Working Pa P er Serie S [602882]

Working Pa P er Serie S
no 849 / J an U ar Y 2008
government Size
Com
P
o
S
ition,
v
olatilit
Y
and
eC
onomi
C

g
ro
W
th
by António Afonso
and Davide Furceri

WORKING PAPER SERIES
NO 849 / JANUARY 2008
GOVERNMENT SIZE, COMPOSITION,
VOLATILITY AND ECONOMIC
GROWTH 1
This paper can be downloaded without charge from
http://www.ecb.europa.eu or from the Social Science Research Network
electronic library at http://ssrn.com/abstract_id=1077767.
1 We are grateful to Peter Claeys, Georgios Karras, Ad van Riet, Jürgen von Hagen, and an anonymous referee for helpful comme nts and to
Simone Ruiz for research assistance. The opinions expressed herein are those of the authors and do not necessarily
refl ect those of the author’s employers.
2 European Central Bank, Directorate General Economics, Kaiserstraße 29, D-60311 Frankfurt am Main, Germany;
e-mail: [anonimizat]. ISEG/TULisbon – Technical University of Lisbon, Department of Economics;
UECE – Research Unit on Complexity and Economics, R. Miguel Lupi 20, 1249-078 Lisbon, Portugal;
e-mail: [anonimizat]. UECE is supported by FCT (Fundação para a Ciência e a Tecnologia, Portugal),
fi nanced by ERDF and Portuguese funds.
3 University of Palermo, Department of Economics, Italy; University of Illinois at Chicago, Department of Economics (M/C 144 ),
601 S. Morgan Street, Chicago, 60607, Illinois, USA; e-mail: [anonimizat], [anonimizat];
Tel/Fax: +1-3129962683/+1-3129963344.
In 2008 all ECB
publications
feature a motif
taken from the
€10 banknote.by António Afonso 2
and Davide Furceri 3

© European Central Bank, 2008
Address
Kaiserstrasse 29 60311 Frankfurt am Main, Germany
Postal address
Postfach 16 03 19 60066 Frankfurt am Main, Germany
Telephone
+49 69 1344 0
Website
http://www.ecb.europa.eu
Fax
+49 69 1344 6000
All rights reserved. Any reproduction, publication and
reprint in the form of a different publication, whether printed or produced electronically, in whole or in part, is permitted only with the explicit written authorisation of the ECB or the author(s).
The views expressed in this paper do not
necessarily refl ect those of the European
Central Bank.
The statement of purpose for the ECB
Working Paper Series is available from the ECB website, http://www.ecb.europa.eu/pub/scientific/wps/date/html/index.en.html
ISSN 1561-0810 (print)
ISSN 1725-2806 (online)

3ECB
Working Paper Series No 849
January 2008Abstract 4
Non-technical summary 5
1 Introduction 7
2 Motivation and literature 8
3 Methodology 12
4 Empirical analysis 16
4.1 Data 16
4.2 Results and discussion 18
4.3 Robustness analysis 21
5 Conclusion 23
References 26
Appendix 1 – Cyclical volatility 30
Appendix 2 – Data sources 32
Tables 34
European Central Bank Working Paper Series 42CONTENTS

4ECB
Working Paper Series No 849
January 2008

Abstract
This paper analyses the effects in terms of size and volatility of government revenue and spending on growth in OECD and EU countries. The results of the paper suggest that both variables are detrimental to growth. In particular, looking more closely at the effect of each component of government revenue and spending, the results point out that i) indirect taxes
(size and volatility); ii) social contributions (size and volatility); iii) government
consumption (size and volatility); iv) subsidies (size); and v) government investment
(volatility) have a sizeable, negative and statistically significant effect on growth. JEL: E62, H50, O40.

Keywords: Fiscal Policy, Government Size, Fiscal Volatility, Economic Growth.

