What Drives Output Volatility? The Role of Demographics and Government Size Revisited
| Published date | 01 August 2019 |
| Author | Hauke Vierke,Martin Iseringhausen |
| DOI | http://doi.org/10.1111/obes.12286 |
| Date | 01 August 2019 |
849
©2018 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 81, 4 (2019) 0305–9049
doi: 10.1111/obes.12286
What Drives Output Volatility? The Role of
Demographics and Government Size Revisited*
Martin Iseringhausen† and Hauke Vierke‡
†Department of Economics, Ghent University, Sint Pietersplein 6, B-9000, Ghent, Belgium
(e-mail: martin.iseringhausen@ugent.be)
‡DG Economic and Financial Affairs, European Commission, CHAR, B-1049 Brussels,
Belgium (e-mail: hauke.vierke@ec.europa.eu)
Abstract
This paper studies the determinants of output volatility in a panel of 22 OECD countries. In
contrast to the existing literature, we avoid ad hoc estimates of volatility based on rolling
windows, and we account for possible non-stationarity. Specifically, output volatility is
modelled within an unobserved components model where the volatility series is the out-
come of both macroeconomic determinants and a latent integrated process. A Bayesian
model selection approach tests for the presence of the non-stationary component. The
results point to demographics and government size as important determinants of macroe-
conomic (in)stability.A larger share of prime-age workers is associated with lower output
volatility, while higher public expenditure increases volatility.
I. Introduction
Since its first documentation by Kim and Nelson (1999) and McConnell and Perez-Quiros
(2000), the persistent reduction in US business cycle volatility during the 1980s has inspired
a large body of literature. Stock and Watson (2002) coined the term ‘Great Moderation’ to
describe the puzzling fall in volatility. Summers (2005) and Del Negro and Otrok (2008)
showed that this moderation has been a global phenomenon with important differences
across countries regarding its magnitude, timing and sources. Much work has been done
on explaining the time variation in business cycle volatility, mainly through advances in
economic policy or changes in structural factors underlying the economy. In particular,
several cross-country studies that have strivedto explain volatility among OECD countries
JEL Classification numbers: C11, E32, E62.
*The viewsexpressed in this paper are those of the authors and do not necessarily reflect the European Commission’s
position. The authors thank Gerdie Everaert, Tino Berger, Gert Peersman, Freddy Heylen, Helmut Herwartz, Joris
Wauters,Lukas Vogel,twoanonymous referees, the editor Anindya Banerjee and the members of the Macroeconomics,
Policy,and Econometrics Research Group at Ghent University for helpful comments, as well as Olaf Posch for sharing
his dataset on taxation. The computational resources (Stevin Supercomputer Infrastructure) and services used in this
work were provided bythe Flemish Supercomputer Center, funded by Ghent University; the Hercules Foundation;
and the Economy, Science, and InnovationDepartment of the Flemish Government. Martin Iseringhausen gratefully
acknowledges financial support from Ghent University’s Special Research Fund (BOF).
850 Bulletin
suggest an important role for government size and the demographic composition of the
labour force in stabilizing the economy.
A seminal contribution on the role of the size of the government sector for output sta-
bilization was made by Gali (1994), who investigates the effect of income taxation and
government purchases on output volatility in a real business cycle model. While the theo-
retical model predicts that the destabilizing effect of income taxes dominates the stabilizing
effect of higher government purchases, the empirical model suggests that both are auto-
matic stabilizers in the Keynesian sense. Fat´as and Mihov (2001) estimate the effect of
government spending on output volatility for 20 OECD countries (1960–97), and report
that regardless of the volatility or government size measure, the effect on output is always
stabilizing. Other studies emphasize that the stabilizing effect of government spending
exhibits time variation (Pisani-Ferry, Debrun and Sapir, 2008), is subject to nonlinearity
depending on the actual level of government expenditures (Crespo Cuaresma, Reitschuler
and Silgoner, 2011; Collard, Dellas and Tavlas, 2017) and could suffer from endogene-
ity (Carmignani, Colombo and Tirelli, 2011). In addtion, Martinez-Mongay and Sekkat
(2005) highlight that the composition of public finances, in particular the tax mix, matters
for the impact of fiscal policy on macroeconomic stability. Posch (2011) derives the ef-
fect of various taxes on volatility within a stochastic neoclassical growth model and tests
the theoretical predictions through a panel regression of 20 OECD countries. Different
tax ratios are found to have different effects: Taxes on labour and corporate income are
stabilizing, while capital taxes increase volatility.
Next to the role of government size, a recent strand of the literature discusses the role
of demographics as a driver of macroeconomic volatility. As the decline in United States
output volatility coincides with a decrease in the number of young workers relative to
prime-age workers, the demographic composition of the labour force arises as a possible
explanation for volatility shifts. Jaimovich,Pr uitt and Siu (2013) arguethat the volatility of
hours worked differs across age groups which cannot be explained by age-specific labour
supply factors alone.1Hence, age-specific demand factors must play an important role
as well. The main idea is the introduction of capital-experience complementarity into the
production function. Older workers are more experienced on the labour market and have
gained firm-specific knowledge that is complementary to physical capital.As a result, firms
tend to hoard ‘older’ labour. A technology shock causes a larger reaction in the demand
for young workers as the older ones are a complement to physical capital, which is more
difficult to adjust in the short term. There exist alternative theoretical explanations on the
link between demographics and aggregate volatility. Studies show that next to a pure com-
positional effect also the endogenous reaction of firms (Lugauer, 2012a) and the education
and gender composition of the workforce (Mennuni, 2016) matter for the overall effect
on volatility. Lugauer (2012c) demonstrates how changes in firms’ hiring strategies can
reduce the labour market’s ability to amplify aggregate shocks and hence reduce volatility.
Jaimovich and Siu (2009) were the first to empirically test the link between demo-
graphics and output volatility. They document that the youngest and oldest among the
workforce experience larger fluctuations of hours worked than prime-age workers. To
1Hoynes, Miller and Schaller (2012) provide evidence based on microdata that labour market outcomes vary
significantly across age groups with younger individuals being much more responsive to cyclicalfluctuations.
©2018 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial