FOUNDATIONS OF INTERNATIONAL MACROECONOMICS Maurice Obstfeld Kenneth Rogoff ‘The MIT Press Cambridge, Massachusetts London, England© 1996 Massachusetts Institute of Technology All rights reserved. No part of this publication may be reproduced in any form by any electronic or ‘mechanical means (including phococopying, recording, or information storage and rettieval) without ‘permission in writing from the publisher. ‘This book was set in Times Roman by Windlall Software using ZZTEX and was printed and bound in the United States of America, Library of Congress Cataloging-in-Publication Data Obstfeld, Maurice. ‘Foundations of international macroeconomics / Maurice Obstfeld, Kenneth Rogoft. p. em, Includes bibliographical references and index. ISBN 0-262-15047-6 1, Intemational economic relations, 1. Rogoff, Kenneth, Iti HFI259.027 1996 397—de20 96-27824 cPContents Preface Introduction Intertemporal Trade and the Current Account Balance La 12 13 14 15 1A A Small Two-Period Endowment Economy Application: Consumption Smoothing in the Second Millennium B.C. ‘The Role of Investment Box 1.1 Nominal versus Real Current Accounts ‘A Two-Region World Economy Application: War and the Current Account Application: Investment Productivity and World Real Interest Rates in the 1980s ‘Taxation of Foreign Borrowing and Lending, International Labor Movements Application: Energy Prices, Global Saving, and Real Interest Rates Stability and the Marshall-Lerner Condition Exercises Dynamics of Small Open Economies 24 2.2 23 2.4 25 2A 2B ‘A Small Economy with Many Periods Application: When Is a Country Bankrupt? Dynamics of the Current Account Box 2.1 Japan's 1923 Earthquake ‘A Stochastic Current Account Model Application: Deaton’s Paradox Application: The Relative Impact of Productivity Shocks on Investment and the Current Account ‘Consumer Durables and the Current Account Firms, the Labor Market, and Investment ‘Trend Productivity Growth, Saving, and Investment: A Detailed Example Speculative Asset Price Bubbles, Ponzi Games, and Transversality Conditions Exercises ‘The Life Cycle, Tax Policy, and the Current Account 3 3.2 Government Budget Policy in the Absence of Overlapping Generations Government Budget Deficits in an Overlapping Generations Model Box 3.1 Generational Accounting 4 18 23 25 35 42 45 51 33 34 59 BAS 6 19 87 116 121 124 129 130 133 142vu Contents 3A Application Do Government Budget Deficits Cause Current Account Deficits? Application Overlapping Generations and Econometric Tests of the Buler Equation Output Fluctuations, Demographics, and the Life Cycle Application How Are Saving and Growth Related? Investment and Growth Application Feldstem and Hortoka’s Saving-Investment Puzzle Aggregate and Intergenerational Gains from Trade Pubhe Debt and the World Interest Rate Apphation Government Debt and World Interest Rates since 1970 Integraung the Overlapping Generations and Representative-Consumer Models Dynamic Inefficiency Exercises ‘The Real Exchange Rate and the Terms of Trade 41 International Price Levels and the Real Exchange Rate 42 — The Price of Nontraded Goods with Mobile Capital Box 41 Empirical Evidence on the Law of One Price Application Sectoral Productivity Differentials and the Relative Prices of Nontradables in Industrial Countnes Application Productivity Growth and Real Exchange Rates 43> Consumption and Production in the Long Run 44 — Consumption Dynamucs, the Price Level, and the Real Interest Rate 45 — The Terms of Trade in a Dynamic Ricardian Model Box 42 The Transfer Effect for Industral Countries 4A Endogenous Labor Supply, Revisited 4B Costly Capital Mobility and Short-Run Relative Price Adjustment Exercises Uncertainty and International Financial Markets 51 Trade across Random States of Nature The Small-Country Case Box 51 Lloyd's of London and the Custom Market for Risks 52 A Global Model Application Comparing International Consumption and, Output Correlations Box 52 Are Markets More Complete within than among Countries? International Portfolio Diversification 144 146 147 152 156 161 164 167 173 174 191 195 199 200 202 203 209 212 216 225 235 256 258 260 264 269 270 274 285 290 295 300Contents 34 56 5c sD Application: International Portfolio Diversification and the Home Bias Puzzle Asset Pricing Application: The Equity Premium Puzzle over the Very Long Run Application: GDP-Linked Securities and Estimates of V" ‘The Role of Nontradables Application: Nontradability and International Consumption Correlations Application: How Large Are the Gains from International Risk Sharing? A Model of Intragenerational Risk Sharing Box 5.3 A Test of Complete Markets Based on Consumption Divergence within Age Cohorts Spanning and Completeness ‘Comparative Advantage, the Current Account, and Gross Asset Purchases: A Simple Example An Infinite-Horizon Complete-Markets Model Ongoing Securities Trade and Dynamic Consistency Exercises . Imperfections in International Capital Markets 61 6A 6B Sovereign Risk Box 6.1 Sovereign Immunity and Creditor Sanctions Application: How Costly Is Exclusion from, World Insurance Markets? Application: How Have Prior Defaults Affected Countries’ Borrowing Terms? Sovereign Risk and Investment Application: Debt Buybacks in Practice Risk Sharing with Hidden Information Moral Hazard in International Lending Application: Financing Constraints and Investment Recontracting Sovereign Debt Repayments Risk Sharing with Default Risk and Saving Exercises Global Linkages and Economic Growth 1 12 The Neoclassical Growth Model Box 7.1 Capital-Output Ratios since World War I International Convergence 304 306 314 317 319 323 329 332 335 335 337 340 343 345 349 349 352 366 378 379 309 401 407 417 419 422 425 429 430 435 454.Contents 74 TA “7B Application: Productivity Convergence, 1870-1979: ‘The Baumol-De Long-Romer Debate Application: Public Capital Accumulation and Convergence Endogenous Growth Application: Can Capital Deepening Be an Engine of Sustained High Growth Rates: Evidence from Fast-Growing East Asia Application: Population Size and Growth Stochastic Neoclassical Growth Models Continuous-Time Growth Models as Limits of Discrete-Time Models A Simple Stochastic Overlapping Generations Model with Two-Period Lives Exercises Money and Exchange Rates under Flexible Prices 81 82 83 84 85 8.6 87 8A 8B Assumptions on the Nature of Money ‘The Cagan Model of Money and Prices Box 8.1 How Important Is Seignorage? Monetary Exchange Rate Models with Maximizing Individuals Application: Testing for Speculative Bubbles Nominal Exchange Rate Regimes Box 8.2. Growing Use of the Dollar Abroad ‘Target Zones for Exchange Rates Speculative Attacks on a Target Zone A Stochastic Global General Equilibrium Model with Nominal Assets A Two-Country Cash-in-Advance Model ‘The Mechanics of Foreign-Exchange Intervention Exercises Nominal Price Rigidities: Empirical Facts and Basic Open-Economy Models 9.1 92 93 94 95 Sticky Domestic Goods Prices and Exchange Rates ‘The Mundell-Fleming-Dombusch Model Empirical Evidence on Sticky-Price Exchange-Rate Models Choice of the Exchange-Rate Regime Models of Credibility in Monetary Policy Application: Central Bank Independence and Inflation Application: Openness and Inflation Exercises 457 467 473 481 492 496 508 510 512 513 514 515 527 530 546 554 555 569 576 579 595 597 599 605 621 631 634 646 653 65710 Contents Sticky-Price Models of Output, the Exchange Rate, and the Current Account 10.1. A Two-Country General Equilibrium Model of International Monetary Policy Transmission Box 10.1 More Empirical Evidence on Sticky Prices Box 10.2 The Role of Imperfect Competition in Business Cycles 10.2 Imperfect Competition and Preset Prices for Nontradables: Overshooting Revisited Application: Wealth Effects and the Real Exchange Rate 10.3 Government Spending and Productivity Shocks 10.4 Nominal Wage Rigidities Application: Pricing to Market and Exchange-Rate Pass-Through Exercises Supplements to Chapter 2 A Methods of Intertemporal Optimization B A Model with Intertemporally Nonadditive Preferences C Solving Systems of Linear Difference Equations Supplement to Chapter 5 A Multiperiod Portfolio Selection Supplement to Chapter 8 A Continuous-Time Maximization and the