5ECB
Working Paper Series No 849
January 2008Non-technical summary
Public spending is widely seen as having an important role in supporting economic
growth. On the other hand, a lower level of spending implies that fewer revenues are needed to achieve balanced budgets, which means that lower taxes can be levied, therefore contributing to stimulate growth and employment. Public spending is a key variable that
influences the sustainability of public finances via effects on fiscal balances and
government debt, and this is relevant for the success of common monetary areas such as the European Monetary Union. Additionally, in the European Union, the so-called Lisbon Agenda also assigned a relevant role to the reform of public finance in order to foster economic growth. For those reasons, a firm control and, where appropriate, reduction of public expenditure is important and a balance has to be drawn between running down public debt, cutting taxes and financing public investment in key areas. Therefore, in this paper we examined the effect of government size and fiscal volatility on growth for a set of OECD and EU countries, from 1970 to 2004. The overall results suggest that both dimensions tend to hamper growth in both country samples. Total revenue and total expenditure seem to impinge negatively on the real growth of per capita GDP both for the OECD and the EU countries. In particular, a percentage point increase in the share of total revenue (total expenditure) would decrease output by 0.12 and 0.13 percentage points respectively for the OECD and for the EU countries. It is worthwhile mentioning that the magnitude of the effect is almost identical in absolute terms to the effect of total investment (private and public) share on growth. Moreover, total expenditure volatility also has a negative and statistically significant effect on growth, at
least for the EU countries.

6ECB
Working Paper Series No 849
January 2008
Breaking up total revenue into direct taxes, indirect taxes and social contributions, our
results suggest that among total revenue the variables that are most detrimental to growth,
both in terms of size and volatility, are indirect taxes and social contributions. At the same time, analysing the components of total spending (transfers, subsidies, government consumption and government investment) the results suggest that, while for both set of
countries both subsidies and government consumption have a significantly negative impact
on growth, government investment does not significantly affect growth, and transfers have a positive and significant effect only for the EU countries. Moreover, for the EU countries, public consumption and investment volatility have a sizeable, negative and statistically significant effect on growth. These results are also in line with some available related empirical evidence pointing to the negative effects on growth of public spending, particularly in the case of developed countries. There are relevant policy implications to be drawn from these results. It seems that revenue reductions that occur mainly in terms of indirect taxes and social contributions, and cuts in government consumption and subsidies may contribute positively to fostering economic growth in the country samples analysed. Moreover, public capital formation may indeed turn out to be less productive if devoted to in efficient projects, or if it crowds out private
investment. These conclusions also provide useful indications to policy makers when deciding which components of public finances to adjust (namely by redirecting spending towards more growth enhancing activities, in a context of limited public resources and fiscal constraints).

7ECB
Working Paper Series No 849
January 2008
1. Introduction
Public spending is widely seen as having an important role in supporting economic
growth. On the other hand, a lower level of spending implies that fewer revenues are needed to achieve balanced budgets, which means that lower taxes can be levied, therefore contributing to stimulate growth and employment. Public spending is a key variable that
influences the sustainability of public finances via effects on fiscal balances and
government debt, and this is relevant for the success of common monetary areas such as the European Monetary Union. Additionally, in the European Union, the so-called Lisbon Agenda also assigned a relevant role to th e reform of public finance in order to foster
economic growth. For those reasons, a firm control and, where appropriate, reduction of public expenditure is important and a balance has to be drawn between running down public debt, cutting taxes and financing public investment in key areas. Moreover, a better control of fiscal variables would eliminate or at least reduce the possibility that fiscal policy itself is a source of macroeconomic volatility. In fact, if we accept that fiscal policy is in some cases driven by considerations which are not linked to macroeconomic stability, then there is the possibility that by limiting such actions the society will gain by having less economic volatility (both in terms of output and investment) and thereby higher economic growth. Generally speaking, one would like to redirect public expenditure towards increasing the importance of capital accumulation – both physical and human – and support such areas as research, development, and innovation. However, in order to understand how to restrict fiscal policy volatility and check government size, it is particularly important to understand which components of

8ECB
Working Paper Series No 849
January 2008
government revenue and spending are most detrimental to growth. In fact, understanding
the channel through which fiscal policy affects growth can help us to understand how to redirect public spending and revenue, and which components should be limited. Therefore, in the main contributions of the paper we provide some answers to this composition issue, and we address the effects of both government size and fiscal policy volatility on
economic growth using the volatility of the cyclical components of the budgetary
variables. Moreover, we also look into several budgetary revenue components in our analysis, something seldom done in the literature. We analyse, for a set of OECD and a subset of EU countries, from 1970 to 2004, the effect of total public revenue and expenditure and of their components on growth. In particular, for each of these components we consider two measures of fiscal activity: the relative size of each fiscal variable as a percentage of GDP, and business cycle volatility. The results of the paper suggest that for several components of general government revenue and spending both size and volatility measures have a negative effect on growth, and that restrictions on these variables should be seen as favourable. The remainder of the paper is organized as follows. In Section Two we position our paper in the related existing literature. In Section Three we explain our methodology. In Section Four we present the empirical analysis and discuss its results. Section Five summarises the paper’s main findings and provides some policy implications.
2. Motivation and Literature
One of the frequently quoted stylised facts of public sector economics is that of “Wagner’s
Law” about the long-run tendency for public expenditure to grow relative to some national

