Taxation, technology, and the user cost of capital

324 Pages·1989·12.378 MB·English

by  Erik Biørn

CONTRIBUTIONS
TO
ECONOMIC ANALYSIS
182
Honorary Editor:
J. TINBERGEN
Editors:
D.W. JORGENSON
J. WAELBROECK
NH
m
NORTH-HOLLAND
AMSTERDAM • NEW YORK • OXFORD • TOKYO
TAXATION,
TECHNOLOGY
AND THE USER COST
OF CAPITAL
E. BI0RN
Department of Economics
University of Oslo
Norway
989
NORTH-HOLLAND
AMSTERDAM • NEW YORK • OXFORD • TOKYO
© ELSEVIER SCIENCE PUBLISHERS B.V., 1989
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording or
otherwise, without the prior written permission of the publisher, Elsevier Science Publishers B.V. /
Physical Sciences and Engineering Division, P.O. Box 1991, 1000 BZ Amsterdam, The Netherlands.
Special regulations for readers in the U.S.A. - This publication has been registered with the
Copyright Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained from
the CCC about conditions under which photocopies of parts of this publication may be made in the
U.S.A. All other copyright questions, including photocopying outside of the U.S.A., should be
referred to the copyright owner.
No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as
a matter of products liability, negligence or otherwise, or from any use or operation of any methods,
products, instructions or ideas contained in the material herein.
ISBN: 0 444 87490 9
Publishers:
ELSEVIER SCIENCE PUBLISHERS B.V.
Sara Burgerhartstraat 25
P.O. Box 211, 1000 AE Amsterdam
The Netherlands
Distributors for the United States and Canada:
ELSEVIER SCIENCE PUBLISHING COMPANY, INC.
655 Avenue of the Americas
New York, N.Y 10010
U.S.A.
PRINTED IN THE NETHERLANDS
INTRODUCTION TO THE SERIES
This series consists of a number of hitherto unpublished studies, which are intro-
duced by the editors in the belief that they represent fresh contributions to economic
science.
The term "economic analysis" as used in the title of the series has been adopted
because it covers both the activities of the theoretical economist and the research
worker.
Although the analytical methods used by the various contributors are not the same,
they are nevertheless conditioned by the common origin of their studies, namely
theoretical problems encountered in practical research. Since for this reason, busi-
ness cycle research and national accounting, research work on behalf of economic
policy, and problems of planning are the main sources of the subjects dealt with, they
necessarily determine the manner of approach adopted by the authors. Their
methods tend to be "practical" in the sense of not being too far remote from appli-
cation to actual economic conditions. In additon they are quantitative.
It is the hope of the editors that the publication of these studies will help to stimulate
the exchange of scientific information and to reinforce international cooperation in
the field of economics.
The Editors
PREFACE
This study attempts to give a unified treatment of the cost of capital ser-
vices and its relationship to the corporate income tax system on the one
hand and to the production technology of the firm on the other. In recent
decades, quantitative information on this cost has become useful for several
purposes in applied economic research, in particular in macroeconomic and
econometric model building and public economics.
The study relates the measurement of the price of capital services to the
measurement of the capital stock of the firm, and gives a parallel treat-
ment of the neo-classical technology with malleable capital and the putty-
clay technology with vintage specific input coefficients. It further discusses
and unifies various concepts of neutrality of corporate income taxation.
Empirical illustrations based on Norwegian data are given. The study
is intended to serve as a reference for researchers in econometric model
building, corporate investment behaviour, tax analysis, and national ac-
counting.
Thanks are due to several persons who have contributed, in various ways,
at several stages during the work with this monograph. First, I want to
express my gratitude to Dale W. Jorgenson for his encouragement, ad-
vice, and kind interest in the project. I have also received several helpful
comments from colleagues who have read, partly or entirely, previous ver-
sions of the manuscript, and I want in particular to thank Petter Prenger
and Kare Petter Hagen for a substantial number of comments, as well as
good advice, on the complete manuscript. My appreciation also goes to
Sigbj0rn Atle Berg, Erling Holm0y, Tor Jakob Klette, Karl Ove Moene,
Erik Offerdal, Oystein Olsen, and Agnar Sandmo for useful comments on
specific points, to J0rgen Ouren for help and advice with the program-
ming of the numerical calculations, to Elisa Holm, Hanne Ramb0l, Anne
Skoglund, Anne Thorstensen, and Solveig Wiig for excellent typing and
word-processing of several versions of the manuscript, and to Bruce Wol-
man who, with impressive skill, set the final camera-ready version in the
viii Taxation, technology and the user cost of capital
computer type-setting system TfeX. Of course, I have the only responsibility
for any remaining errors and shortcomings.
