the above economic environment to illustrate a number of distinct sources of ineﬃcient policies. In this section, political institutions correspond to “the dictatorship of the elite” in the sense that they allow the elite to decide the policies, so the focus will be on the elite’s desired policies. The main (potentially ineﬃcient) policy will be a tax on middle class producers, though more generally, this could correspond to expropriation, corruption or entry barriers. As discussed in the introduction, there will be three mechanisms leading to ineﬃcient policies; (1) Resource Extraction; (2) Factor Price Manipulation; and (3) Political Consolidation. To illustrate each mechanism in the simplest possible way, I will focus on a subset of the parameter space and abstract from other interactions. Throughout, I assume that there is ¯ and τ e ¯, where τ ¯ ≤ 1. This limit can be an upper bound on taxation, so that τ m t ≤ τ t ≤ τ institutional, or may arise because of the ability of producers to hide their output or shift into informal production. The timing of events within each period is as follows: ﬁrst, taxes are set; then, investments are made. This removes an additional source of ineﬃciency related to the holdup problem whereby groups in power may seize all of the output of other agents in the economy once it has been produced. Holdup will be discussed below. To start with, I focus on Markov Perfect Equilibria (MPE) of this economy, where strategies are only dependent on payoﬀ-relevant variables. In this context, this means that 405

14.451: Introduction to Economic Growth strategies are independent of past taxes and investments (since there is full depreciation). In the dictatorship of the elite, policies will be chosen to maximize the elite’s utility. Hence, a

m w m e political equilibrium is given by a sequence of policies {τ e t , τ t , Tt , Tt , Tt }t=0,1,...,∞ (satisfying

(18.3)) which maximizes the elite’s utility, taking the economic equilibrium as a function of the sequence of policies as given. More speciﬁcally, substituting (18.5) into (18.2), we obtain elite consumption as ¸ ∙ α e e 1/α e e (1 − τ t ) A − wt lt + Tte , (18.9) ct = 1−α with wt given by (18.7). This expression follows immediately by recalling that the ﬁrst term in square brackets is the after-tax proﬁts per worker, while the second term is the equilibrium wage. Total per elite consumption is given by their proﬁts plus the lump sum transfer they receive. Then the political equilibrium, starting at time t = 0, is simply given by a sequence

m w m e of {τ e t , τ t , Tt , Tt , Tt }t=0,1,...,∞ that satisﬁes (18.3) and maximizes the discounted utility of P t e the elite, ∞ t=0 β ct .

The determination of the political equilibrium is simpliﬁed further by the fact that in the

MPE with full capital depreciation, this problem is simply equivalent to maximizing (18.9). We now characterize this political equilibrium under a number of diﬀerent scenarios.

18.2.4

Revenue Extraction

To highlight this mechanism, suppose that Condition (ES) holds, so wages are constant at zero. This removes any eﬀect of taxation on factor prices. In this case, from (18.6), we also

j = λ for all producers. Also assume that φ > 0 (for example, φ = 1). have lt

It is straightforward to see that the elite will never tax themselves, so τ e t = 0, and will redistribute all of the government revenues to themselves, so Ttw = Ttm = 0. Consequently 406

14.451: Introduction to Economic Growth taxes will be set in order to maximize tax revenue, given by Revenuet = φ m (1−α)/α m A λθm + R τ (1 − τ m t ) 1−α t (18.10)

m at time t, facedownwhere the ﬁrst term is obtained by substituting for lt = λ and for (18.5) m into (18.2) and multiplying it by τ m t , and taking into account that there are θ middle class

producers and a fraction φ of tax revenues can be redistributed. The second term is simply the revenues from natural resources. It is clear that tax revenues are maximized by τ m t = α. In other words, this is the tax rate that puts the elite at the peak of their Laﬀer curve. In contrast, output maximization would require τ m t = 0. However, the output-maximizing tax rate is not an equilibrium because, despite the distortions, the elite would prefer a higher tax rate to increase their own consumption. At the root of this ineﬃciency is a limit on the tax instruments available to the elite. If they could impose lump-sum taxes that would not distort investment, these would be preferable. Ineﬃcient policies here result from the redistributive desires of the elite coupled with the absence of lump-sum taxes. It is also interesting to note that as α increases, the extent of distortions are reduced, since there are greater diminishing returns to capital and investment will not decline much in response to taxes. Even though τ m t = α is the most preferred tax for the elite, the exogenous limit on taxation may become binding, so the equilibrium tax is

RE τm ≡ min {α, τ ¯} t = τ

(18.11)

for all t. In this case, equilibrium taxes depend only on the production technology (in particular, how distortionary taxes are) and on the exogenous limit on taxation. For example, 407

14.451: Introduction to Economic Growth as α decreases and the production function becomes more linear in capital, equilibrium taxes decline. This discussion is summarized in the following proposition (proof in the text): Proposition 53 Suppose Assumption 15 and Condition (ES) hold and φ > 0, then the

RE unique political equilibrium features τ m ≡ min {α, τ ¯} for all t. t = τ

18.2.5

Factor Price Manipulation

I now investigate how ineﬃcient policies can arise in order to manipulate factor prices. To highlight this mechanism in the simplest possible way, let us ﬁrst assume that φ = 0 so that there are no direct beneﬁts from taxation for the elite. There are indirect beneﬁts, however, because of the eﬀect of taxes on factor prices, which will be present as long as the equilibrium wage is positive. For this reason, I now suppose that Condition (ES) does not hold, so that equilibrium wage is given by (18.7). Inspection of (18.7) and (18.9) then immediately reveals that the elite prefer high taxes in order to reduce the labor demand from the middle class, and thus wages, as much as possible. The desired tax rate for the elite is thus τ m t = 1. Given constraints on taxation,

FPM ≡τ ¯ for all t. We therefore have: the equilibrium tax is τ m t = τ

**Proposition 54 Suppose Assumption 15 holds, Condition (ES) does not hold, and φ = 0,
**

FPM then the unique political equilibrium features τ m ≡τ ¯ for all t. t = τ

This result suggests that the factor price manipulation mechanism generally leads to higher taxes than the pure revenue extraction mechanism. This is because, with the factor price manipulation mechanism, the objective of the elite is to reduce the proﬁtability of the 408

In the next subsection. Revenue extraction. whereas for revenue extraction. The role of φ = 0 also needs to be emphasized. diﬀerently from the pure revenue extraction case. It is also worth noting that.451: Introduction to Economic Growth middle class as much as possible. 18. whose wages are being reduced because of the tax policy. In particular. by directly impoverishing the middle class. I discuss how the factor price manipulation mechanism works in the presence of an instrument that can directly raise revenue from the middle class. By itself the factor price manipulation eﬀect led to the extreme result that the tax on the middle class should be as high as possible. but it is also doing so indirectly from the workers. Why is there not a more eﬃcient way of transferring resources to the elite? The answer relates to the limited ﬁscal instruments available to the elite. the elite would like the middle class to invest and generate revenues.2. so they are forced to use ineﬃcient means of increasing their consumption. will serve to reduce the power of the factor price manipulation 409 .6 Revenue Extraction and Factor Price Manipulation Combined I now combine the two eﬀects isolated in the previous two subsections. φ = 0 implies that they cannot use taxes at all to extract revenues from the middle class.14. This will illustrate that the absence of any means of transferring resources from the middle class to the elite is not essential for the factor price manipulation mechanism (though the absence of non-distortionary lump-sum taxes is naturally important). though typically another motive for imposing taxes on the middle class. the tax policy of the elite is not only extracting resources from the middle class. Taxing the middle class at the highest rate is clearly ineﬃcient.