Maximum Principle References Notation Guide and Symbol Glossary Author Index Subject Index 659 660 676 689 689 694 696 706 m1 13 115 15 m2 726 742 742 745 745 155 781 789 195Preface ‘The goal of this book is to develop a broad coherent framework for thinking about all of the fundamental problems in international macroeconomies. We are unaware of any previous attempt to accomplish this goal using modern analytical methods. This project arose from our concern over the widespread view that open- economy macroeconomics (also known as international finance) has become an extremely scattered field, defined more by the people who consider themselves working in it than by any set of unifying ideas. The problem was brought home forcefully in an informal survey conducted by Professor Alan Deardorff of the University of Michigan. In 1990, Deardorff gathered graduate international finance reading lists from eight top economics departments, hoping to find a consensus on which readings should be deemed most essential. To his surprise, he found strikingly little agreement, with only one article appearing on more than half the reading lists. Such idiosyncrasies will come as no surprise to those who have stud- ied international macroeconomics in graduate school. Whereas most students find the major issues compelling, @ constant complaint is that the material lacks a uni- fied theme and that radical gear shifts are needed to move from one topic to the next. Some senior economists at leading departments view the state of the field with such dismay that they teach mostly from articles written in the 1960s and 1970s, (or at least articles that should have been written then), disdaining the modern lit- erature for its alleged lack of policy relevance. Unfortunately, that approach is not productive for graduate students who need a vision of the field's present and its fu- ture. While the classic literature from twenty and thirty years ago can be admired for articulating and attempting to formalize a number of central policy issues, its limitations are many. The classic approach lacks the microfoundations needed for internal consistency, it fails to deal with dynamics in any coherent way, and its vi- sion of capital market integration may generously be described as narrow. Perhaps most importantly, the older literature simply doesn’t deal sensibly with many ques- tions that are central to today’s policy world, such as current accounts, government budget deficits, speculative attacks, and the implications of the expanding global markets for securities and derivatives. ‘Our goal in this book is to show that one can address virtually all the core issues in international finance within a systematic modern approach that pays attention to the nuances of microfoundations without squeezing all life out of this fascinating topic, We knew at the outset that our project was an ambitious one, and we plunged ahead despite having no idea how we would cross some of the vast chasms in the field. We realized that many areas would have to be substantially reformulated, and frankly hoped that some of the gaps would be fortuitously filled in by other researchers over the life of the project. This did happen, but only in small doses. ‘One important missing link we tackled involved bridging the gap between modern fiexible-price maximizing models of the current account and the older sticky-pricexiv Preface Keynesian models that still dominate policy debate. The approach we took is the subject of Chapter 10, The reader will have to judge for himself or herself how successful we have been in our goal of offering a unified framework, Whereas we do not pretend that any one model can encompass all issues, you should find that most models and themes we develop appear several times in the book, sometimes in seemingly widely disparate contexts. The authors of any manuscript this length must stand ready to be accused of self-indulgence. We trust the reader will agree that part of the length is explained by our deliberate attempt to make the chapters easy to follow. We fill in many intermediate steps to guide students through the central models, hoping to save them the frustrations we ourselves often feel when trying to follow the mathematics of elegant but tersely expressed articles. We also devote a considerable amount of attention to providing empirical motivation for the main concepts and them: introduce. Our notion of empirical evidence is a broad one. If there exists a careful econometric study that casts light on a question we ask, we try to discuss it, But if the best evidence is a suggestive diagram or a relatively unsophisticated regression, we still offer it to illustrate our concepts if not to test them. How did we ultimately decide on the specific topies to be covered? Two cri- teria strongly influenced our choices. We biased our selection toward topics that we felt we could integrate with our central microfoundations approach and toward topics for which there seemed to exist an interesting empirical literature. With little exception, nothing appears in the book that is not based on explicit microfounda- tions. But we have not been so dogmatic as to exclude the most influential ad hoc aggregative models completely. You will find Solow’s growth model in Chapter 7, ‘Cagan’s model of hyperinflation in Chapter 8, and Dombusch’s classic extension of the Mundell-Fleming open-economy [S-LM model in Chapter 9. In each case, however, we go on to show how similar ideas could be expressed in models with microfoundations. Though we try to be up to date, we have also tried not to be slaves to fashion. Target zone exchange rate models are great fun and have gener~ ated a tidal wave of papers; we provide an introduction in Chapter 8. But we think the relatively small number of pages we devote to the topic is the right proportion for this book, and we make no apologies for not including an entire chapter on the subject. Other topics such as European Monetary Union (on which both of us have written) are of great current interest, but the relevant academic literature consists largely of narratives and applications of seasoned theory. We examine optimal cur- rency areas in Chapter 9, but not nearly as extensively as others might have chosen to do, We genuinely regret not having more space left over to deal with questions of political economy. We do cover these to some extent in our discussion of inter- national debt and exchange rate policy. But again much of the literature uses rather stylized descriptive models that did not easily fit into our general approach, and so we leave that important subject for later work. In addition to the book’s core ma- terial in Chapters 1 through 10, we have also included a number of Supplementsxv Preface (five in all) covering the mathematical methods applied in the main exposition, ranging from difference equations to dynamic programming. We were especially concerned to provide useful primers for material that is “folk wisdom” in the field but for which one often cannot find an easily accessible reference. Some inwardly oriented macroeconomists may look at our coverage and feel that we have somehow usurped standard topics from macro such as growth or inflation. Our simple answer to them is that we do not consider “closed-economy” macroe- conomics to be an independent field. We do not see how anyone can. The United States now trades extensively, and even Albania has opened up to the outside world. There are no closed economies, and there are no virtually closed economies. There are only open national economies and the global economy. Macroeconomics is not the only field our analysis crosses. We freely draw on finance and interna- tional trade theory throughout. However, we do not necessarily expect our readers to be conversant with all of these topics, and we generally develop any necessary concepts from the ground up. Finally, the question may arise as to how we intend that this book be used. Is it a treatise? Is it for the core of a graduate international finance course? Or is it aimed at first-year graduate macroeconomics? The answer, we admit, is all three, The book contains a great deal of new research, most