9ECB
Working Paper Series No 849
January 2008
income aggregate such as GDP.1 This implies that public expenditure can be treated as an
outcome, or an endogenous factor, rather than a cause of growth in national income. On
the other hand, Keynesian propositions treat public expenditure as an exogenous factor, which could be utilised as a policy instrument. In the former approach, the causality runs from national income to public expenditure whereas in the latter proposition, causality runs
from public expenditure via domestic demand to national income. Evidence concerning
this topic is not conclusive.
2
Additionally, Lucas (1988) argues that public investment in education increases the level of human capital and this can be seen as a main source of long-run economic growth. Moreover, Barro (1990) mentions the importance of government expenditure in public infrastructure for economic growth and Romer (1990) stresses the relevance of research and development expenditure. Therefore, composition of public spending is also a relevant issue, and if the aim is to promote growth, the focus should be put on the more productive items of the budget, even if the balance between the various functional items of the budget can vary according to the particular circumstances and priorities of each country.
3
The increase in total expenditure in most developed countries (see Table 1 for an illustration) must be seen against a background where governments gradually tried to fulfil “Musgravian” goals: macroeconomic stabilization, income redistribution, and more efficient resource allocation. In fact, it was during the 1970s and 1980s that most European

1 Adolf Wagner formulated in the 19th century a “law” re garding the expansion of government. Based on the
data he had he argued that as the wealth of society increases, so does the size of government.
2 A number of time series empirical studies have in the past found support for the so-called Wagner’s Law.
These, however, might not be reliable due to measurement errors and because they did not employ co-
integration tests to establish stationarity in the relevant variables (see for example, Peacock and Wiseman
(1961), Musgrave (1969 ), Bird (1971) and Beck (1982)). More recen t work seems to point to the existence of
a positive relation between per capita income an d public spending (see Martinez-Mongay (2002)).
3 For a survey on fiscal policy and economic growth see Zagler and Durnecker (2003), while Hemming, Kell
and Mahfouz (2002) review the effectiven ess of fiscal policy in economic activity.

10ECB
Working Paper Series No 849
January 2008
countries increased the coverage of social benefits such as unemployment insurance. In
addition, pension benefits related to public pension insurance were also reinforced in the 1970s and in the 1980s. On the other hand, from Table 1 from 1995 onwards expenditure ratios declined in most countries in order to contain rising public debt ratios.
[Insert Table 1]
Some increase in the size of the public sector is to be expected when taking into account past rising population and also to meet the broadening requirements of the welfare state in most countries. However, a larger public sector, as measured here by the share of government expenditures in GDP, does not necessarily imply a better satisfaction of public requirements or, for that matter, a more efficient approach to providing the minimum required benefits of the welfare state.
4
Fiscal volatility is another important issue regarding fiscal policy and its effect on growth. From a theoretical point of view, restrictions on government expenditure volatility may have both positive and negative effects on long-run growth. A crucial variable to determine the sign of these effects is business-cycle volatility.
5 On the one hand, since
governments can smooth out business-cycle fluctuations by the use of discretionary changes in fiscal policy and by the use of automatic stabilisers
6, fiscal policy may
positively affect private investment and long-run growth. On the other hand, fiscal policy itself might be a source of business-cycle fluctuations and exacerbate macroeconomic volatility, e. g. in case of pro-cyclical measures.