To characterize the equilibrium in this case again necessitates the maximization of (18. Intuitively. As before.7). The interesting case is the one where (ES) does not hold. the maximization problem can be written as ∙ ¸ ∙ ¸ α φ m 1 e e m (1−α)/α m m m A lt θ + R . The solution to this problem can take two diﬀerent forms depending on whether (18.12) (18.7). while wages are given by (18. and θe lt m 1/α m = λ if (1 − τ m A ≥ Ae . and elite producers make zero proﬁts and their only income is derived from transfers.12) is the elite’s net revenues and the second term is the transfer they receive. transfers are obtained from (18.7) and e m + θm lt = 1. this corresponds to the case where the elite prefer to let the middle class producers undertake all of the proﬁtable 410 .13) is the market clearing constraint. so that wages are not equal to zero.14) The ﬁrst term in (18.10). while (18. and are given by the minimum of the two expressions in (18. When Condition (ES) holds and there is excess supply of labor.14) holds in the solution.9). lt t ) (18. This is simply the same as maximizing transfers minus wage bill for each elite producer. The reason is that high taxes also reduce the revenues extracted by the elite (moving the economy beyond the peak of the Laﬀer curve).14) ensures that middle class producers employ as much labor as they wish provided that their net productivity is greater than those of elite producers.14. then w = αAe / (1 − α). A − wt lt + e τ (1 − τ t ) max τm 1−α θ 1−α t t subject to (18.13) (18. Equation (18. and are costly to the elite. If it does. Incorporating the fact that the elite will not tax themselves and will redistribute all the revenues to themselves. wages are equal to zero.451: Introduction to Economic Growth eﬀect. and we obtain the same results as in the case of pure resource extraction.

which is the desired tax rate in the case of pure factor price manipulation. The maximization of (18. α. θe . (18. (18. so if θe = θm and Ae = Am . κ (λ. φ) ≡ 1 − τm 1−α (1 − λθe ) φ t (18. In this case w = α(1 − τ m )1/α Am / (1 − α).15) where I have used the fact that all elite producers will employ λ employees. φ) is always less than ∞. when Assumption 16 holds. this condition is always satisﬁed). and from (18. Consequently. If.e. on the other hand. we have wt = α(1 − τ m t ) and the elite’s problem simply boils down to choosing τ m t to maximize ∙ ¸ 1 φ m α m (1−α)/α m m m 1/α m τ t (1 − τ t ) (1 − τ m A l θ +R − A λ. This implies that τm t is always less than 1. so that τ m t is 411 . α. then the elite generate revenues both from their own production and from taxing the middle class producers. This assumption ensures that the solution will always take the latter form (i.. θe . Moreover.14) does not hold). α. I impose the following additional assumption: Assumption 16 A ≥ φ(1 − α) e (1−α)/α A mθ m θe . Intuitively. The ﬁrst interesting feature is that κ (λ. Rather than provide a full taxonomy. φ) is strictly greater than α/ (1 − α). = κ (λ. t ) e θ 1−α 1−α lm = (1 − λθe ) /θm .14. this condition makes sure that the productivity gap between the middle class and elite producers is not so large as to make it attractive for the elite to make zero proﬁts themselves (recall that φ(1 − α)(1−α)/α < 1.14) does not hold.13). 1/α m m A τ t / (1 − α).15) gives µ ¶ λθe α τm t e 1+ . θ .451: Introduction to Economic Growth activities and maximize tax revenues.

but τ COM declines. among other things. Naturally. and as commented above. i. One interesting implication of this discussion is that when the factor price manipulation eﬀect is more important. since one may have expected that as taxation becomes less costly. First.14. Intuition for this result follows from the observation that an increase in φ raises the importance of revenue extraction. as φ increases. taxation becomes more beneﬁcial (generates greater revenues). the desired tax rate with pure resource extraction. taxes should increase. τm t =τ COM ≡ min ½ ¾ κ (λ. if this level of tax is greater than τ ¯. without directly impoverishing them.. α.16) It is also interesting to look at the comparative statics of this tax rate. revenue extraction is a force towards lower taxes (it makes it more costly for the elite to move beyond the peak of the Laﬀer curve).e.451: Introduction to Economic Growth always greater than α. because greater state capacity enables the elite to extract revenues from the middle class through taxation. greater state capacity enables more eﬃcient forms of resource extraction by the groups holding political power. in this case. this comparative static result suggests that higher state capacity will translate into lower taxes. This might at ﬁrst appear paradoxical. The reason for this eﬀect is again the interplay between the revenue extraction and factor price manipulation mechanisms. the factor price manipulation motive always increases taxes above the pure revenue maximizing level (beyond the peak of the Laﬀer curve). to state capacity. Therefore. the equilibrium tax will be τ ¯. α. as θe increases and the number of elite producers increases. while the revenue maximization motive reduces taxes relative to the pure factor price manipulation case. 1 + κ (λ. Since the parameter φ is related. reducing factor prices becomes more important relative to gathering tax revenue. In other words. Second. When there are more elite producers. φ) (18.τ ¯ . θe . θe . φ) . there will 412 . taxes also increase.

in particular pt = p (θm cm t ) ∈ [0. the previous analysis still applies. This mechanism has been absent so far. Once they come to power. 18. taxes create fewer distortions and this increases the revenue-maximizing tax rate.16) for all t = τ t. Finally. Let us assume that p is continuous and diﬀerentiable with p0 > 0. the model needs to be modiﬁed to allow for endogenous switches of power. (18. the desire of the elite to preserve their political power. let us assume that there is a probability pt in period t that political power permanently shifts from the elite to the middle class. Condition (ES) does not hold. Equilibrium taxes are increasing in θe and α and decreasing in φ. For now. 1] . since the elite were assumed to always remain in power.17) where I have used the fact that income is equal to consumption. Interesting economic interactions arise when this probability is endogenous.2.451: Introduction to Economic Growth typically be greater ineﬃciencies. an increase in α raises taxes for exactly the same reason as above. Institutional change will be discussed in greater detail later.7 Political Consolidation I now discuss another reason for ineﬃcient taxation.14. and COM φ > 0. Once again summarizing the analysis: Proposition 55 Suppose Assumptions 15 and 16 hold. the middle class will pursue a policy that maximizes their own utility. When this probability is exogenous. To illustrate it. Here I will use a simple (reduced-form) model to illustrate the trade-oﬀs and assume that this probability is a function of the income level of the middle class agents. which captures the fact that when the middle 413 . Then the unique political equilibrium features τ m as given by (18.

since the structure of the problem makes it clear that these values will be constant in equilibrium.14.451: Introduction to Economic Growth class producers are richer.18) V (E ) = max τm ⎩ +β £V e (E ) − p ¡ α (1 − τ m )1/α Am θm λ¢ (V e (E ) − V e (M ))¤ ⎭ t t 1−α 414 . suppose instead that Condition RE (ES) holds. wt = 0 and the optimal static policy is τ m ≡ min {α. This reduced-form formulation might capture a variety of mechanisms.17).7). t ) The ﬁrst observation is that if the solution to the static problem involves cm t = 0. This implies that. The dynamic maximization problem then becomes ⎫ ⎧ ¤ £ φ m m α 1 e m (1−α)/α m ⎬ ⎨ A λ + θe 1−α τ t (1 − τ t ) A λθ + R 1−α e . τ ¯} as t = τ discussed above and implies positive proﬁts and consumption for middle class agents. To see the role of the political consolidation mechanism. For example. then the same ﬁscal policy is optimal despite the risk of losing power. the political consolidation mechanism does not add an additional motive for ineﬃcient taxation. we have ⎫ ⎧ £ ⎨ α Ae − w ¤ le + 1e £ φ τ m (1 − τ m )(1−α)/α Am lm θm + R¤ ⎬ t t t t 1−α θ 1−α t V e (E ) = max τm ⎭ ⎩ e e t +β [(1 − p ) V (E ) + p V (M )] t t subject to (18. (18. as long as Condition (ES) does not hold and Assumption 16 holds. ¡ α (1 1−α ¢ 1/α m m m m m − τm A lt θ − wt lt θ . and denote it by V e (E ) when they are in power and by V e (M ) when the middle class is in power. they have greater de facto political power. To investigate this issue we now write the utility of elite agents recursively. they may be more successful in solving their collective action problems or they may increase their military power. In this case.14) and (18.13). when the middle class are richer. Naturally. This modiﬁcation implies that the ﬁscal policy that maximizes current consumption may no longer be optimal. with pt = p I wrote V e (E ) and V e (M ) not as functions of time. (18. (18.

