Professional Documents
Culture Documents
Textbook An Econometric Model of The Us Economy Structural Analysis in 56 Equations John J Heim Ebook All Chapter PDF
Textbook An Econometric Model of The Us Economy Structural Analysis in 56 Equations John J Heim Ebook All Chapter PDF
https://textbookfull.com/product/an-econometric-model-of-the-us-
economy-structural-analysis-in-56-equations-1st-edition-john-j-
heim-auth/
https://textbookfull.com/product/crowding-out-fiscal-stimulus-
testing-the-effectiveness-of-us-government-stimulus-programs-1st-
edition-john-j-heim-auth/
https://textbookfull.com/product/why-fiscal-stimulus-programs-
fail-volume-1-the-limits-of-accommodative-monetary-policy-in-
practice-1st-edition-john-j-heim/
https://textbookfull.com/product/econometric-analysis-global-
edition-william-h-greene/
Efficacy Analysis in Clinical Trials an Update Efficacy
Analysis in an Era of Machine Learning Ton J. Cleophas
https://textbookfull.com/product/efficacy-analysis-in-clinical-
trials-an-update-efficacy-analysis-in-an-era-of-machine-learning-
ton-j-cleophas/
https://textbookfull.com/product/cause-and-correlation-in-
biology-a-user-s-guide-to-path-analysis-structural-equations-and-
causal-inference-hernan-m-a/
https://textbookfull.com/product/hrm-in-africa-understanding-new-
scenarios-and-challenges-in-an-emerging-economy-john-e-opute/
https://textbookfull.com/product/partial-differential-equations-
of-classical-structural-members-a-consistent-approach-andreas-
ochsner/
https://textbookfull.com/product/why-fiscal-stimulus-programs-
fail-volume-2-statistical-tests-comparing-monetary-policy-to-
growth-effects-1st-edition-john-j-heim/
An Econometric Model of the US Economy
John J. Heim
An Econometric
Model of the
US Economy
Structural Analysis in 56 Equations
John J. Heim
University at Albany-SUNY
Albany, New York, USA
I left academic life in 1972, after getting my Ph.D. At that time large-scale
econometric modeling of the economy was the rage; everyone thought it
would be just a matter of time before we had “done enough science” to
allow economists to discuss economics in the classroom, not in terms of
the alphas and betas of theoretical models, but in terms of the real-world
coefficients they represent. Economics would become the next branch of
engineering, or so many thought.
Much to my surprise, when I returned to academic life 25 years later
things had not much progressed. Most economists were still using alphas
and betas to describe how one variable affects another in economics. For
lack of vigorous, concerted effort over those 25 years to pursue the hard
numbers underlying the theories, and their statistical significance, econom-
ists were still just discussing theories with the best “numbers” we had – the
abstract alphas and betas of pure theoretical discourse. Because we hadn’t
disciplined our presentation of theories to those scientifically proven to
work, even more theories abounded than was the case in 1972. Worse, the
overriding emphasis in economic theory was not on “what works?”, but
on “what’s new?”.
My engineering students knew the difference. When I tried to describe
macroeconomics as real science, and then described the coefficients that
connect one variable to another in alphas and betas, instead of real num-
bers, they just snickered. “Yes, but what is the real relationship?” they
would ask, meaning what are the real numbers? “And if you don’t have
them, why do you call this science?” they would ask. Certainly in their
vii
viii PREFACE
ix
x ACKNOWLEDGMENTS
Nor could the book have been written without the strong support of
my wife Sue. This book required 2 years full-time work, and before that,
considerable part-time work. The problems to be resolved required endless
long hours at work, and endlessly preoccupied my mind, even at home.
Sue was always willing to make the sacrifices necessary to cope with all
that.
Finally, I must acknowledge the secretarial assistance provided by
Annemarie Hebert. She has helped pull together, duplicate, and send out
endless drafts of this work.
