Labour Costs and the Size of Government
| Date | 01 April 2017 |
| Author | Mickael Melki,Andrew Pickering,François Facchini |
| DOI | http://doi.org/10.1111/obes.12140 |
| Published date | 01 April 2017 |
251
©2016 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
Labour Costs and the Size of Government*
Franc¸ois Facchini†, Mickael Melki‡,and Andrew Pickering§
†MSE, Centre d’Economie de la Sorbonne, University of Paris 1 Pantheon-Sorbonne, Paris,
France (e-mail: francois.facchini@u-psud.fr)
‡Department of Economics, Fribourg University, Boulevard de P´erolles 90, 1700 Fribourg,
Switzerland (e-mail: mickael.melki@unifr.ch)
§University of York, Heslington, York YO10 5DD, UK (e-mail: Andrew.Pickering@
york.ac.uk)
Abstract
Given inelastic demand for labour-intensive public services, the size of government de-
pends positivelyon labour costs. OECD data exhibit a strong statistical association between
government size and the business-sector labour share of income. When the labour share is
instrumented with measures of technological change, institutional variation and predeter-
mined data it continues to positively impact government size. In contrast, transfer spending
is unaffected by the labour share. The evidence is consistent with the idea that the recent
decline in the labour share has contributed to the slowdown in the growth of government
witnessed in much of the post-war era.
I. Introduction
Explanations of the size of government have historically focussed on its growth. However,
as shown in Figure 1, the share of government expenditure as a percentage of GDP in the
OECD has if anything declined in recent years. The 10-year average up until 2007 was
lowerthan that up to 1998 in all but five out of 29 countries in the sample.1Notwithstanding
important cyclical features in many of the countries, it is remarkable howg rowthin relative
governmentsize was the norm from the early 1960s to around the mid-1980s, and thereafter,
more-or-less universally ceased – albeit at different levels in different countries.
The literature distinguishes between demand- and supply-side explanations for the
growth of government.2Pickering and Rockey(2011) attribute the increases and the diver-
JEL Classification numbers: H10, H50, O41
*We thank seminar participants at the Assisi Workshop on Economics and Institutions, Thema (University of
Cergy), Economix (Paris 10), the European Public Choice Society Conference, the University of Fribourg, the
French Economic Association Conference, the Louis-Andr´eG´erard-Varet Conference in Public Economics, the
Public Choice Society Conference, and the University ofYork. We are also grateful to two anonymous referees for
helping to substantially improve the paper.Remaining errors are our own.
1The exceptions are France, Iceland, Japan, Korea (a notable outlier in the sample) and Portugal. The data are
truncated at 2007 because after this date financial bail-outs and fiscal stimuli have led to significant increases in total
outlays in many countries.
2Holsey and Borcherding (1997), Lybeck (1988) and Shelton (2007) surveythe extensive literature on the size of
government.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 79, 2 (2017) 0305–9049
doi: 10.1111/obes.12140
252 Bulletin
0
10
20
30
40
50
60
70
80
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
Total Government Outlays as a % of GDP
Year
The size of government, 1960-2007
AUSTRALIA AUSTRIA
BELGIUM CANADA
CZECH REP DENMARK
ESTONIA FINLAND
FRANCE GERMANY
HUNGARY ICELAND
IRELAND ISRAEL
ITALY JAPAN
KOREA LUXEMBOURG
NETHERLANDS NEW ZEALAND
NORWAY POLAND
PORTUGAL SLOVAK REP
SLOVENIA SPAIN
SWEDEN USA
UK
Figure 1 The size of government, 1960–2007
gence in government size observed in the earlier part of the sample as due principally to
demand-side factors. In their analysis the income elasticity of demand for public services
exceeds unity, as in Wagner’s (1893) law, once the median voter has reached a certain level
of income, and income elasticity differs with ideology. Government grows with income,
converging to a steady state that depends on ideology. This theory successfully explains
the growth and divergence observed in the data up until 1998, but the recent downward
movements noted above suggest other factors at work.3
This paper argues that supply side factors, in particular labour costs, are also important,
and thus proposes a new explanation that can help to account for the more recent contr-
actions in government size. The seminal supply-side explanation of government growth is
Baumol’s (1967) cost disease: costs are pushed up over time because of rising wages and
stagnant productivity in the public sector. Given inelastic demand for public services the
relative size of government grows.4However, at the present time of writing private sector
3Other demand side explanations include demographic change (e.g. Sanz and Vel´azquez, 2007), and the median
voter’s relative income (Meltzer and Richard, 1981). However, in both cases, data would point towards ongoing
increases rather than decreases in relativegovernment size. In the former instance, OECD populations have continued
to age, and in the instance of the Meltzer and Richard mechanism, measured inequality in the before-tax income
distribution has increased in many countries. Both trends would point to increased demand for redistribution rather
than government shrinkage.
Another important potential driver is globalization, which is controlled for in the analysis below.
4Borcherding (1985) estimates that 31% of the observed growth of total government size in the US between1902
and 1978 was due to the Baumol effect. Borcherding, Stephen Ferris and Garzoni (2004) and Sanz and Vel´azquez
(2007) find similar evidence in the panel of OECD countries.
©2016 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial