1
The Macroeconomic Effects of Fiscal
Policy in Portugal: a Bayesian SVAR
Analysis
*
António Afonso
#
and Ricardo M. Sousa
$
December 2010
Abstract
With a new quarterly dataset we estimate a Bayesian Structural Autoregression model
and a Fully Simultaneous System approach to analyze the macroeconomic effects of
fiscal policy. Results show that positive government spending shocks, in general, have a
negative effect on real GDP; lead to “crowding-out” effects of private consumption and
investment; have a persistent and positive effect on the price level and a mixed impact
on the average financing cost of government debt. Explicitly considering the
government debt dynamics in the model is also important. A VAR counter-factual
exercise confirms that unexpected positive spending shocks create relevant “crowding-
out” effects.
Keywords: B-SVAR, Fully Simultaneous System, fiscal policy, debt dynamics,
Portugal.
JEL classification: E37, E62, H62, G10.
*
We are grateful to Ad van Riet, to participants to the 4
th
Meeting of the Portuguese Economic Journal,
and to an anonymous referee for helpful comments. The opinions expressed herein are those of the
authors and do not necessarily reflect those of the ECB or the Eurosystem.
#
European Central Bank, Directorate General Economics, Kaiserstraße 29, D-60311 Frankfurt am Main,
Germany. ISEG/TULisbon - Technical University of Lisbon, Department of Economics; UECE –
Research Unit on Complexity and Economics; R. Miguel Lupi 20, 1249-078 Lisbon, Portugal. UECE is
supported by FCT (Fundação para a Ciência e a Tecnologia, Portugal), financed by ERDF and Portuguese
funds. E-mails: antonio.afonso@ecb.europa.eu, aafonso@iseg.utl.pt.
$
University of Minho, Department of Economics and Economic Policies Research Unit (NIPE), Campus
of Gualtar, 4710-057 - Braga, Portugal; London School of Economics, Financial Markets Group (FMG),
Houghton Street, London WC2 2AE, United Kingdom. E-mails: rjsousa@eeg.uminho.pt;
rjsousa@alumni.lse.ac.uk.
2
1. Introduction
In the last twenty years, public spending control has been a major problem in
Portugal. The gains from the drop in interest rates and, consequently, in the interest
payments on the outstanding government debt were not accompanied by a sustained
consolidation of public finances. Moreover, the episodes of fiscal improvement that
occurred in the 1980s and in the 1990s have been short-termed and mostly not
successful. Following the introduction of the Stability and Growth Pact (SGP), Portugal
was the first country in the Economic and Monetary Union to breach the 3% of GDP
reference value for the government deficit in 2001. Consequently, it became subject to
the Excessive Deficit Procedure (EDP) in 2002, a situation that occurred again in 2005.
Therefore, given past performance and outcomes, it seems fair to say that after
entering the European Union (EU) in 1986, joining the Exchange Rate Mechanism
(ERM) of the European Monetary System (EMS) in 1992 and entering EMU in January
1999, Portugal’s fiscal track record could have been better.
In this context, the evaluation of the effects of fiscal policy on economic activity
in Portugal becomes relevant and is the major goal of this paper. Additionally, we look
at its impact on the composition of GDP, therefore, analyzing potential “crowding-out”
effects on private consumption and private investment.
Fiscal policy shocks are identified using a recursive partial identification
scheme
1
and we assess the posterior uncertainty of the impulse-response functions by
estimating a Bayesian Structural Vector Autoregression (B-SVAR) model. We also
account for the automatic response of fiscal policy to the economic activity, and use a
Fully Simultaneous System approach in line with the works of Blanchard and Perotti
(2002), Leeper and Zha (2003) and Sims and Zha (1999, 2006). In addition, we consider
1
Christiano et al. (2005) identify the monetary policy shock using the same procedure.
3
the response of fiscal variables to the level of the government debt following Favero
and Giavazzi (2007) and Afonso and Sousa (2009a).
Another important contribution of the paper is the use of a set of quarterly fiscal
data, which we build by drawing on the higher frequency (monthly) availability of fiscal
cash data. This allows us to identify more precisely the effects of fiscal policy.
