The Growth of the Public Sector: Theories and International Evidence.

Schap, David

Edited collections of essays are unruly as a rule. If a well-organized collection therefore pleasantly surprises, then the thorough plenning and fine execution so obvious in this book is very pleasing indeed. The order is not spontaneous. Rather, it is created using simple organizing principles by which specific goals are first identified then pursued via the scientific method. Here we have a book with a threefold mission: document the size and growth of

the public sector using clear definitions and recent international data (observation); review alternative hypotheses (theory), carefully partitioning those that address public sector size from those that address public sector growth; and present evidence on the various and sometimes competing hypotheses (empirical testing). The intended audience ranges from students who have advanced beyond intermediate theory to scholars researching and/or teaching public economics (i.e., public finance, public choice, or public policy).

One reads in the editor's opening chapter that public sector growth can be measured in various ways. Those interested in growth patterns measured in terms of the amount of resources governments use, own, control and/or produce will have to read elsewhere. This book focuses more narrowly on the most conventional measure of growth, namely changes in the amount governments spend. A twofold rationale is given for the limited focus: convenience of measure and economists' interest in nonmarket (or more particularly government) provision of goods and services.

The next two chapters present data concerning government expenditure. Paul Saunders writes on "Recent Trends in the Size and Growth of Government in OECD Countries," first describing a set of comparative measures of the level and growth of government activity, then applying principally one measure, namely government outlays relative to GDP. The data indicate that after several decades of steady increase, general government outlays relative to GDP for most OECD countries peaked during the 1980s with slight declines thereafter. Also documented is the general drift toward expenditures with ever more substantial welfare state/transfer components. The second chapter presenting summary data is similar to the first in structure as well as title, though its findings are dissimilar. In "Recent Trends in the Size and Growth of Government in Developing Countries," David Lira revisits the issue of measurement before presenting data and findings on developing countries. In these countries, especially the lower income economies, using most any conventional measure of relative size one finds smaller public sectors than those observed in developed countries. Lim highlights the rather slight decline in the last twenty-five years in the proportion of central government expenditure going to defense in developing countries, contrasting it with the rather dramatic declines witnessed in developed countries.

The next six chapters discuss a variety of theoretical issues. Magnus Henrekson finds four testable versions of "The Peacock-Wiseman Hypothesis" in the literature: (1) the strong version--real absolute government expenditure per capita evolves in a steplike pattern accompanying major social disturbances like war; (2) the semi-strong version--government expenditure as a proportion of national income evolves in a steplike pattern induced by social disturbances: (3) the weak version--the ratio of government expenditures to GDP, which follows an upward trend, is upwardly and permanently shifted as a consequence of social disturbance; and (4) the amorphous version--social disturbances induce a change in the values of the parameters in the model relevant to explaining government expenditure over time. Flaws in previous empirical testing of the Peacock-Wiseman Hypothesis are identified; new testing fails to confirm any of the four testable versions developed.

Alan Hamlin, in "Public Expenditure and Political Process," argues that the institutions and factors that comprise the political process are essential components in any model purporting to explain government expenditure levels or growth rates. Among the factors highlighted are median voter demand, interest group formation and log-rolling, and bureaucratic discretionary power. The last of these is explored in the next chapter. "The Economics of Bureaucracy," by John C. Cullis and Philip R. Jones. The authors are critical of Niskanen-style models of bureaucracy on theoretical grounds and ultimately dismiss their applicability for the post-war UK based on empirical testing. Their test, crudely described here, involves a relative public sector growth version of Wagner's Law in which bureaucratic power (measured as civil service and defense department employment relative to electorate size) is hypothesized as driving public sector growth (measured as government expenditure relative to GNP). The volume editor returns in Chapter 7 to discuss "Wagner's Law and Musgrave's Hypotheses." Alternative interpretations of Wagner's Law are identified in the literature and relevant empirical evidence presented. Among the findings are that government goods do have a positive income elasticity of demand, but not greater than unity; and increases in per capita income are indeed associated with a rise in government expenditure relative to national income, especially so if one is generous in allowing for lags in the association.

Those interested in a synthesized analysis of public sector growth will find "Modelling Public Expenditure Growth: An Integrated Approach," by Peter M. Jackson most satisfying. Analytically rich, yet still accessible to advanced undergraduate students, this chapter incorporates many insights and techniques from the public choice literature. Suitable attention is given to the limitations of the analysis, especially with respect to dynamic aspects.

Rounding out the six-chapter section termed Theoretical Perspectives is a chapter with an inverted perspective called "Government Consumption: Its Effects on Productivity Growth and Investment," by Steve Dowrick. Government spending is no longer the thing to be explained; rather, its consequences are. But if you think that growth in the size of government induces productivity increases, think again--conversely. Or consider this: government size must be understood as affecting economic growth, if at all, only through its effects on investment. Twist and turn, cause and effect--the plot line here reads like a good "whodunit", which I will not spoil with further elaboration.

The volume concludes with five case studies in public expenditure growth which align themselves into two groups. The first set of three are thematic in exploring particular kinds of expenditures: "Social Security Expenditure," by John Creedy; "Health Care Expenditure," by David K. Whynes; and "Higher Education Expenditure," by Paul Ryan. None of these suffer a typical fate of case studies, namely limited applicability. The first contains a discussion sufficiently general as to be applicable in a variety of countries. The latter two explicitly and repeatedly draw international comparisons. The final two selections are thematic in that the case studies involve particular countries: "The Political Economy of Public Sector Growth in the United States," by Terry L. Anderson and Thomas Stratmann; and "The Public Sector in Australia: A Quantitative Analysis," by Franz Hackl, Friedrich Schneider and Glenn Withers. Not surprisingly, in both cases the authors reject any unicausal explanation of long run government growth.

One leaves this volume with a better understanding of the multifaceted nature of public sector growth. Small wonder Peter M. Jackson described the multiple approaches to understanding public sector growth in these terms: "Each of these approaches focuses attention on one set of particular variables which might influence the growth of public spending; each contributes to the overall mosaic. Taken on their own they are an incomplete specification of the whole picture".

David Schap College of the Holy Cross