notably in Chapter 4, where we present a dynamic version of the classic Dornbusch-Fischer Samuelson continuum ‘model, and in Chapter 10, where we develop a sticky-price dynamic model that fi- nally updates the Mundell-Fleming-Dornbusch model, We also develop a number of useful diagrams and analytical devices, including microfoundations for the clas- sic Metzler saving-investment diagram in Chapter 1, and a “GDP-GNP” diagram for analyzing nontradable goods production under capital mobility in Chapter 4. In Chapter 5, we demonstrate the remarkable analogy between modern finance models incorporating nonexpected utility preferences and standard open-economy intertemporal models that allow for multiple consumption goods. Chapter 6 on in- ternational debt contains a number of new results, for example, on the relationship between investment and foreign debt levels. In Chapter 8, we show how to develop a simple tractable general equilibrium model for nominal asset pricing that is more plausible empirically than the alternatives in the existing literature. Second-year graduate international finance students are certainly an important constituency. We appreciate that this book cannot easily be covered in a semester, but one can conveniently choose parts of it as the basis for a course and use read- ings to cover other recent research developments. Second-year students may be able to skim fairly quickly past some parts of Chapters 1 through 3, provided this, material was adequately covered in their first-year macroeconomics course. As for a first-year macro course, our coverage is fairly comprehensive, although naturally the choice of topics is biased toward open-economy issues. We have endeavored to ‘make most of the chapters relatively self-contained and, except perhaps for Chap- ters 4 and 10, most material in the other chapters should be accessible to first-yearPreface graduate students, Sections making the greatest technical demands are marked with asterisks (*), and the book is written so that these can be skipped without break- ing the flow of the development. We hope that both first- and second-year students will appreciate our efforts at working through the models algorithmically, provid- ing many intermediate steps along the way. Apologies: Although our book includes extensive citations, space limitations simply make it impossible to provide the intellectual history behind every idea There are numerous excellent surveys (many of which we reference) that are more suitable for this purpose. We also regret (and accept full responsibility for) the hopefully small number of erroneous attributions, equation typos, and other mis- takes that may have crept into the manuscript. We owe thanks to a large number of individuals: + Rudiger Dombusch, from whom we both took our first course in open-economy macroeconomics, and who was a major influence on our decisions to pursue re~ search in the field. + Our graduate students, for helping us think through the material, for assem- bling data, and for catching numerous mistakes. In particular we thank Geun Mee Ahn, Brian Doyle, Cristian Echeverria, Fabio Ghironi, Harald Hau, Min Hwang, Matthew Jones, Greg Linden, Sydney Ludvigson, Ilian Mihov, Lorrie Mitchell, Giovanni Olivei, Giovanni Peri, Don Redl, Peter Simon, Cedric Tille, Susanne Trimbath, Clara Wang, Keong Woo, and Ning Zhang. * Reviewers and others who offered helpful comments, including Henning Bohn, Jeremy Bulow, Richard Clarida, Mick Devereux, Charles Engel, Mark Gertler, Michael Huggett, Peter Kenen, Michael W. Klein, Karen Lewis, Enrique Mendoza, Gian-Maria Milesi Ferretti, Jaime Marquez, Paolo Pesenti, Assaf Razin, Gregor Smith, Federico Sturzenegger, Linda Tesar, and Aaron Tornell. In this regard we are especially grateful to Kiminori Matsuyama, who provided valuable criticisms of early drafts of Chapters 1 through 3 and 9, and to Reuven Glick, who read through the entire manuscript. * Our colleagues, including Ben Bernanke and Michael Woodford, who test drove some of the chapters in first-year graduate macro at Princeton, and Luisa Lamber- tini, who used several draft chapters for graduate international finance at UCLA. (The authors themselves have, of course, used the manuscript to teach international finance at Berkeley and Princeton.) * Our assistants, including Barbara Aurelien, Sherri Ellington, Kazumi Uda, and Annie Wai-Kuen Shun, for help in countless dimensions ranging from processing mountains of photocopying to organizing the bibliography to compiling the author index. + Terry Vaughn, economics editor at MIT Press, for his advice and encouragement.Preface * The production team of Peggy Gordon (editor) and Paul Anagnostopoulos (com- positor) for their high level of skill and for patiently working with us through our extensive fine tuning of the manuscript. We also thank TEX expert Jacqui Scarlott for expertly processing and correcting our original Scientific Word files. + Finally, and most importantly, our spouses, Leslie Ann and Natasha, for pro- viding essential moral support and cheerfully tolerating the endless hours of work needed to bring this project to fruition. MO. KR. World Wide Web addresses (as of August 1996) For us: hup:iwww.wws.princeton.edwObstfeldRogoffBook. html This book at MIT Press: up://www-mitpress.mit.edu/mitp/recent-books/econ/obsth.htmn! MIT Press Web Site: http:/iwww-mitpress.mitedw/Introduction International macroeconomics is alive with great practical questions. What are the long-term implications of sustained United States current account deficits and. Japanese current account surpluses? How do government budget deficits affect in- terest rates, trade balances, and exchange rates? Is increasing global capital market integration affecting the nature and international propagation of business cycles, and how is it changing the susceptibility of economies to sudden shifts in investor sentiment? What if Europe really goes ahead and adopts a single currency” Is there any tendency for the per capita incomes of developing countries to converge over time to industrial-country levels? How important is monetary policy, and what are the channels through which it affects the economy? These are issues that interest policymakers, business people, and researchers alike. It is no wonder that interna- tional macroeconomic issues command the attention of the world’s financial press. This book offers a framework and a general approach for thinking about inter- national macroeconomics. We believe the framework is valuable from both the theoretical and practical perspectives. Central to it is the role of international asset, markets in allowing countries to trade consumption goods over time by borrowing, from and lending to each other. This intertemporal approach certainly illuminates the economics of current account imbalances. But it also discloses the intimate relationship between dynamic possibilities and international trade within periods. trade of distinct commodities, of consumption indexed to uncertain contingencies, and so on. We do not pretend to have definitive answers to all the questions listed at the start of this introduction, but you should find the approach developed here useful for thinking about all of them. First, a brief road map of the topics and questions covered in the various chapters (for more detail, see the table of contents). Chapters 1 through 7 of the book are all concerned with the “real” side of international macroeconomics. The models in Chapters 8 through 10 for the most part build closely on those of the preceding seven, but they bring money into the picture The first three chapters all assume one good on each date and view trade purely from an intertemporal perspective. Chapter 1 covers the many basic insights that can be developed in a simple two-period model with a single asset. Chapter 2 looks at implications of deterministic and stochastic economies with infinitely-lived rep- resentative consumers. Richer demographic assumptions are introduced in Chap- ter 3, which explores various overlapping-generations models. Throughout the book we focus often—though far from exclusively—on the case of a small coun- try that takes the world interest rate (and possibly other external prices) as given by world markets. As the body of international trade theory amply demonstrates, this can be a powerful and illuminating simplifying assumption, We nonetheless systematically study global equilibrium models as well, both to show how world prices are determined and to understand more completely the routes by which var- ious economic shocks are transmitted across national borders.xx Introduction Chapters 4 through 6 extend and apply the basic approach to myriad new ques- tions that cannot be addressed in a one-good, riskless-lending framework. In the fourth chapter, we introduce the possibility of several consumption goods on a given date, including nontraded goods (one of which could be leisure) and multiple tradable goods (so that the static terms of trade play a role). As we see, the theory illuminates some important facets of the long-term evolution of economic struc- ture. We conclude Chapter 4 by studying the terms of trade within an intertemporal Ricardian model in which nontradability is determined endogenously by interna- tional transport costs. Chapter 5 looks at the economic role of trade in risky assets, showing how models with uncertainty can be understood in terms of the fundamental principles covered in the previous four chapters. The chapter illustrates how international fi- nancial markets may dramatically affect the dynamics of the current account and the international transmission of business cycles. It also covers the essentials of international portfolio diversification and the empirical puzzles surrounding ob- served diversification patterns. One vital application of the models of this section is to international asset pricing, which we explore in depth. International debt problems are frequently at the center of policy concerns. Chapter 6 shows how they can arise when there is sovereign risk. The chapter also looks at other realistic departures from the idealized complete markets world of Chapter 5, notably models where asymmetric information is the central distortion. The results of this chapter throw light both on the degree to which the international capital market can perform its potential role and on the nature of the instruments, that will be traded. A revival of economic growth theory has been one of the major developments in macroeconomics over the past decade. Chapter 7 surveys this literature, As is appropriate for a book focusing on international macroeconomics, however, we place a relatively heavy stress on open-economy issues and intercountry compar- isons, Thus we highlight the role of cross-border capital mobility in economic convergence, the interplay between capital flows and market imperfections, and the growth effects of international diversification under uncertainty. In an earlier era, the function of money in trade and capital movements was considered the defining characteristic of international finance as a field (as opposed o now, when nonmonetary aspects of international exchange are viewed as at Jeast as important for understanding macro issues). Chapter 8 introduces monetary ‘models, and it does so in the simplest possible setting, one with fully flexible nominal prices. The chapter (which can be read before the first seven) highlights such classic or soon-to-be-classic subjects as seignorage, endogenous price-level instability, speculative exchange rate crises, target zones for exchange rates, and the pricing of risky nominal assets. Chapter 9 introduces sticky-price monetary models, with an emphasis on empiri- cal stylized facts and the workhorse Keynesian models. The conceptual frameworkxxi Introduction that grew out of Keynes’ General Theory (1936) furnished the dominant paradigm for more than a half century. The Keynesian mode of analysis started with the pio- neering work of Metzler (1942a,b) and Machlup (1943), included Meade’s (1951) brilliant and prescient synthesis, and culminated in the systematic incorporation of capital mobility by Fleming (1962) and Mundell (1963, 1964) and of forward- looking expectations by Dornbusch (1976).! Chapter 9’s coverage of Dornbusch’s celebrated overshooting model of exchange rates builds intuition and illustrates, some first implications of sticky prices. The chapter concludes with one of the most fruitful applications of the simple Keynesian framework, the strategic analysis of credibility in monetary policy. Chapter 10 is our attempt to provide a dynamic sticky-price model that preserves, the empirical wisdom embodied in the older Keynesian tradition without sacrific- ing the theoretical insights of modern dynamic macroeconomics. This chapter is 0 and w(c') <0.! Let y' denote the individual's output and r the real interest rate for borrowing or lending in the world capital market on date 1. Then consumption must be chosen e budget constraint a 2) This constraint restricts the present value of consumption spending to equal the present value of output. Output is perishable and thus cannot be stored for later consumption? We assume, as we shall until we introduce uncertainty about future income in Chapter 2, that the consumer bases decisions on perfect foresight of the future. This is an extreme assumption, but a natural one to make whenever the complexities introduced by uncertainty are of secondary relevance to the problem being stud- ied, Perfect foresight ensures that a model's predictions are driven by its intrinsic logic rather than by ad hoc and arbitrary assumptions about how people form ex- pectations. Unless the focus is on the economic effects of a particular expectational assumption per se, the deterministic models of this book therefore assume perfect foresight? To solve the problem of maximizing eq. (1) subject to eq. (2), use the latter to substitute for c, in the former, so that the individual's optimization problem reduces to max u(e}) + Bul(l + r)(y} - ef) + y$ The first-order condition for this problem is 1. Until further notice, we also assume that Tim we ‘The purpose of this assumption isto ensure that individuals always desire atleast alittle consumption in every period, so that we don't have to add formal constraints of the form c! > 0 to the utility ‘maximization problems considered later. ‘Whenever we refer to the subjective time-preference rate in this book, we will mean the parameter 8 such that = 1/(1 48), that is, 8 = (1 — B)/B. 2. Ata positive rate of interest r, nobody would want to store output in any case. In section 1.2 we will ‘ee how this intertemporal allocation problem changes when output can be invested, that is, embodied in capital to be used in producing future output 3. Even under the perfect foresight assumption we may sometimes loosely refer to an individual’ pectation” or (worse) “expected value” of a variable. You should understand that in a nonstochastic environment, these expectations are held with subjective certainty. Only when there is real uncertainty, asin later chapters, are expected valUes averages over nondegenerate probability distribution.1.1 A Small Two-Period Endowment Economy w(c) = (1 + r)Bu'(e)), @) which is called an intertemporal Euler equation. This Euler equation, which will recur in many guises, has a simple interpretation: at a utility maximum, the con- sumer cannot gain from feasible shifts of consumption between periods. A one- unit reduction in first-period consumption, for example, lowers U; by u'(c}). The consumption unit thus saved can be converted (by lending it) into 1 + r units of second-period consumption that raise Uy by (1 + r)Bu'(c}). The Euler equation (3) thus states that at an optimum these two quantities are equal. An alternative and important interpretation of eq. (3) that translates it into lan- guage more closely resembling that of static price theory is suggested by writing itas Buc) 1 wd) itr 4 ‘The left-hand side is the consumer's marginal rate of substitution of present (date 1) for future (date 2) consumption, while the right-hand side is the price of future consumption in terms of present consumption, ‘As usual, individual i’s optimal consumption plan is found by combining the first-order condition (3) {or (4)] with the intertemporal budget