The findings of this paper can be summarized as follows. On the one hand,
government spending shocks: (i) have a negative effect on real GDP; (ii) generate
substantial “crowding-out” effects and lead to a fall in both private consumption and
private investment; (iii) have a persistent and positive impact on the price level; and (iv)
have mixed impacts on the average cost of refinancing the debt. Therefore, and from a
policymaking perspective, increasing government spending does not emerge as an
obvious instrument to help fostering economic activity.
On the other hand, government revenue shocks: (i) have a negative impact on
GDP; (ii) crowd-out private consumption and private investment, although the response
emerges with a lag of about four quarters; and (iii) is normally followed by a somewhat
less disciplined fiscal policy.
The consideration of the feedback from government debt makes the effects of
fiscal policy on (long-term) interest rates and GDP more persistent and these variables
are also more responsive to the shocks. Moreover, the results do not seem to support the
existence of a significant stabilizing response of the budget balance to the debt level. In
fact, there is only weak evidence suggesting that: (i) government spending falls when
the debt-to-GDP ratio is above its mean (in particular, in the period 1979:1-1993:3); and
(ii) government revenue increases when the debt-to-GDP ratio is above its mean
(namely, in the period 1993:4-2007:4, that is, after the Maastricht Treaty). Therefore,
there was a possible Ricardian behaviour after the beginning of the 1990s, although the
4
past fiscal stabilization attempts have been mostly biased towards increases in
government revenue, without tackling sufficiently the spending side. Moreover, and as
we will see below, the breaching of the SGP by Portugal in 2002 cannot be
disconnected from such past fiscal responses. Finally, a VAR counterfactual exercise
shows that unexpected increases in government spending lead to important “crowding-
out” effects.
The rest of the paper is organized as follows. Section two reviews the related
literature. Section three presents fiscal developments in Portugal. Section four explains
the empirical strategies used to identify the effects of fiscal policy shocks. Section five
describes the data and provides the empirical analysis. Section six concludes.
2. Literature Review
Despite the large literature on the impact of monetary policy on economic
activity, the importance of fiscal policy for economic stabilization has received less
attention. This section provides a brief review of the existing evidence of the effects of
fiscal policy on GDP, the aggregate price level and the composition of output, that is,
private consumption and private investment.
For the U.S., different approaches have been used in the identification of the
fiscal policy shock. The “narrative approach” developed by Ramey and Shapiro (1998)
isolates political events and finds that, after a brief rise in government spending,
durables consumption falls while nondurable consumption displays a small decline.
Fatás and Mihov (2001) use a Cholesky ordering and show that increases in government
expenditures are expansionary, but lead to important changes in the composition of
output in the form of an increase in private investment that more than compensates for
the fall in private consumption. Blanchard and Perotti (2002) identify the automatic
5
response of fiscal policy by using information about the elasticity of fiscal variables,
and find that fiscal shocks are expansionary, have a positive effect on private
consumption, and a negative impact on private investment. More recently, Mountford
and Uhlig (2009) relying on sign restrictions for the fiscal impulse-response functions
find a negative effect in residential and non-residential investment for the U.S..
At the international level, the evidence is scarce due to the limited availability of
quarterly public finance data. Perotti (2004) finds that fiscal policy leads to no response
of private investment and a relatively large and positive effect on private consumption
in a set of five countries (Australia, Canada, Germany, the U.S. and the U.K.). For
France, Biau and Girard (2005) find a positive effect on both private consumption and
private investment. For Spain, De Castro and Hernández de Cos (2008) show that a
positive spending shock lead to higher inflation and lower output in the medium and
long-term, but can be expansionary in the short-term. Heppke-Falk et al. (2006) and
Giordano et al. (2007) find that government spending has expansionary effects on both
output and private consumption for, respectively, Germany and Italy. Afonso and Sousa
(2009a, 2009b) show that, for the U.S., the U.K., Germany and Italy, quarterly fiscal
policy shocks have important macroeconomic effects while also impacting on housing
and stock prices. In addition, Burriel et al. (2010), using a quarterly standard SVAR,
report that expenditure shocks are more persistent in the US than in the euro area, while
the negative response from net tax increases is shorter lived in the euro area.
As can be inferred from the abovementioned studies, data availability – in
particular, high frequency data –, remains a major drawback in the literature on fiscal
policy. Therefore, we try to overcome this issue, by building also a fiscal quarterly
dataset for Portugal.