e. yet now not to increase their revenues. greater state capacity. τ m t = τ V e (E ) − V e (M ) > 0 is the case is immediate since when the middle class are in power. by increasing the political stakes. Consequently. R. But in the presence of political competition. but to consolidate their political power. because remaining in power has now become more valuable. Intuitively. a higher φ. the gap between V e (E ) and V e (M ) increases. it leads to greater conﬂict and more distortionary policies. the elite become more willing to sacriﬁce tax revenue (by overtaxing the middle class) in order to increase the probability of remaining in power. they get to tax the elite and receive all of the transfers. Intuitively. 1−α RE It is clear that when p0 (·) = 0. also increases the gap between V e (E ) and V e (M ) (because this enables the group in power to raise more tax revenues) and thus implies a higher tax rate on the middle class. Intuitively. the party in power receives the revenues from natural resources. τ ¯} as above. That when p0 (·) > 0. we obtain τ m ≡ min {α.451: Introduction to Economic Growth The ﬁrst-order condition for an interior solution can be expressed as 1 − α τm t φ−φ + βθe p0 α 1 − τm t µ ¶ α m 1/α m m (1 − τ t ) A θ λ (V e (E ) − V e (M )) = 0. as with the factor price manipulation mechanism. However. by allowing more eﬃcient forms of transfers. These high taxes reduce the income of the middle class and their political power. τ ¯} as long as V e (E ) − V e (M ) > 0. 415 . the elite tax beyond the peak of the Laﬀer curve. t = τ PC > τ RE ≡ min {α. improves the allocation of resources.14. there is a higher probability that the elite remain in power in the future. i. enjoying the beneﬁts of controlling the ﬁscal policy. When R increases. and the tax that the elite sets increases as well. greater state capacity. More interestingly. when there is no political competition. This contrasts with the results so far where R had no eﬀect on taxes.. An interesting comparative static is that as R increases.

SPEs will coincide with the MPEs.14. In this case. This will not be true in the next subsection. Therefore.2. This tax rate is increasing in R and φ. If there is no room for such history dependence. Given the factor demands.451: Introduction to Economic Growth Summarizing this discussion: Proposition 56 Consider the economy with political replacement. Therefore. In the models analyzed so far. so it is useful to brieﬂy discuss why it is the case here. such punishment strategies are not possible even in the SPE. the payoﬀs from various policy sequences are also uniquely pinned down. Intuitively. Consequently.8 Subgame Perfect Versus Markov Perfect Equilibria I have so far focused on Markov perfect equilibria (MPE). there cannot be any SPEs other than the MPE characterized above. This means that the returns to various strategies for the elite are independent of history. such a focus can be restrictive.6) determine the factor demands uniquely in any equilibrium. Loosely speaking. SPEs that are not Markovian will be supported by some type of “history-dependent punishment strategies”. we have: Proposition 57 The MPEs characterized in Propositions 53-56 are the unique SPEs. (18. t = τ 18. 416 . it can be proved that subgame perfect equilibria (SPE) coincide with the MPE.5) and (18. however. then the political equilibrium features PC τm > τ RE for all t. Suppose also that Assumption 15 and Condition (ES) hold and φ > 0. each individual is inﬁnitesimal and makes its economic decisions to maximize proﬁts. MPE are a subset of the SPE. In general.

(18. The problem with holdup is that the elite will be unable to commit to a particular tax rate before middle class producers undertake their investments (taxes will be set after investments). there is a unique MPE where τ m ≡τ ¯ for t = τ all t. if expropriation (or taxation) happens after investments. Using a term from organizational economics.6) still determine factor demands. so in all cases. and in particular. Holdup (lack of commitment to taxes or policies) changes the qualitative implications of the model.14. with the only diﬀerence that τ m and τ e now refer to “expected” taxes. What is diﬀerent is the calculus of the elite in setting taxes. Since. 417 .9 Lack of Commitment–Holdup The models discussed so far featured full commitment to taxes by the elites. Previously. To illustrate this possibility. revenues generated by investments can be ex post captured by others.451: Introduction to Economic Growth 18.2. this corresponds to the situation without any “holdup ”. this eﬀect is absent. they took into account that higher taxes would discourage investment. This lack of commitment will generally increase the amount of taxation and ineﬃciency. in the MPE. the elite will always want HP to tax at the maximum rate. in equilibrium expected and actual taxes coincide. These types of holdup problems are likely to arise when the key investments are long-term. This establishes (proof in the text): HP ≡τ ¯ for Proposition 58 With holdup. there is a unique political equilibrium with τ m t = τ all t. now. so that various policies will be determined and implemented after these investments are made (and sunk). taxes are set after investment decisions. but change the timing of events such that ﬁrst individual producers undertake their investments and then the elite set taxes. As a result. The economic equilibrium is unchanged.5) and (18. I consider the same model as above. Naturally.

τ m t = α and there is positive economic activity by the middle class producers. the most proﬁtable deviation for them is to set The trigger-strategy proﬁle will be an equilibrium as long as (18. In the MPE.5). imagine a situation in which Condition (ES) holds so that with the original timing of events (without holdup). This is naturally very costly for the elite as well since they lose all their tax revenues. the elite set τ m = 1 and the middle class producers invest zero. In this model. Consider the extreme case where τ ¯ = 1. For example. If there is any other action in the history. consider the example where Condition (ES) holds and τ ¯ = 1.19) is greater than or equal 418 . To illustrate this.14. the elite raise no tax revenue from the middle class producers. 1−α (18.19) If. it is no longer true that the MPE is the only SPE. and receive transfers worth φ α(1 − α)(1−α)/α Am λθm . the equilibrium tax is τm t = 1 and the middle class stop producing. since there is room for an implicit agreement between diﬀerent groups whereby the elite (credibly) promise a diﬀerent tax rate than τ ¯. and they will raise φ (1 − α)(1−α)/α Am λθm .5) with τ m = α as long as the history consists of τ m = α and investments have been consistent with (18. Instead. Recall that the history of the game is the complete set of actions taken up to that point. with holdup. With this strategy proﬁle. In contrast. Now without holdup. they deviate at any point. consider the following trigger-strategy combination: the elite always set τ m = α and the middle class producers invest according to (18.20) (18. (1 − β ) (1 − α) τ m = 1. the equilibrium tax rate is τ m t = α. in contrast.451: Introduction to Economic Growth It is clear that this holdup equilibrium is more ineﬃcient than the equilibria characterized above. the elite raise a tax revenue of φα(1 − α)(1−α)/α Am λθm / (1 − α) in every period.