constraint (2). An important special case is the one in which B = 1/(1 +7), so that the subjective discount factor equals the market discount factor. In this case the Euler equation becomes u'(c}) = u'(ch), which implies that the consumer desires a flat lifetime consumption path, c = c}. Budget constraint (2) then implies that consumption in both periods is 2, where ld +r)yi +95] +r (6) Equilibrium of the Small Open Economy We assume that all individuals in the economy are identical and that population size is 1. This assumption allows us to drop the individual superscript / and to identify per capita quantity variables with national aggregate quantities, which we denote by uppercase, nonsuperscripted letters. Thus, if C stands for aggregate consump- tion and ¥ for aggregate output, the assumption of a homogeneous population of size | implies that c! = C and y! = ¥ for all individuals i. Our assumed demo- graphics simplify the notation by making the representative individual's first-order conditions describe aggregate dynamic behavior. The Euler equation (3), to take 4, The Swiss mathematician Leonhard Euler (1707-1783) served at one time as the court mathemati- cian to Catherine the Great of Russia, The dynamic equation bearing his nazme arose originally in the problem of finding the so-called brachistochrone, which is te least-time path ina vertical plane for an ‘object pulled by gravity between two specified points.Intertemporal Trade and the Current Account Balance one instance, will also govern the motion of aggregate consumption under our con- vention. We must keep in mind, however, that our notational shortcut, while innocuous in this chapter, is not appropriate in every setting. In later chapters we reintro- duce individually superscripted lowercase quantity variables whenever consumer heterogeneity and the distinction between per capita and (otal quantities are im- portant. Since the only price in the model is the real interest rate r, and this is exoge- nously given to the small economy by the world capital market, national aggregate quantities are equilibrium quantities. That is, the small economy can carry out any intertemporal exchange of consumption it desires at the given world interest rate 7, subject only to its budget constraint. For example, if the subjective and market discount factors are the same, eq. (5), written with C in place of c! and ¥ in place of y’, describes aggregate equilibrium consumption. The idea of a representative national consumer, though a common device in modem macroeconomic modeling, may seem implausible. There are, how- ever, three good reasons for taking the representative-consumer case as a start- ing point. First, several useful insights into the macroeconomy do not depend on a detailed consideration of household differences. An instance is the pre- diction that money-supply changes are neutral in the long run. Second, there are important cases where one can rigorously justify using the representative- agent model to describe aggregate behavior. Finally, many models in inter- national macroeconomics are interesting precisely because they assume differ- ences between residents of different countries. Sometimes the simplest way to focus on these cross-country differences is to downplay differences within coun- tries. We have seen [in eq, (5)] that when 6 = 1/(1 +r), the time path of aggregate ‘consumption is flat. This prediction of the model captures the idea that, other things the same, countries will wish to smooth their consumption, When the subject time-preference rate and the market interest rate differ, the motivation to smooth ‘consumption is modified by an incentive to tilt the consumption path, Suppose, for example, that 6 > 1/(1 +r) but C) = Cp. In this case the world capital market of- fers the country a rate of return that more than compensates it for the postponement of a little more consumption. According to the Euler equation (3), u'(C1) should exceed u/(Cz) in equilibrium; that is, individuals in the economy maximize utility by arranging for consumption to rise between dates 1 and 2. The effects of a rise in 5. One does not need to assume literally that all individuals are identical to conclude that aggregate consumption will behave as if chosen by a single maximizing agent. Under well-defined but rather strin- ‘gent preference assumptions, individual behavior can be aggregated exactly, as discussed by Deaton and ‘Muellbauer (1980, ch. 6). We defer a formal discussion of aggregation until Chapter 5. Fora perspective ‘on ways in wiih the representative-agent paradigm can be misleading, however, see Kirman (1992),S 1.1. A Small Two-Period Endowment Economy. + on initial consumption and on saving are rather intricate. We postpone discussing them until later in the chapter. 1.1.3 International Borrowing and Lending, the Current Account, and the Gains from Trade Let's look first at how intertemporal trade allows the economy to allocate its con- sumption over time. 1.1.3.1. Defining the Current Account Because international borrowing and lending are possible, there is no reason for an open economy’s consumption to be closely tied to its current output. Provided all loans are repaid with interest, the economy’s intertemporal budget constraint (2) is respected. In the special case 6 = 1/(1 +r), consumption is flat at the level Ci = C=C ineg, (5), but output need not be. If, for example, ¥i < Yo, the coun- try borrows C — Y; from foreigners on date 1, repaying (1 +r)(€ — ¥;) on date 2. Whenever date 2 consumption equals output on that date less the interest and prin- cipal on prior borrowing—that is, C = ¥2 — (1 +r)(Ci — ¥;)—the economy's intertemporal budget constraint obviously holds true. A country's current account balance over a period is the change in the value of its net claims on the rest of the world—the change in its net foreign assets. For example, in our initial simple model without capital accumulation, a country’s first- period current account is simply national saving. (In section 1,2 we will see that in general a country’s current account is national saving less domestic investment.) ‘The current account balance is said to be in surplus if positive, so that the economy as a whole is lending, and in deficit if negative, so that the economy is borrow- ing. Our definition of a country’s current account balance as the increase in its net claims on foreigners may puzzle you if you are used to thinking of the current account as a country’s net exports of goods and services (where “service” exports include the services of domestic capital operating abroad, as measured by interest and dividend payments on those assets). Remember, however, that a country with positive net exports must be acquiring foreign assets of equal value because it is selling more to foreigners than it is buying from them; and a country with negative net exports must be borrowing an equal amount to finance its deficit with foreigners. Balance-of-payments statistics record a country’s net sales of assets to foreigners under its capital account balance. Because a payment is received from foreigners for any good or service a country exports, every positive item of its net exports is associated with an equal-value negative item in its capital account— namely, the associated payment from abroad, which is a foreign asset acquired. ‘Thus, as a pure matter of accounting, the net export surplus and the capital account surplus sum identically to zero. Hence, the capital account surplus preceded by aIntertemporal Trade and the Current Account Balance minus sign—the net increase in foreign asset holdings—equals the current account balance Despite this accounting equivalence, there is an important reason for focusing on the foreign asset accumulation view of the current account. It plainly shows that the current account represents trade over time, whereas the net exports view draws attention to factors determining gross exports and imports within a single time period. Those factors are far more than unimportant details, as we shall see in subsequent chapters, but to stress them at the outset would only obscure the basic principles of intertemporal trade. ‘To clarify the concept of the current account, let Br .1 be the value of the econ- ‘omy’s net foreign assets at the end of a period r. The current account balance over period r is defined as CA, = B,+) — By. In general, the date r current account for a country with no capital accumulation or government spending is CAr = Bry — By = Yr +7: By — Cr, © where 7B; is interest earned on foreign assets acquired previously. (This conven- tion makes r; the one-period interest rate that prevailed on date t — 1.) 