On the size of government spending multipliers in Europe

Abstract

This article compares the size of government spending multipliers in Europe by applying a panel structural vector autoregression analysis on 11 eurozone and 8 non-eurozone countries using quarterly data from 1991Q1 to 2012Q4. We find that (i) spending multipliers are smaller in eurozone compared to non-eurozone countries, (ii) across the euro area the impact of government spending on GDP has been higher before than after the introduction of the euro, (iii) spending multipliers are larger in the eurozone periphery than in the core countries and (iv) since the beginning of the recent financial crisis, spending multipliers have become larger both for eurozone and for non-eurozone countries. We relate these results to an emerging theoretical literature linking the size of fiscal multipliers to the monetary policy stance. We also discuss the implications of our findings for the effectiveness of fiscal policy in Europe.

Keywords:

• euro area
• government spending multiplier
• panel vector autoregression
• crowding out

JEL Classification:

• C32
• E62
• H30
• H50

Notes

¹ Ilzetzki et al. (2013) present similar evidence for a large sample of 44 countries. They also find that the output effect of an increase in government consumption is larger in industrial than in developing countries, and is bigger in economies operating under fixed relative to more flexible exchange rate regimes.

² The baseline model is extended to other variables in Section V.

³  This normalization is without loss of generality as long as the diagonal elements of ${\mathbf{B}}_0$ remain unrestricted. For details, see Kilian (2011, pp. 2–5.)

⁴  This assumption is common in the literature cited in Section I, and is based on the observation that spending decisions are frequently made without regard to their implications on the government budget.

⁵ According to the definition of the European system of national and regional accounts (ESA95), government revenue consists of the following variables: (i) market output, output for own final use, payment for other nonmarket output; (ii) taxes on production and imports; (iii) property income; (iv) current taxes on income and wealth (receivable); (v) social contributions (receivable); (vi) other current transfers (receivable) and (vii) capital transfers (receivable).

⁶  These transfer items consist of social transfers (payable), other current transfers (payable) and capital transfers (payable).

⁷ The importance of including the debt-to-GDP ratio has been pointed out by Favero and Giavazzi (2007) who show that even if the VAR includes all the variables that affect the government’s intertemporal budget constraint, excluding the debt level can result in biased estimates of the effect of fiscal policy shocks on macrovariables.

⁸ Changing the ordering of the additional variables in the SVAR only had a negligible impact on the results reported in Table 2.

⁹  The ordering of the additional variables again had no bearing on the results reported in Table 3.