Once investments in technology are made.10 Technology Adoption and Holdup Suppose now that taxes are set before investments. Another interesting question is whether. In particular. there is room for coordinating on a subgame perfect equilibrium supported by an implicit agreement (trigger strategy proﬁle) between the elite and the rest of the society. there exists a subgame perfect equilibrium where τm t = α for all t. and suppose that Assumption 15 and Condition (ES) hold and τ ¯ = 1.451: Introduction to Economic Growth to (18. because typical investments involve longer horizons. the game proceeds as before. middle class agents can invest to increase their productivity. The function Γ is non-negative. An important implication of this result is that in societies where there are greater holdup problems. Instead.20). Therefore we have: Proposition 59 Consider the holdup game. the equilibrium allocations are the same as in the results presented above. continuously diﬀerentiable and convex. Then for β ≥ 1 − α. so the source of holdup in the previous subsection is absent. if they could. suppose that there is a cost Γ (Am ) of investing in productivity Am . This investment is made once and the resulting productivity Am applies forever after. suppose that at time t = 0 before any economic decisions or policy choices are made. 18.2. for example. Since investments in technology are sunk after date t = 0. which requires β ≥ 1 − α.14. The analysis of this case follows closely that of the baseline model. and I simply state the results (without proofs to save space): 419 . the elite would prefer to commit to a tax rate sequence at time t = 0.

then the unique political equilibrium FPM ≡τ ¯ for all t. they would choose lower taxes. Once again. Once the middle class producers have made their technology decisions.21) where τ m is the constant tax rate that they will face in all future periods. Moreover. and it is impossible to create history-dependent punishment strategies to support a tax rate diﬀerent than the static optimum for the elite. For contrast. Condition (ES) does not hold. In the pure ¯}. this is also the unique SPE. To illustrate this. Nevertheless. recall that the equilibrium is τ m = τ RE ≡ min {α. Therefore. the optimization problem of 420 . τ same arguments as before. It is also intuitive that it is the unique SPE. this is not necessarily the allocation that the elite prefer. In fact. As a result. but this restriction saves on notation). suppose that they can commit to a constant tax rate (it is straightforward to show that they will in fact choose a constant tax rate even without this restriction. the ﬁrst-order condition for an interior solution to the middle class producers’ technology choice is: Γ0 (Am ) = α 1 (1 − τ m )1/α 1−β1−α (18. The reason is as follows: in the case of pure factor price manipulation.451: Introduction to Economic Growth Proposition 60 Consider the game with technology adoption and suppose that Assumption 15 holds.14. then they would still choose τ m t = τ That this is the unique MPE is quite straightforward. If the elite could commit to a tax rate sequence at time t = 0. the elite would choose exactly this tax rate even if they could commit at time t = 0. and φ = 0. the MPE is identical to before. let us next consider the pure revenue extraction case with Condition (ES) satisﬁed. if the elite could commit to a tax sequence at features τ m t = τ FPM ≡τ ¯. With the revenue extraction case. time t = 0. the only objective of the elite is to reduce the middle class’ labor demand. there is no history-dependent action left. so they have no interest in increasing the productivity of middle class producers.

If the elite could commit to a tax policy at time t = 0. An important feature is that in contrast to the pure holdup problem where SPE could prevent the additional ineﬃciency (when β ≥ 1 − α. Summarizing this discussion: Proposition 61 Consider the game with technology adoption. In other words. and suppose that Assumption 15 and Condition (ES) hold and φ > 0. The constraint (18.21) as given. then the unique political equilibrium features τ m t = ¯} for all t. the elite would like to commit to a lower tax rate in the future in order to encourage the middle class producers to undertake technological improvements. τ ¯}. If they could.21) as: 1 (1 − τ m )(1−α)/α 1 dAm < 0. This expression can be obtained from (18. The ﬁrst-order condition for an interior solution can be expressed as Am − m 1 − α τm m m dA A + τ =0 α 1 − τm dτ m where dAm /dτ m takes into account the eﬀect of future taxes on technology choice at time t = 0. Their inability to commit to such a tax policy leads to greater ineﬃciency than in the case without technology adoption. the ineﬃciency survives the SPE.14. = − dτ m 1−β1−α Γ00 (Am ) This implies that the solution to this maximization problem satisﬁes τ m = τ T A < τ RE ≡ min {α.21). they will solve: max φτ m (1 − τ m )(1−α)/α Am λθm / (1 − α) subject to (18. with the technology adoption game. they would τ RE ≡ min {α.21) incorporates the fact that (expected) taxes aﬀect technology choice. τ prefer to commit to τ T A < τ RE . The reason is that.451: Introduction to Economic Growth the elite is to maximize tax revenues taking the relationship between taxes and technology as in (18. recall Proposition 59). since middle 421 .

so that there is a credible threat against the elite if they deviate from the promised policies. which concerns direct or indirect factors aﬀecting the productivity of producers. preferences over institutions are derived from preferences over policies and economic allocations. there is greater need for economic institutions to play the role of placing limits on future policies. we can think of this as determining the level of τ ¯. (2) Regulation of technology. I now discuss the implications of these mechanisms for ineﬃcient institutions. but also frequent investments by the middle class. in particular middle class producers. let 422 . 18. there is no possibility of using historydependent punishment strategies. This illustrates the limits of implicit agreements to keep tax rates low. To illustrate the main economic interactions. In terms of the model above. I consider two prototypical economic institutions: (1) Security of property rights. which possesses political power. and consequently.2. Since the elite prefer to implement ineﬃcient policies to transfer resources from the rest of the society (the middle class and the workers) to themselves. to redistribute resources towards themselves. Such agreements not only require a high discount factor (β ≥ 1 − α). the main role of institutions is to provide the framework for the determination of policies.451: Introduction to Economic Growth class producers invest only once at the beginning. they will also prefer ineﬃcient economic institutions that enable and support these ineﬃcient policies.11 Ineﬃcient Economic Institutions The previous analysis shows how ineﬃcient policies emerge out of the desire of the elite. When such implicit agreements fail to prevent the most ineﬃcient policies. there may be constitutional or other limits on the extent of redistributive taxation and/or other policies that reduce proﬁtability of producers’ investments. As pointed out in the introduction. Bearing this in mind.14.

423 . and in particular. suppose that the timing of taxation decision is after the investment decisions (so that there is the holdup problem).451: Introduction to Economic Growth us now discuss the determination of economic institutions in the model presented here.14. i. Therefore. say from τ ¯H to some level in the interval [0. The results are diﬀerent when there are holdup concerns. τ ¯H ].e. putting further restrictions on the taxes can only reduce the elite’s utility. This proposition implies that if economic institutions are decided by the elite (which is the natural benchmark since they are the group with political power). the elite prefer τ ¯=τ ¯H . and start with security of property rights. they will in general choose not to provide additional security of property rights to other producers. and consider the case with revenue extraction and factor price manipulation combined. the underlying economic institutions will support the ineﬃcient policies discussed above. with the only diﬀerence that at time t = 0. thus creating an upper bound on taxes and providing greater security of property rights to the middle class. the elite would like to commit to a lower tax rate than τ ¯H in order to encourage the middle class to undertake greater investments. and this creates a useful role for economic institutions (to limit future taxes): Proposition 63 Consider the game with holdup and suppose Assumptions 15 and 16 hold. In this case. To simplify the discussion. throughout this section. for the rest of the analysis. The proof of this result is immediate. before any decisions are taken. I focus on MPE. since without holdup or technology adoption. To illustrate this. The environment is the same as in the previous section. the elite can reduce τ ¯. The key question is whether the elite would like to ¯<τ ¯H do so.. whether they prefer τ ¯=τ ¯H or τ The next three propositions answer this question: Proposition 62 Without holdup and technology adoption.