1.1.3.2 Gross National Product and Gross Domestic Product Equation (6) shows that a country’s current account (or net export surplus) is the difference between its total income and its consumption, The national income of an economy is also called its gross national product (GNP) and is measured as the sum of two components: the value of the final output produced within its borders and net international factor payments. Here, these factor payments consist of interest and dividend earnings on the economy's net foreign assets, which are capital operating abroad.° (In line with the definition of net viewed as domest exports given earlier, a country’s earnings on its foreign assets are considered part of its national product despite the fact that this product is generated abroad.) In terms of our formal model, GNP over any period t is ¥; + r:By, as just indicated. The first component of national product, output produced within a country’s ge- ographical borders, is called gross domestic product (GDP). In the present model 6. Strictly speaking, national income equals national product plus net unrequited transfer payments from abroad (including items like reparations payments and workers’ remittances to family members in other countries). Workers’ remittances, which represent payment for exported labor services. are not truly unrequited and are completely analogous to asset earnings, which are payments for capital services. We will eat them as such in section 1.5. In practice, however, national income accountants usually don't teat remittances as payments for service exports, The term “ross” in GNP reflects its failure to account for depreciation of capital—a factor absent from our theoretical model. When depreciation occurs, net national product (NNP) measures national income less depreciation. Empirical {economists prefer to work with GNP rather than NNP data, especially in international comparisons, because actual national account estimates of depreciation are accounting measures heavily infiuenced bby domestic fax laws. Reported depreciation figures therefore are quite unreliable and can differ widely from country 10 country. For the United States, a ballpark estimate of annual depreciation would be around 10 percent of GNP.1.1 A Small Two-Period Endowment Economy Table L1 GNP versus GDP for Selected Countries, 1990 (dollars per capita) Country cpp GNP Percent Difference Australia 17327 17,000 -19 Brazil 2,753 2,680 27 Canada 21515 20,470 -49 Saudi Arabia 5,429 7,050 299 Singapore 11,533 11,160 32 United Arab Emirates 17,669 19,860 a United States 21,569 21,790 10 Source: World Bank, World Development Report 1992, GDP is ¥;. Typically the difference between national and domestic product is a rather small number, but for some countries, those which have amassed large stocks of foreign wealth or incurred substantial foreign debts, the difference can be significant. Table 1.1 shows several of these cases. 1.1.3.3 The Current Account and the Budget Constraint in the ‘Two-Period Model Our formulation of budget constraint (2) tacitly assumed that B; =0, making CA) = ¥; — C; on the formal model’s date 1 (but not in general). By writing con- straint (2) as a strict equality, we have also assumed that the economy ends period 2 holding no uncollected claims on foreigners. (That is, B3 = 0. Obviously foreign- ers do not wish to expire holding uncollected claims on the home country either!) Thus, CAz = Yo +1rBy— C2 =¥2 + r(¥1 — C1) — Co -% = Cr —B,=-CAl, where the third equality in this chain follows from the economy's intertemporal budget constraint, eq. (2). Over any stretch of time, as over a single period, a country’s cumulative current account balance is the change in its net foreign assets, but in our two-period model with zero initial and terminal assets, CA; + CA2 = B3~ By A Figure 1.1 combines the representative individual's indifference curves with the intertemporal budget constraint (2), graphed as Q ¥2 — (1+ r)(Ci — ¥1). It provides a diagrammatic derivation of the small economy's equilibrium and the implied trajectory of its current account. (The figure makes no special assump- tion about the relation between B and 1 +r.) The economy's optimal consump- tion choice is at point C, where the budget constraint is tangent to the highestIntertemporal Trade and the Current Account Balance Period 2 consumption, C, Budge tne, y= Ye (14016, -¥4) Cy Y, +¥ei(t48) Period 1 consumption, C, Figure Lt Consumption over time and the current account attainable indifference curve. The first-period current account balance (a deficit in Figure 1.1) is simply the horizontal distance between the date 1 output and con- sumption points, As an exercise, the reader should show how to read from the figure’s vertical axis the second-period current-account balance. Economic policymakers often express concern about national current account deficits or surpluses. Our simple model makes the very important point that an un- balanced current account is not necessarily a bad thing. In Figure 1.1, for example, the country clearly does better running an unbalanced current account in both pe- riods than it would if forced to set Cy = ¥; and C2 = ¥p (the autarky point A), Intertemporal trade makes possible a less jagged time profile of consumption. The utility gain between points A and C illustrates the general and classic insight that countries gain from trade. Application: Consumption Smoothing in the Second Millennium B.C. ‘An early anecdote concerning the consumption-smoothing behavior underlying this chapter's model comes from the story of Joseph in the Book of Genesis. Schol- ars of the biblical period place the episode somewhere around 1800 3. ‘The Pharaoh of Egypt summoned Joseph, then an imprisoned slave, to interpret two dreams. In the first, seven plump cattle were followed and devoured by seven11 A Small Two-Peniod Endowment Economy lean, starving cattle In the second, seven full ears of corn were eaten by seven thin ears After heanng these dreams, Joseph prophesied that Egypt would enjoy seven years of prospenty, followed by seven of famine He recommended a consumption- smoothing strategy to provide for the years of famine, under which Pharaoh would appropriate and store a fifth of the grain produced during the years of plenty (Gen- esis 41 33-36) According to the Bible, Pharaoh embraced this plan, made Joseph his prime minister, and thereby enabled Joseph to save Egypt from starvation Why did Joseph recommend storing the grain (a form of domestic investment yielding a rate of return of zero before depreciation) rather than lending 1t abroad at a positive rate of interest? Cuneiform records of the period place the interest rate on loans of grain in Babyloma in a range of 20 to 33 percent per year and show clear evidence of international credit transactions within Asia Minor (Heichelheim 1958, pp 134-135) At such high interest rates Egypt could have earned a handsome return on its savings It seems hkely, however, that, under the military and pohtieal conditions of the second millenium BC, Egypt would have found st difficult to compel foreign countries to repay a large loan, particularly during a domestic famme Thus storing the grain at home was a much safer course The model in this chapter assumes, of course, that international loan contracts are always respected, but we have not yet examined mechamsms that ensure compliance with their terms We will study the question in Chapter 6 . 