Finally, I gratefully acknowledge financial support from the Norwegian
Research Council for Science and the Humanities.
Oslo, November 1988
Erik Bi0rn
CHAPTER 1
INTRODUCTION
1.1 General background
The capital accumulation process plays a crucial role in economic growth.
This study is concerned with the concept and measurement of one of the
basic determinants of capital accumulation and factor allocation, and hence
of efficiency and growth, in a market economy, namely the cost of using
real capital as an input in production. This topic, like the problem of
measuring capital input itself, has been much discussed among economists.
The definition and measurement of the cost of using real capital input has
been approached in several ways in the public finance, the macroeconomic,
and the econometric literature. In business economics, investment criteria
based on the present value of an 'investment project' or on its internal rate
of return have had predominance. In empirical econometrics, a measure of
the cost of capital services may be essential for the modelling of capital ac-
cumulation and the estimation of relations explaining corporate investment
behaviour.
In recent decades, the effect of the tax system on the efficiency of cap-
ital allocation, through its distortionary effects on the incentives to save
and invest and the discrimination between different types of capital assets
it creates, has become an issue of increasing importance not only in the
economic and econometric literature, but also in the tax policy debate.
In several western countries, major reforms of the capital taxation have
been accomplished or are under way. A close examination of the literature
shows, however, a substantial diversity in approach when analyzing such
tax reforms.
The present study attempts to provide a unified treatment of the cost
of capital services, with emphasis on its relation to the corporate income
tax system on the one hand, and to the production technology of the firm
on the other. In the greater part of the study, the cost of capital services
will be considered an input price which, loosely speaking, has a similar
relationship to capital input as the wage rate has to labour input. It is a
2 Taxation, technology and the user cost of capital
rental price, representing the cost of using one unit of capital input per unit
of time in the same way as the wage rate is a rental price for labour input.
But unlike the wage rate, the capital service price cannot — with some
exceptions — be observed directly in the market. From an econometric
point of view, this limited observability of the capital service price raises
particular problems.
The tax imposed on corporate income, or net profit, and the tax on
corporate wealth (if the tax system includes such a tax), are essentially
taxes on the capital stock of the firm. This is the reason why the corporate
tax system may affect the cost of using capital services as perceived by
the firm, and may lead to a distortion of its input and output decisions.
Capital, unlike for instance labour services and energy, is a durable input.
Each capital good invested "produces" services to its user over a period
which is longer than the conventional accounting period, usually one year,
which underlies the corporate income (and wealth) taxation. Since capital
is durable, the tax code — inter alia via the rules for deduction of financial
costs, the depreciation rules, and the rules for taxation of capital gains —
usually gives a prescription of how the firm's real investment expenditures
should be accounted over time. This "periodization" of the investment
cost imposed by the tax code may be different from the "periodization"
which would be "correct" if the purpose is to define the true net profit of
the firm. There may be several reasons why such differences arise between
the tax accounted income and the true income. Measurement problems
is one such reason. The tax system will then affect the shadow value of
capital services for an optimizing firm. The textbook conclusion that a
profit tax does not distort the producer's input and output decisions is
valid in a stylized world of static profit maximization and with a particular
definition of taxable profit only.
Analytic expressions for the cost of capital services which reflect the effect
of the tax system are needed for numerous purposes in macroeconomic
model building, public economics, and econometrics. Such expressions may
be defined, interpreted, and measured in several ways depending on the
specific features of the tax system, the purpose of the investigation, and
the availability of data. Seminal contributions to this literature were given
by Jorgenson (1963, 1967), Jorgenson and Stephenson (1967), and Hall
and Jorgenson (1967, 1971). They denned the user cost of capital as the
opportunity cost of using real capital input in the presence of an income
tax and depreciation rules, and incorporated it into a neo-classical model of
producer behaviour. Since the term 'user cost of capital' over the years has
become familiar in the literature, we shall use it, or its abbreviated form
Introduction 3
the 'user cost', in this monograph. The term 'user price (of capital)' would
perhaps have been more appropriate, since the variable is essentially a price
variable, whereas the word cost is commonly referring to the product of a
price and a quantity variable. The work of Jorgenson on the impact of the
corporate tax system on the user cost of capital in a neo-classical setting
will be one of the starting points for this study.