The reason is that because of holdup. then as long as τ T A < τ prefer τ ¯ = τ T A. equilibrium taxes are too high even relative to those that the elite would prefer. when we look at SPE. SPE and MPE coincide. for similar reasons. it can exactly commit to the tax rate that maximizes their utility). we see that economic institutions also have and major eﬀect on the environment for technology adoption or more directly the technology choices of producers. the elite will again prefer to change economic institutions to restrict future taxes: Proposition 64 Consider the game with holdup and technology adoption. By manipulating economic institutions. the elite prefer τ The proof is again immediate. the elite may approach their desired policy (in fact. with pure holdup. there may not be a need for changing economic institutions. and the elite can beneﬁt by using economic institutions to manipulate equilibrium taxes. by providing infrastructure or protection of intellectual 424 . τ ¯H . the elite Assumption 15 and Condition (ES) hold and φ > 0. While τ COM maximizes the elite’s utility. then as long as τ COM given by (18. As before. parallel to the results above. in the presence of holdup the MPE involves τ = τ ¯H . so a change in economic institutions is necessary for a credible commitment to a low tax rate (here τ T A ). in the technology adoption game. However.14. Finally. and φ > 0. in the economy with technology adoption discussed above.451: Introduction to Economic Growth Condition (ES) does not hold. and suppose that ¯H . This result shows that the elite may provide additional property rights protection to producers in the presence of holdup problems.16) is less than ¯ = τ COM . since credible implicit promises might play the same role (as long as β ≥ 1 − α as shown in Proposition 59). For example. Turning to the regulation of technology now.

1}. it imposes no costs on the elite). φ = 0. i. so that they generate greater tax revenues. Am = Am (g ). and has no other inﬂuence on payoﬀs (and in particular.e. the elite would like the producers to be as productive as possible.. with Am (1) > Am (0). and when the middle class is more productive. the elite generate greater tax revenues.451: Introduction to Economic Growth property rights. which inﬂuences only the productivity of middle class producers. Consequently. a society may improve the technology available to its producers. Condition (ES) does not hold. the elite may want to block. Assume that the choice of g is made at t = 0 before any other decisions. there is no competition between the elite and the middle class (either in factor markets or in the political arena). Intuitively. take active actions against. The proof follows immediately since g = 1 increases the tax revenues and has no other eﬀect on the elite’s consumption. Therefore the question is: do the elite have an interest in increasing the productivity of the middle class as much as possible? Consider the baseline model. then w = 0 and the the elite always choose g = 1. Conversely. i. the answer is that the elite would like the middle class to have the best technology: Proposition 65 Suppose Assumption 15 and Condition (ES) hold and φ > 0. Will the elite always choose g = 1.. The situation is diﬀerent when the elite wish to manipulate factor prices: Proposition 66 Suppose Assumption 15 holds. in this case. Suppose that there exists a government policy g ∈ {0. or will they try to block technology adoption by the middle class? When the only mechanism at work is revenue extraction. then the elite choose g = 0. the technological improvements of the middle class.e. 425 .14. and τ ¯ < 1. increasing the middle class producers’ productivity.

14. since there is excess labor supply and wages are equal to zero. and their only objective is to reduce the labor demand from the middle class and wages as much as possible. The next proposition shows that a similar eﬀect is in operation when the political power of the elite is in contention. then the elite prefer g = 0. Overall. This makes g = 0 the preferred policy for the elite. the elite will choose economic institutions so as to reduce the productivity of competing (middle class) producers. With τ ¯ < 1.451: Introduction to Economic Growth Once again the proof of this proposition is straightforward. there are no labor market interactions. labor demand from the middle class is high enough to generate positive equilibrium wages. Suppose also that Assumption 15 and Condition (ES) hold and φ = 0. They achieve this by creating an environment that reduces the productivity of middle class producers. introducing economic institutions that limit taxation or put other constraints on policies provides no beneﬁts to the elite. When there are no holdup problems. In this case. this section has demonstrated how the elite’s preferences over policies. when the elite are unable to commit to future taxes (because of holdup problems). Since φ = 0. the elite would like the proﬁts from middle class producers to be as low as possible so as to consolidate their political power. Proposition 67 Consider the economy with political replacement. translate into preferences over ineﬃcient– non-growth enhancing–economic institutions. However. when it is within their power. taxes raise no revenues for the elite. the elite cannot raise any taxes from the middle class since φ = 0. But diﬀerently from the previous proposition. Consequently. Nevertheless. the factor price manipulation mechanism suggests that. and in particular their desire to set ineﬃcient policies. equilibrium taxes may be too high even from the viewpoint of the 426 .

18. An alternative is to have “the dictatorship of the middle class. Hence. the elite may want to discourage or block technological improvements by the middle class. I now brieﬂy discuss the possibility of a switch from the dictatorship of the elite to one of these two alternative regimes. It is clear that whether the dictatorship of the elite or that of middle class is more eﬃcient depends on the relative numbers and productivities of the two groups. and in this case. and they will pursue policies to maximize their own income. this section will ﬁrst characterize the equilibrium under these alternative political institutions. Interestingly. I simplify the discussion by imposing the following assumption: 427 .” a set of political institutions that gave all political power to the elite producers.451: Introduction to Economic Growth elite. Finally. this never happens when the main mechanism leading to ineﬃcient policies is revenue extraction. and whether elite control or democracy is more eﬃcient depends on policies in democracy. then the majority are the workers.14..3 Modeling Political Institutions The above analysis characterized the equilibrium under “the dictatorship of the elite. a system in which the middle class makes the key policy decisions (this could also be a democratic regime with the middle class as the decisive voters).” i. the analysis reveals that the elite may want to use economic institutions to discourage productivity improvements by the middle class. another possibility is democracy in which there is voting over diﬀerent policy combinations.e. for part of the analysis in this subsection. using economic institutions to manipulate future taxes may be beneﬁcial. If θe + θm < 1. Similarly. Instead. Moreover. when factor price manipulation and political consolidation eﬀects are present.

combining revenue extraction and factor price manipulation. (18.451: Introduction to Economic Growth Assumption 17 1 θm = θe < . with the roles reversed.22) 1 + κ (λ. The analog of Assumption 16 in this case is: Assumption 18 A ≥ φ(1 − α) m (1−α)/α θe A m. φ) Proposition 68 Suppose Assumptions 15 and 17 hold. To avoid repetition. θm . it is apparent that the elite equilibrium will be more eﬃcient when Ae and θe are large relative 428 .14. Condition (ES) does not hold. the middle class will tax the elite and will redistribute the proceeds to themselves. and they are in the minority relative to workers. and moreover. θ e Given this assumption.τ ¯ . the political equilibrium is identical to the dictatorship of the elite. a similar proposition to that above immediately follows.1 Dictatorship of the Middle Class With the dictatorship of the middle class.e. then the unique political equilibrium with middle class control features τ e t = τ given by (18..22) for all t. Instead. I will not provide a full analysis. φ) e COM τt = τ ˜ ≡ min . α. let me focus on the case. 2 This assumption ensures that the number of middle class and elite producers is the same. θm . Ttw = Tte = 0. and ˜COM as φ > 0. Comparing this equilibrium to the equilibrium under the dictatorship of the elite. 18. α. the same analysis as above gives their most preferred tax rate as ½ ¾ κ (λ. i.3.

taxes have no eﬀect on wages. More speciﬁcally. then aggregate output is higher with the dictatorship of the elite than the dictatorship of the middle class if Ae > Am and it is higher under the dictatorship of the middle class if Am > Ae . so that workers care about equilibrium wages and transfers. with wt given by (18.3). similar to the case of revenue extraction for the elite above. Tt ..451: Introduction to Economic Growth to Am and θm . each w worker’s consumption is cw t = wt + Tt . and the middle class equilibrium will be more eﬃcient when the opposite is the case. Proposition 69 Suppose Assumptions 15-18 hold.. so the workers will tax at the revenue maximizing rate. This result is stated in the next proposition (proof omitted): 429 .1. 18. Workers will then choose the sequence of policies P∞ t w m w m e {τ e t . ¤ φ £ m (1−α)/α m m m e e (1−α)/α e e e τ t (1 − τ m ) A l θ + τ (1 − τ ) A l θ +R t t t 1−α As before.2 Democracy Under Assumption (A4).. Tt . we obtain that democracy will solve the following maximization problem to determine policies: τ t . when Condition (ES) holds. Intuitively. these taxes are equal) and not on themselves. Tt }t=0.. It is straightforward to see that the workers will always set Ttm = Tte = 0. and have the power to tax the elite and the middle class to redistribute themselves. Substituting for the transfers from (18. τ t . so aggregate output is higher when the group with greater productivity is in power and is spared from distortionary taxation.14.3.7).3) to maximize t=0 β ct .∞ that satisfy (18.τ t max wt + e m with wt given by (18. the group in power imposes taxes on the other group (and since θm = θe . workers are in the majority in democracy.7).