114 Autarky Interest Rates and the Intertemporal Trade Pattern Diagrams ike Figure | | can illuminate the main factors causing some countnes to run initial current account deficits while others run surpluses The key concept we need for this analysis 1s the autarky real interest rate, that 1s, the interest rate that would prevail in an economy barred from international borrowing and lending Were the economy restricted to consume at the autarky point A in Figure I 1, the only real interest rate consistent with the Euler eq (3) would be the autarky interest rate r*, defined by eq (4) with outputs replacing consumptions Bu'(¥2) 1 wi) Tre a This equation also guves the autarky price of future consumption in terms of present consumption Figure I 1 shows that when the latter autarky price 1s below the world rela- tuve price of future consumption—which 1s equivalent to r* being above r—future Consumption 1s relatively cheap in the home economy and present consumption relatively expensive Thus the home economy will “import” present consumption from abroad 1n the first period (by running a current account deficit) and “export” future consumption later (by repaying its foreign debt) This result 1s m accordLs Intertemporal Trade and the Current Account Balance with the principle of comparative advantage from international trade theory, which states that countries tend to import those commodities whose autarky prices are high compared with world prices and export those whose autarky prices are com- paratively low.’ It is the opportunity to exploit these pretrade international price differentials that explains the gains from trade shown in Figure 1.1 A rise in present output or a fall in future output lowers the autarky real interest, rate: either event would raise desired saving at the previous autarky interest rate, but since the residents of a closed endowment economy cannot save more in the aggregate without lending abroad, r* must fall until people are content to consume their new endowment. Similarly, greater patience (a rise in B) lowers r*. By modi- fying Figure 1.1, you can check that when r* is below the world interest rate r, the country runs a first-period current account surplus followed by a deficit, but still gains from trade. It may come as a surprise that the existence of gains from intertemporal trade does not depend on the sign of the country’s initial current account balance. The reason is simple, however. What produces gains is the chance to trade with some- one different from oneself, Indeed, the greater is the difference, the greater the gain. The only case of no gain is the one in which, coincidentally, it happens that her. This reasoning also explains how changes in world interest rates affect a coun- try’s welfare. In Figure 1.1 the economy reaps trade gains by borrowing initially because its autarky interest rate is above the world rate, r. Notice, however, that, were the world interest rate even lower, the economy's welfare after trade would be higher than in Figure 1.1. The basic reason for this welfare gain is that a fall in the world interest rate accentuates the difference between the home country and the rest of the world, increasing the gains from trade. A small rise in the world inter- est rate (one that doesn’t reverse the intertemporal trade pattern) therefore harms a first-period borrower but benefits a first-period lender. ‘Temporary versus Permanent Output Changes A suggestive interpretation of the preceding ideas leads to a succinct description of how alternative paths for output affect the current account. ‘The natural benchmark for considering the effects of changing output is the case B=1/(1 +r). The reason is that, in this case, eq. (7) becomes ler ter’ ‘which implies that the sole factor responsible for any difference between the world and autarky interest rates is a changing output level. 7. For adetailed discussion, see Dixit and Norman (1980),u 1.1.6 1.1 A Small Two-Period Endowment Economy Imagine an economy that initially expects its output to be constant over time. The economy will plan on a balanced current account. But suppose Yi rises. If Ya does not change, the economy’s autarky interest rate will fall below the world interest rate: a date 1 current account surplus will result as people smooth their consumption by lending some of their temporarily high output to foreigners. If Yp rises by the same amount as Yj, however, the autarky interest rate does not change, and there is no current account imbalance. Alternatively, consumption automatically remains constant through time if people simply consume their higher output in both periods. ‘One way to interpret these results is as follows: permanent changes in output do not affect the current account when f = 1/(1 +7), whereas temporary changes do, temporary increases causing surpluses and temporary declines producing deficits. Likewise, a change in future expected output affects the sign of the current account in the same qualitative manner as an opposite movement in current output. We will generalize this reasoning to a many-period setting in the next chapter. Adding Government Consumption So far we have not discussed the role of a government. Government consumption however, easy to introduce. Suppose government consumption per capita, G, enters the utility function ad- ditively, giving period utility the form u(C) + v(G). This case is, admittedly, a simple one, but it suffices for the issues on which we focus. For now, it is easiest to suppose that the government simply appropriates G; in taxes from the private sector for ¢ = 1, 2. This policy implies a balanced government budget each period (we will look at government deficits in Chapter 3). The representative private indi- vidual’s lifetime budget constraint is thus 7, C2 Y2 — G2 C+ an G+ Bo, tT Oat (8) Government spending also enters the date ¢ current account identity, which is now CAp = Bist ~ Br =i + Br — Cr - Ga The new feature here is that both government and private consumption are sub- tracted from national income to compute the current account. (Plainly we must account for all domestic expenditure—public as well as private—to reckon how much a country as a whole is saving.) Since G is beyond the private sector's control we can follow the same steps as in section 1.1.1 to conclude that the Euler equation (3) remains valid. In- deed, introducing government consumption as we have done here is equivalent12 Intertemporal Trade and the Current Account Balance to relabeling the private sector's endowment ¥ as output net of government con- sumption, ¥ — G. How do government consumption decisions affect the current account? A nat- ural benchmark once again is the case in which 6 = I/(1 +7), and output is con- stant at ¥) = Y2 = ¥. Absent government consumption, private consumption would be constant in this case at C = ¥, with the current account balanced. Suppose, however, that G1 > 0 while G2 =0. Now the private sector will want to borrow against its relatively high second-period after-tax income to shift part of the bur- den of the temporary taxes to the future. The country therefore will run a deficit in period 1 and a surplus in period 2 Replacing Y with Y ~ G in eq. (5) implies that in the preceding example e- Wtnd -Gy4F7l_p_ C4nG ee Qtr Government consumption in period | lowers private consumption, but by an amount smaller than G1. The reason is that the government consumption is tem- porary: it drops to zero in period 2. Thus the current account equation presented earlier in this subsection (recall that B; = 0 here) implies that CA, =¥ -€-G,=- In contrast, suppose that G; = G2 = G. Then consumption is constant at @ = ¥ — G in both periods, and the current account is balanced always. Government consumption affects the current account here only to the extent that it tilts the path of private net income. 