The problems encountered in defining and measuring the cost of capi-
tal services are closely related to those that arise when measuring the real
capital stock, or the flow of capital services. They are in a sense dual
problems. The capital measurement problem has for a long time been rec-
ognized as a complex and controversial problem in economic research. In-
deed, Hicks has described it as "one of the nastiest jobs that economists has
set to statisticians" [Hicks (1969, p. 253)]. At the core of the problem lies
the fact that 'the capital stock' is a multi-dimensional economic concept.
First, it is a capacity measure, representing the (potential) flow of services
which can currently be "produced" by a production unit by means of the
stock of capital goods existing at the time. Second, capital has a wealth
characteristic. The value of a capital good reflects its ability to produce
capital services today and in the future. From this it follows that capital
goods which are equivalent in terms of current capacity may be widely
different when characterized by their future usefulness in the production
process. While the capacity dimension of the capital stock is in focus when
specifying the production technology, its wealth dimension is the relevant
dimension to consider when defining corporate income, depreciation, taxes
assessed on income and wealth, and hence income after taxes. As a conse-
quence, both the capacity and the wealth dimension of the capital will be
involved in the measurement of the cost of capital services in the presence
of a corporate tax system. The incorporation of the two-dimensional na-
ture of corporate capital is a distinctive feature of the approach to be taken
in this study. The problem of measuring the cost of capital services then
becomes no less complicated than the capital measurement problem. In
fact, at some stages in our analysis, three capital concepts will be involved,
one related to capacity, one related to wealth, and one related to the way
in which capital is accounted by the tax system. The duality between the
measurement of the capital stock and the measurement of its service price
is a second starting point for the present study.
A basic consequence of this duality is that the measurement of the cost
of capital services and the measurement of the capital stock should utilize
the same representation of the production technology. The specification
of the technology has two aspects, one related to the deterioration of the
4 Taxation, technology and the user cost of capital
capital input, the other related to input substitution. First, assume that a
measure of (the capacity dimension of) the capital stock can be constructed
by cumulating a weighted sum of past investment flows, i.e. the so-called
'perpetual inventory method'. The deterioration and retirement of capi-
tal goods over time will then be represented by the form of the weighting
function used. The deterioration process is an equally important compo-
nent in the user cost of capital as it is in the capital stock itself. It follows
from the duality that this weighting function, to be denoted as the survival
function of capital in this study, should be represented in exactly the same
way when constructing capital stocks as when constructing their user costs.
Second, the specification of factor substitution — notably between capital
and other inputs — should be the same. We get one user cost of capital
concept if the description of the technology is based on the neo-classical
model, with malleable capital and perfect substitutability between differ-
ent vintages of capital after installation. This is different from the concept
which we get when assuming a vintage technology characterized by imper-
fect substitutability between different capital vintages after installation, for
instance the putty-clay technology, with fixed input coefficients ex post.
The large majority of theoretical and applied works in this field — fol-
lowing Jorgenson's seminal contributions from the 1960's — have been
based on a neo-classical description of the technology and on a particu-
lar parametric form of the survival function, namely the exponential decay
specification. Under this assumption, the retirement will be a constant,
time invariant share of the volume of the capital stock. This implies that
the rate of retirement is independent of the age distribution of the capital
stock. Over the years, this has become the standard representation of the
retirement process not only in the literature on capital measurement, but
also in the growing literature on the neutrality or lack of neutrality of the
corporate income tax system.
Several objections can, however, be raised against this approach. First,
the assumption of an exponentially declining survival function has been em-
pirically contested by several authors [e.g. Griliches (1963), Feldstein and
Foot (1971), Eisner (1972), Feldstein and Rothschild (1974), and Hulten
and Wykoff (1981b)], although the issue is still open. Second, even if ex-
ponential decay may be an acceptable simplifying approximation for long-
term analysis, for instance in modelling economic growth [confer Jorgenson
(1963, p. 251) (1965, pp. 50-51)], it may give an inadequate description of
the deterioration process in the short run. Econometricians working with
quarterly or annual data and short-term models as well as researchers in
empirical tax analysis may feel it as a strait-jacket. Third, taking exponen-