unique political equilibrium with democracy features τ m t = τt = τ Therefore. workers realize that by taxing the marginal group they are reducing their own wages. and τ m t = τ e De If Am < Ae . when Condition (ES) does not hold and wages are positive. if Am > Ae . then the e RE ≡ min {α. 430 . In this case. The same is not the case. we will have τ m will be such that (1 − τ De )1/α Ae = Am or t = 0. Therefore. The intuition is that when Condition (ES) does not hold. and τ t = τ τ De = α and (1 − α)1/α Ae ≥ Ae . however. As a result. τ ¯}. More speciﬁcally. The most interesting implication of this proposition comes from the comparison of the case with and without excess supply. we will have τ e A = Ae or t = 0.451: Introduction to Economic Growth Proposition 70 Suppose Assumption 15 and Condition (ES) hold and φ > 0. when there is no excess supply. taxes always reduce wages more than the revenue they generate because of their distortionary eﬀects. we will have τ e t = 0. Dm will be such that (1 − τ Dm )1/α Am = Ae or τ Dm = α and (1 − α)1/α Am ≥ Ae . workers will only tax the group with the higher marginal productivity. if Am > Ae . in this case democracy is more ineﬃcient than both middle class and elite control. In fact. workers understand that high taxes will depress wages and are therefore less willing to use distortionary taxes. for m m 1/α m example. we have: τm t = α and (1 − α) Proposition 71 Suppose Assumptions 15 and 18 hold and Condition (ES) does not hold.14. democracy taxes both groups of producers and consequently generates more ineﬃciency than the dictatorship of the elite or the middle class. since it imposes taxes on both groups. Then in the unique political equilibrium with democracy. and τ t will be such that (1 − τ t ) 1/α m A ≥ Ae . While in the presence of excess labor supply. it is in general less distortionary than the dictatorship of the middle class or the elite.

political institutions that lead to more ineﬃcient policies will persist even though alternative political institutions leading to better outcomes exist. they 431 . However. since the net present value of the beneﬁt of holding political power often exceeds any transfer that can be made. (i) such promises will not be credible. to relinquish their power to the middle class. To discuss why (and why not). Consequently. the middle class will have no incentive to keep on making such transfers. and once they have political power.3 Ineﬃciency of Political Institutions and Inappropriate Institutions Consider a society where Assumption 18 is satisﬁed and Ae < Am so that middle class control is more productive (i.e. First. (ii) since there are no other. so that the alternative political institutions will not be as eﬃcient in the ﬁrst place..451: Introduction to Economic Growth 18. One possibility is a Coasian deal between the elite and the middle class. This type of solution will run into two diﬃculties. perhaps the elite can relinquish political power and get compensated in return. to compensate the elite. in the same way as preferences over ineﬃcient policies translate into preferences over ineﬃcient economic institutions. Second. For example. Thus. ﬁscal instruments. the elite may relinquish power in return for a promise of future transfers.3. Despite this. let us distinguish between two alternative approaches. such deals are in general not possible. Such a solution is also not possible in general. generates greater output). the middle class will have to impose similar taxes on itself. the desire of the elite to implement ineﬃcient policies also implies that they support political institutions that enable them to pursue these policies. less distortionary. the elite will have no incentive.14. the elite may relinquish power in return for a lump-sum transfer from the middle class. In this case. without some type of compensation.

The above analysis shows that the elite will choose a high tax rate on the middle class.4 below. Imagine an economy in which the elite are in power.14. but then become “inappropriate” and costly for economic activity later. In contrast. Northeastern United States developed as a settler colony.451: Introduction to Economic Growth also lead to preferences towards ineﬃcient political institutions. for example. output will be relatively high. starting in the late 18th century. and few rights for the slaves that made up the majority of the population. Ae is relatively high and Am is relatively small to start with. they lagged behind the United States and many other more democratic societies. approximating a democratic society with signiﬁcant political power in the hands of smallholders and a broader set of producers. φ is small. and almost certainly richer and more productive than the Northeastern United States. because the elite will undertake the right investments themselves. The Caribbean colonies were clear examples of societies controlled by a narrow elite. with political power in the monopoly of plantation owners. which took advantage of new investment opportunities.3. Nevertheless. This question might be motivated. and the distortion on the middle class will be relatively small since Am is small. by the contrast of the Northeastern United States and the Caribbean colonies between the 17th and 19th centuries. While in both the 17th and 18th centuries. Another interesting question is whether a given set of economic institutions might be “appropriate” for a while. I will discuss how political institutions can change from the “ground-up” in Section 18. The baseline model used above suggests a simple explanation along these lines. the Caribbean societies were among the richest places in the world. This raises the question as to whether the same political and economic institutions that encouraged the planters to invest and generate high output in the 17th and early 18th centuries then became a barrier to further growth. 432 . particularly in industry and commerce. Condition (ES) does not hold.

De jure political power is determined by political institutions. This is reminiscent of the planter elite controlling the economy in the Caribbean. we need an equilibrium model of institutional change. if at some point the environment changes so that Am increases substantially relative to Ae . that did not generate much distortion or may have even encouraged growth) later caused the society to fall substantially behind other economies. 433 . still in power. the dictatorship of the elite may generate greater income per capita than an alternative society under the dictatorship of the middle class. Another society where the middle class have political power will now generate signiﬁcantly greater output. This simple example illustrates how institutions that were initially “appropriate” (i. De facto political power. I will conceptualize institutional change as resulting from the interplay of de jure and de facto political power. but now these policies have become very costly because they distort the investments of the more productive group. The elite.3. However. de jure political power is in the hands of the elite. The simplest example of de facto political power is when a group manages to organize itself and poses a military challenge to an existing regime or threatens it with a revolution. will continue to impose high taxes on the middle class.. 18. It is ﬁrst useful to draw a distinction between de jure and de facto political power. since the political institutions give them the right to set taxes and determine the economic institutions. In the baseline model.4 Institutional Change and Persistence To develop a better understanding for why ineﬃcient institutions emerge and persist. the situation changes radically.e. did not feature so far in the model (except in the discussion of political consolidation). which comes from other sources.451: Introduction to Economic Growth Consequently.14. I now brieﬂy discuss such a model.