C ONTENTS
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Modern Macroeconomics: Moving from
the Methods of Economic Philosophy to Those of
Economic Science 5
1.2 Summary of Ways in Which This Large-
Scale Econometric Model Improves on Past Work 9
1.3 The 56-Equation Model: 30 Behavioral Equations,
15 Identities (Product Side of National Income
and Product Accounts (NIPA)), and 8 Behavioral
Equations, 3 Identities (Income Side of NIPA) 14
1.4 The 38 Behavioral Equations: Coefficients,
Significance, R2 , and Durbin Watson Tests:
(Summary of Results: Detailed Explanations of
Findings Presented in Chapters 4–20) 16
2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.1 General Methodological Issues 40
2.2 Choosing Between VAR, DSGE, and
Cowles Commission Models 48
xi
xii CONTENTS
3.2 Otto Eckstein’s (1983) The DRI Model of the U.S. Economy 125
3.3 Ray Fair’s Estimating How the Macroeconomy Works (2004) 131
3.4 Federal Reserve Board/U.S. Model (1996) 140
3.5 Literature Review Summary 144
15 Endogeneity of Government
Spending Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
15.1 The Model for Total Government Spending for All
Purposes: Goods, Services, and Transfers 309
15.2 The Model for Government Spending on Goods
and Services Only 313
18 Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
18.1 Introduction 363
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
L IST OF F IGURES
xvii
xviii LIST OF FIGURES
xix
xx LIST OF TABLES
The book has two parts: Part I contains 45 equations describing in detail
the “product side” of the National Income and Product Accounts (NIPA).
It contains tested models of the GDP and its major components, and the
determinants of their level of production (Chapters 4–19). Part II provides
11 additional equations describing how the value of the product generated
producing the GDP is distributed among the factors of production. For
each factor of production there are two equations. The first describes the
variables that were found to determine each factor’s percentage share of
national income. The second describes the variables found to determine
the total amount (the level) of each factor’s total income. These models
describe the variables whose own changes cause the distribution of income
among factors to shift from one factor to another over time (Chapter 20).
Chapter 19 provides a summary of the substantive findings as to the
determinants of GDP and its components. Chapter 20, Section 20.5,
summarizes the determinants of factor shares and levels of income.
xxvii
xxviii SUMMARY
METHODOLOGY
Good science requires replicability of results. This chapter’s goal was to
provide, to the best extent possible, models whose results meet the rep-
licability standard. Largely, this goal appears to be achieved, though in
some areas more remains to be done. Hopefully, future generations of
researchers will find it worthwhile to take up where this study leaves off.
In particular, in some equations we were not able to fully resolve the “left
out” variables and multicollinearity problems that affects the credibility of
parameter estimates in any economic model.
In most models 85–95% of the variance is explained. However, in some
models, there are definitely some “left out” explanatory variables remain-
ing to be found. Less of the total variance in the model than we would
like is explained by the variables. Models with this problem are identified
in the text.
In addition, the problem of multicollinearity needs to be better
resolved. It is perhaps the most serious impediment to doing good science
in economics today. To mitigate the problem in this study, we use first dif-
ferencing, and careful selection of combinations of explanatory variables
used. In addition, we do extensive robustness testing, by adding and sub-
tracting explanatory variables to a model, to ensure (reasonable) model
changes do not cause marked changes in other parameter estimates. For
most of our parameter estimates we are able to show these techniques
achieved the desired level of stability, but not for all. For some models,
parameter estimates are still sensitive to exactly what other variables are
included in the model (these models are identified in the text). Economists
needs to develop better scientific methods for dealing with this problem.
CHAPTER 1
Introduction
2×6
This “2 × 6” was to show the length and width of the grave they
would have. Not only that, but the negroes that they could impose
upon and get to vote the democratic ticket received, after they had
voted, a card of safety; and here is that card issued to the colored
people whom they had induced to vote the democratic ticket, so that
they might present it if any white-leaguers should undertake to
plunder or murder them:
New Orleans, Nov. 28, 1874.
WILLIAM ALEXANDER,
President 1st Ward Col’d Democratic Club.
Attest:
J. H. HARDY, Ass’t Sec. Parish Committee.