1.1.7 A Digression on Intertemporal Preferences Equation (1) assumes the representative individual's preferences are captured by a very particular lifetime utility function rather than an unrestricted function U; = U(Cy, Ca). In eq, (1) consumption levels for different dates enter additively; more- over, the period utility function u(C) is constant over time. With consumption occurring over T rather than just two periods, the natural generalization of utility function (1) is r u=D BWC). o tat Preferences that can be represented by an additive lifetime utility function are called intertemporally additive preferences. The key property implied by intertem- poral additivity is that the marginal rate of substitution between consumption on any two dates ¢ and s [equal to B°"w'(C;)/u'(C,) for the preferences described by eq, (9)] is independent of consumption on any third date, This property is re-1.1 A Small Two-Period Endowment Economy strictive (provided T > 2, of course). It rules out certain kinds of intertemporal consumption dependencies, such as complementarity between total consumption levels in different periods. Such dependencies are at the heart of recent models of habit persistence in aggregate consumption.® Although we will discuss particular alternative assumptions on tastes at several points in the book, the assumption of intertemporally additive preferences with an unvarying period utility function will form the backbone of our formal analysis, ‘There are several reasons for this choice: 1. Itis true that some types of goods, such as refrigerators and automobiles, are durable goods typically consumed over many periods rather than just one. This type of consumption linkage, however, is fundamentally technological. By defin- ing utility over the flow of services from durables, and by imputing their rental cost, one can easily incorporate such goods within the umbrella of intertemporally additive preferences, (We show this in Chapter 2.) 2. For some types of goods, consumption at one point in time clearly does influ- ence one’s utility from consuming in closely neighboring periods. After eating a large meal, one is less inclined to want another an hour later. The time intervals of aggregation we look at in macroeconomic data, however, typically are measured in months, quarters, or years, periods over which many types of intertemporal depen- dencies fade. 3. Admittedly, even over long periods, habit persistence can be important. Drug addiction is an extreme example; watching television is a closely related one. In macroeconomics, however, one should think of preferences as being defined over consumption variables that really represent aggregate spending on a wide array of different goods. While we may have some intuition about the persistence effect of consuming certain items, it is harder to see obvious and quantitatively signifi- cant channels through which the totality of consumption has long-lived persistence effects, 4, One can think of some types of goods that most individuals would prefer to consume only once, such as marriage services. But even though consumption of such services is lumpy for an individual, it is relatively smooth in the aggregate 8. If G() is any monotonically increasing function, then the utility function 6 [= orate} naturally represents the same preferences as U1 does, i., # monotonically increasing transforma tion of the lifeume utiicy function does not affect the consumer's underlying preference order ing over different consumption paths. Intetemporally additive preferences take the general form G[uiCy +--+ ur(Cr)} (with period uiility functions possibly distinct). They aso go by the name sronglyintertemporally separable preferences. For further discussion of thir implications, see Deaton ‘and Muellbauer (1980, ch. 5.3)4 Intertemporal Trade and the Current Account Balance People get married all the time. Similarly, people may take vacation trips only at infrequent intervals, but this is not the case in the aggregate. (Seasonality can be important in either of these examples, but such effects are easily dealt with.) 5. Fundamentally, a very general intertemporally nonadditive utility function would yield few concrete behavioral predictions. If consumptions on different dates are substitutes, one gets dramatically different results from the case in which they are complements, Because maximal generality would lead to an unfalsifi- able macroeconomic theory with little empirical content, macroeconomists have found it more fruitful to begin with a tractable basic setup like eq. (9), which has very sharp predictions. The basic setup can then be amended in parsimo- nious and testable ways if its implications seem counterintuitive or counterfac- tual 6. In any event, while empirical research has raised interesting questions about the simplest time-additive preference model, it does not yet clearly point to a superior nonadditive alternative. 1.2 The Role of Investment 124 Historically, one of the main reasons countries have borrowed abroad is to finance productive investments that would have been hard to finance out of domestic sav- ings alone. In the nineteenth century, the railroad companies that helped open up the Americas drew on European capital to pay laborers and obtain rails, rolling stock, and other inputs. To take a more recent example, Norway borrowed exten- sively in world capital markets to develop its North Sea oil resources in the 1970s after world oil prices shot up. So far we have focused on consumption smoothing in our study of the current account, identifying the current account with national saving. In general, however, the current account equals saving minus investment. And because, in reality, in- vestment usually is much more volatile than saving, to ignore investment is to miss much of the action. Adding Investment to the Model Let's modify our earlier model economy to allow for investment, We now assume that output is produced using capital, which, in turn, can be accumulated through investment. The production function for new output in either period is Y=F(K). (10) As usual production is strictly increasing in capital but subject to diminishing marginal productivity: F’(K) > 0 and F"(K) < 0. Furthermore, output cannot be15 1.2. The Role of Investment produced without capital: F (0) = 0. We will think of the representative consumer as having the additional role of producer with direct access to this technology.” A unit of capital is created from a unit of the consumption good. This process is reversible, so that 4 unit of capital, after having been used to produce output, can be “eaten.” You may find these assumptions unrealistic, but they help us sidestep some technical issues that aren’t really central here, One key simplification due to ‘our assumptions is that the relative price of capital goods in terms of consumption always equals 1 Introducing investment requires that we rethink the budget constraints individu- als face, because now saving can flow into capital as well as foreign assets. Total domestic private wealth at the end of a period r is now By) + Ky1, the sum of net foreign assets B,41 and the stock of domestic capital K,+1.!° How is capital investment reflected in the date ¢ current account? The stock of capital K,+; accumulated through the end of period ¢ is the sum of preexisting cap- ital K, and new investment during period t, /; (we ignore depreciation of capital): Ki =Ki th. ab Nothing restricts investment to be nonnegative, so eq. (11) allows people to eat part of their capital Next, the change in total domestic wealth, national saving, is Bist + Kiri — (Bp + Kr) = Yr + Br = Cr = Gr, Finally, rearranging terms in this equation and substituting (11) shows that the current account surplus is CA = Brat — B= ¥; + Be ~ Cr — Gr — (12) A very useful way to interpret the preceding current account identity is to label national saving as S,: Sp =e +B — CO, - Gy, (13) 9. As we discuss in later chapters, i is reasonable to think of labor as being an additional production input slongside capita. A production function ofthe form (10) stills valid as Tong as labor is supplied inelastically by the individual producer. We assume im, Fu) to ensure a positive capital stock. 10. Its simplest to suppose that all domestic capital is owned by domestic residents. The statement that (otal domestic wealth equals B + K is tre even when foreigners own part of the domestic capital stock, however, because domestic capital owned by foreigners is subtracted in calculating net foreign assets B, AS long as perfect foresight holds, so that the ex post returns to assets are equal, the ownership of the domestic capital stock is irelevant, The ownership pattem is not irelevant, as we see later, when. ‘unexpected shocks can occur.