I focus on MPE. M } denotes whether the elite or the middle class are in control. Also denote the state at time t by the tuple (Pt . the middle class has no de facto political power. which is costly. each middle class agent incurs a cost of ψ in the process. In each period. st ). I assume that following a violent overthrow. the elite need to respond in some way. (ii) they can give up power. there will never be any further institutional change. since letting the middle class overthrow the existing regime is excessively costly for them. which were characterized above. In a MPE.. I assume that repression costs µ for the elite as a whole. in the process. to simplify the discussion. with probability q the middle class solve the collective action problem among its members and gather suﬃcient de facto political power to overthrow the existing regime and to monopolize political power (establish a dictatorship of the middle class). (iii) they can use repression. but manages to prevent the regime from falling to the middle class.451: Introduction to Economic Growth Imagine a society described by the baseline model above where de jure political power is initially in the hands of the elite. is an absorbing state and once the middle class comes to power. I assume throughout that Condition (ES) is satisﬁed. violently overthrowing the existing regime is still costly. When the middle class amass de facto political power. Throughout this section. etc. the elite will set the policies that maximize their utility. Let us assume that the dictatorship of the middle class. where Pt ∈ {E. strategies are only a function of the state st . Moreover. such as reducing taxes on the middle class. and in particular. The elite can respond in three diﬀerent ways: (i) they can make temporary concessions. so when st = L. Moreover. and st ∈ {H. if established. So the interesting actions take place in the state st = H . In particular. so that the 434 . With probability 1 − q . the elite receive zero utility.14. the elite are harmed substantially. However. L} denotes the level of threat (high or low) against the regime controlled by the elite.

Similarly. and the second is the distribution from the revenue obtained by taxing the elite and from natural resources. during periods of low threat. we have V m ¡ ¡ ¢ ¢ π m (0) + T e τ RE + R /θm . V (M ) = 1−β the no threat state: ¡ ¢ V m (E. H ) .451: Introduction to Economic Growth main motive for ineﬃcient policy is revenue extraction.25) (18. αAm λ/ (1 − α). Let us ﬁrst calculate the value of a middle class agent when the middle class is in power. What happens when st = H ? As noted above. (18. L) + βqV m (E. in the MPE. and redistribute all the revenue to themselves.14. The low threat state recurs with probability 1 − q . the elite will follow their most preferred policy.11). (M ) = 1−β (18. Then.24) What about the dictatorship of the elite? Let us write this value recursively starting in This expression incorporates the fact that. the above analysis shows that they will not tax themselves. there are 435 . To write the resulting value function. and π j (τ ) ≡ α (1 − τ )1/α Aj λ/ (1 − α) as the proﬁt of a producer in group j facing the tax rate τ . τ m = τ RE and T m = 0. let us introduce the following notation: T j (τ ) ≡ φτ (1 − τ )(1−α)/α Aj θj λ/ (1 − α) as the tax revenue raised from group j at the tax rate τ . The ﬁrst term in the numerator is their own revenues. the value of an elite producer in this case is ¡ ¢ π e τ RE e . using M to indicate a value function under the dictatorship of the middle class.23) where τ RE is given by (18. set a tax of τ e = τ RE on the elite in every period. The term 1 − β provides the net present discounted value of this stream of revenues. L) = π m τ RE + β (1 − q) V m (E. Since Condition (ES) is satisﬁed.

H ) (18. H ). i.27) and (18.28) makes it clear. H ).25). Therefore. When this constraint holds.25). L) is given by expression (18. ¯ m (E. V (18. Note that the suﬃcient to satisfy the middle class. the condition for concessions within the given political regime to prevent V action by the middle class is simply ¯ m (E..26) where recall that ψ is the cost of regime change for the middle class. with V ¯ m (E. H ) replacing V m (E.28) V (1 − β ) This is the maximum credible utility that the elite can promise the middle class within the existing regime. H ) = π m (0) + T e τ RE + R /θm + β (1 − q ) V V right hand side. L) + βqV ¯ m (E. τ e = τ RE . τ m = 0. The reason why they cannot give them greater utility is because of commitment problems. we simply need to calculate V best concession that the elite can do is to adopt a policy that is most favorable for the middle ¡ ¡ ¢ ¢ class.27) ¯ m (E. Therefore. Combining (18. Let us ﬁrst start by investigating whether the elite can prevent a switch of political power by making concessions in the high threat state. we obtain: ¡ e ¡ RE ¢ ¡ RE ¢ £ m ¢ m¤ m (0) + T τ + (1 − β (1 − q )) π τ + R /θ β (1 − q ) π m ¯ (E. Even if they promise to make further transfers or not tax 436 . the elite could make suﬃcient concessions to keep the middle class happy within the existing regime. For this purpose.451: Introduction to Economic Growth three possibilities. to determine whether concessions within the dictatorship of the elite will be ¯ m (E. Then. and T m = T e τ RE + R /θm .e. As (18. H ) = (18. let us denote the highest possible value to the middle class under the dictatorship of the elite by ¯ m (E.14. the elite transfer resources to the middle class only in the state st = H . H ) on the where V ¡ ¡ ¢ ¢ ¯ m (E. H ) ≥ V m (M ) − ψ.

They will clearly not follow the most preferable policy for the middle class. then the elite can prevent a violent overthrow by making concessions within the existing regime. Denote V e (O.29) ˆ e (E. Such repression is always eﬀective. Nevertheless. when the state st = L arrives. the elite may not necessarily prefer such concessions. H ) = V Whether the elite will make these concessions or not then depends on the values of other options available to them. ³ ´ ˆm . L)+ βqV e (O. By standard arguments. T V m (M ) − ψ. T ˆe such that V m (E. when using the repression strategy. These two expressions incorporate 437 the fact that. H ) = ¡ ¡ ¢ ¢ π e (0)+ T m τ RE + R /θm − µ + β (1 − q ) V e (O. H ). st ) the value function to the elite it uses repression and the state is st . τ Instead. H ) = ˆm . we can obtain this value by writing the following standard recursive formulae: V e (O. If given this expression. Another alternative is the use of repression whenever there is a threat from the middle class.e. L) = π e (0)+ ¡ m ¡ RE ¢ ¢ T τ + R /θm + β (1 − q ) V e (O. so the only cost of this strategy for the elite is the cost they incur in the use of repression. we ﬁrst need to determine the exact concessions that the elite will make. the elite will choose a policy combination τ ˆe . by similar arguments. µ. the elite will always choose their most for .26) is satisﬁed. they will make the middle class just indiﬀerent between overthrowing the regime or accepting the concessions. given by: i h ¡ m ¡ RE ¢ ¢ e¢ ¡ e e e e ˆ τ )+T τ + R /θ + (1 − β (1 − q )) π (ˆ β (1 − q ) π (0) + T (1 − β ) (18.. since this will give more than suﬃcient utility to prevent an overthrow.451: Introduction to Economic Growth them in the state st = L. H ) and V e (O. (18. L) + βqV e (O.14. and in the MPE. i. To investigate this issue. The value of such concessions to the elite is. these promises will not be credible (they cannot commit to them). the elite will choose their most preferred policy of taxing the middle class and transferring the resources to themselves.

so comparing these values essentially amounts to comparing nonlinear functions of the underlying parameters. H ) = .30) ˆ e (E. regime change will only happen when (18.τ ˆ . for similar reasons. H ). we obtain the following proposition: (18. i. or if (18.24). H ). H ). If (18. Combining these two equations. 1−β V e (O. Note that all of the values here are simple functions of parameters. H ). H ). since in the latter case they only given by (18. Evidently..451: Introduction to Economic Growth preferred policy combination. and obtain V e (M ) as ˆ e (E.26) holds and V always remains the dictatorship of the elite. τ e = 0 and T m = 0.e. If (18. H ) = V (M ) − ψ. Putting all these pieces together and assuming for convenience that when indiﬀerent the elite opt against repression. they adopt the policy τ ˆ . it needs to be the case that V Proposition 72 Consider the above environment with potential regime change and suppose that Condition (ES) holds.26) does not hold. for regime change to take place. V e (M ) is less than V make concessions (in fact limited concessions) with probability q. the elite set their most preferred policy of τ m = τ RE . then the regime always remains the dictatorship of the elite.26) holds but V V e (O. Finally. H ). the third alternative for the elite is to allow regime change.26) does not hold and V e (M ) < 2.14. In addition. The elite always set 438 . ˆ e (E. we need V e (M ) ≥ V e (O. and will use repression when st = H to defend their regime. When st = L.T . H ) ≥ V e (O. for the elite to prefer concessions. the elite make concessions ´ ³ m e ˆm ˆe m m ˆ suﬃcient to ensure V (E. H ) ≥ Consequently. Then there are three diﬀerent types of political equilibria: ˆ e (E. and when st = H .T . we obtain: ¡ m ¡ RE ¢ ¢ m e (0) + T τ + R /θ − (1 − β (1 − q )) µ π V e (O. Therefore. in the unique equilibrium the regime 1. H ) < V e (O.

since institutional change gives de jure political power and thus the right to set ﬁscal policy in the future to the middle class). the elite voluntarily pass political control to the middle class. τ e = 0 and T m = 0. The comparative statics of regime change are also interesting. Moreover: ¡ RE ¢ ¡ m ¡ RE ¢ ¢ m e e (0) − π τ + T τ + R /θ − (1 − β (1 − q)) µ π V e (O.. The most interesting case is 3. τ e = 0 and T m = 0. and when st = H . where there is equilibrium institutional change as a result of the elite voluntarily relinquishing political control. and acts as a credible promise of future policies that favor the middle class (the promise is credible. If (18. institutional change is more likely. they use repression against the middle class. and when st = H . then there is equilibrium institutional change. First.451: Introduction to Economic Growth their most preferred policy of τ m = τ RE . This discussion highlights that institutional change has two requirements: (i) that concessions within the existing regime are not suﬃcient to appease the middle class. H ) − V e (M ) = 1−β is increasing in R and φ. This proposition illustrates how various diﬀerent institutional equilibria can arise. H ). Why would the elite give up their dictatorship? The reason is the de facto political power of the middle class. µ is higher. The elite then prevent such a violent overthrow by changing political institutions to transfer de jure political power to the middle class. which threatens the elite with a violent overthrow–an outcome worse than the dictatorship of the middle class. (ii) that repression is suﬃciently costly for the elites to accept regime change. the elite set their most preferred policy of τ m = τ RE . 3.26) does not hold and V e (M ) ≥ V e (O.e. i. When st = L. This transfer exploits the role of political institutions as a commitment device (a commitment to a diﬀerent distribution of de jure political power). when repression is more costly. so that there are greater rents from 439 . This implies that when R is high.14.

even the possibility of collective action by the middle class is not suﬃcient. since the elite can use costly methods to defend the existing regime.26) more likely to be violated. This analysis also illustrates the conditions for institutional persistence. This implies that. which will in turn be the case when there is signiﬁcant distributional conﬂict between the elite and the middle class. Similarly. are high. i. since greater state capacity enables greater tax revenues in the future. which typically leads to less distortionary policies. high φ) or because rents from natural resources..2. Therefore. However. greater φ. as already suggested by the results in subsection 18. R. has the same impact on institutional equilibrium.e.14. greater state capacity. when alternative institutional arrangements are costly for those who currently hold political power and have the means to use force to maintain the existing political institutions. making it more diﬃcult for the elite to use concessions to appease the middle class. and the elite now prefer to use repression rather than allowing institutional change. since they also make (18. the success of the middle class in solving their collective action problem and amassing de facto political power. Persistence is the natural course of things and something unusual.e. when the trade-oﬀ for the elite is between repression and institutional change. greater R and φ make repression more likely. Therefore. increases in R or φ do not make institutional change unambiguously less likely. 440 . creates the platform for institutional change. H ) becomes less likely. Therefore. a set of political institutions will persist when political stakes are high. for example because tax revenues are important (i. which corresponds to greater state capacity.451: Introduction to Economic Growth natural resources. institutions will be more persistent when the elite are unwilling to give up the right to determine policies in the future. while when the trade-oﬀ is between concessions and institutional change.7.. they may encourage institutional change. Nevertheless. V e (M ) ≥ V e (O. also increases political stakes and makes the use of repression by the elite the more likely.

14. When the ability of the middle class to solve their collective action problem is endogenous (as in the model used above to illustrate the political consolidation eﬀect). for example because tax revenues or rents from natural resources are high. in turn. reducing political stakes. suppose that the probability q that the middle class will be able to pose an eﬀective threat to the regime is endogenous and depends on the proﬁts of the middle class. When political power is very valuable. increase institutional persistence by making it more diﬃcult for the middle class to solve their collective action problem and mount challenges against the dictatorship of the elite. In this case. These higher taxes will. This implies that control of ﬁscal policy will generate only limited gains.451: Introduction to Economic Growth The model also suggests the possibility of interesting interactions between economic forces and institutional equilibria. The ﬁrst is an interaction between economic and political institutions. there will be a further interaction between economics and politics. 441 . In particular. the elite will wish to “overtax” the middle class to impoverish them and to reduce their political power. institutional persistence might be more of an issue in societies where economic institutions enable those with political power to capture greater rents. and the elite will have less reason to use repression in order to defend the existing regime. Consequently. the greater the threat from them in the future. Suppose that economic institutions impose a low τ ¯. this time between ineﬃcient policies and institutional persistence. This suggests another interesting interaction. the elite realize that the richer are the middle class.

HỒI KÝ BẢO ĐẠI TRONG GIAI ĐOẠN LÀM CỐ VẤN TỐI CAO CHO HỒ CHÍ MINH

Hồi-ký của Bà Ngô Đình Nhu

LỬA TRONG CÁI TRÍ

Process Values, International Law, And Justice

An Introduction to the Philosophy of Law

Critique of Hegel’s Philosophy of Right

Reason and Natural Law

The Natural Law in the Buddhist Tradition

The Superiority or Integrity of Natural Law for Our Time

Natural Law - The Modern Tradition

The Concept of Law and Justice in Ancient Egypt

State and Government

Trung Quốc của Mao Trạch Đông

Teaching the Vietnam War

Tet In Hue

Triết Học Chính Trị - Xã Hội Phật Giáo

Nguyễn Ước Triết học chính trị-Ch08

Human Rights in International Relations-3rd_edition

Binh Pháp Tôn Tử và 36 kế sách

Ma Chiến Hữu

White Paper on the Hoang Sa (Paracel) and Truong Sa (Spratly) Islands

Khai Sáng Kỷ Nguyên Thứ Hai

Good Muslim, Bad Muslim

Dao Thien Chua va Chu Nghia Thuc Dan phan 1.

The Lessons of the Vietnam War

- Read and print without ads
- Download to keep your version
- Edit, email or read offline

Acemoglu Solution

Jones - Introduction to Economic Growth

Economic Growth

Solucionario - Introduction to Economic Growth

Economic growth 3rd edition

Solutions to Exercises in Introduction to Economic Growth-Charles Jones

Advanced Macroeconomics (Solutions) - David Romer

Introduction to Statistics and Econometrics

microeconomic theory

Macroeconomics

Microeconometrics for Policy, Program and Treatment Effects

Bayesian Econometric Methods

Economics

Acemoglu

Economics - Making Sense of the Modern Economy

Introduction to the Mathematical and Statistical Foundations of Econometric s

MIcroeconomics Book

Jehle and Reny Solutions

Robert Boyer - The Future of Economic Growth

Mas Colell, Whinston & Green - Microeconomic Theory

Macroeconomics

Mas-Colell, A., Whinston, M. D. Et Green, J. R. (1995). Microeconomic Theory

Solutions to Introduction to Economic Growth

Handbook of Mathematical Economics-sAyYied

Introduction to Economic Growth - Jones

Mathematical Statistics - Knight

Globalization and International trade

calculus economics

Bayesian Econometrics

STOKEY & LUCAS - Recursive Methods in Economic Dynamics

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

CANCEL

OK

You've been reading!

NO, THANKS

OK

scribd