4 Afonso, Schuknecht and Tanzi (2005) report that so-called “big governments” seem to perform less
efficiently than “smaller governments.” See also Afonso et al. (2005) for a di scussion on spending
composition and growth.
5 In fact, as documented by several studies, busi ness-cycle volatility is harmful for growth and its
determinants. See, for example, Al tman (1992, 1995), Aghi on et al. (2005), Fatás (2002) and Furceri
(2007a), Ramey and Ramey (1995).
6 See, for example, Sachs and Sala-i-Martin (1991), Bayoumi and Masson (1995), Asdrubali, Sorensen and
Yosha (1996), von Hagen (1998), Afonso and Furceri (2007).

11ECB
Working Paper Series No 849
January 2008

Several papers have looked at the specific mechanism through which fiscal policy can affect business cycles. Most of these papers have analysed the effects of government expenditure restrictions on the ability to smooth economic fluctuations. For example, Roubini and Sachs (1989), Alt and Lowry (1994), Poterba (1995), Levinson (1998) and
Lane (2003) show that restrictions on government expenditure, and thus lower government
expenditure volatility, result in a slower adjustment of the economy to unexpected shocks. However, other papers that have directly addressed the effects of fiscal restrictions on the volatility of business cycles provide contradictory results. For example, Alesina and Bayoumi (1996) show that fiscal policy restrictions tend to have a negligible effect on business cycles, and they argue that the positive and negative effects discussed above cancel each other out.
7 In addition, Canova and Pappa (2005) show, for the US states, that
the presence of tighter budget constraints, debt related or institutional restrictions, does not appear to matter much for business cycle fluctuations. In contrast, Fatás and Mihov (2006) illustrate that although the two effects are statistically significant, the first effect predominates. In particular, Fatás and Mihov (2003) show that the volatility of output caused by discretionary changes in fiscal policy lowers economic growth by more than 0.8 percentage points for every percentage point increase in volatility. However, it is important to stress that not only discretionary changes but also transitory (and cyclical) changes in fiscal policy may increase output volatility and thereby reduce output growth. In fact, as has been suggested by Ayagari, Christiano and Eichenbaum (1992), temporary changes in fiscal policy may have a significant impact on

7 For instance, a report from the EC (2007) on the euro area mentions th at fiscal criteria and the Stability and
Growth Pact have dampened growth volatili ty in European and Monetary Union (EMU).

12ECB
Working Paper Series No 849
January 2008
interest rate volatility and this, in turn, will reduce long-run growth.8 To this point, Furceri
(2007b) analyzing a panel of 99 countries from 1970-2000, shows that a 1 percent increase
in government expenditure business cycle volatility determines a decrease of 0.78 percentage points in the long-run rate of growth.

However, although the effect of government expenditure volatility has been widely
analysed, the effect of volatility in the components of public spending and revenue has not so far been widely addressed in the literature.
9 This paper tries to fill this gap. In particular,
it analyses the effect on growth of several government spending and revenue items, both in terms of government size and volatility.
3. Methodology
Several studies in the growth literature have found a negative bivariate relationship
between growth and the measure of government size.
10 It is well known that the inclusion
of particular control variables in a growth regression can wipe out this bivariate relationship (e.g., Easterly and Rebelo, 1993). Thus, it is necessary to consider which information to include in such growth regressions as control variables. Sala-i-Martin (1997) running two million regressions found 60 variables to be significant in at least one growth regression. In a more robust analysis, Levine and Renelt (1992), applying the Extreme Bound Analysis initially proposed by Leamer (1983), found robust cross-country growth correlates to be: (i) the average investment share of GDP; (ii) the initial log of GDP per capita; (iii) initial human capital; and (iv) the average growth rate of the population. The initial level of GDP is not only a robust and significant variable for growth (in terms

8 For example, Bernanke (1983), Pi ndyck and Solimano (1993), and Blac kburn and Varvarigos (2005) show
that in models with investment irreversibility and fina ncial frictions, higher uncertainty regarding investment
prices will determine a lower level of investment and growth.
9 Some examples of somewhat sim ilar analysis are provided by Brunetti (1998) and Gong and Zou (2002).
10 See Plosser (1993).

13ECB
Working Paper Series No 849
January 2008
of conditional convergence, Barro and Sala-i-Martin, 1995), but output is generally
correlated with tax revenues and government expenditure. Openness is another variable found to be significant in many cross-country growth regressions, moreover it is somewhat related to fiscal policy. In fact, if open economies are
especially exposed to shocks, as argued by Rodrick (1998), it may be especially important
for the government to facilitate private consumption smoothing by operating a countercyclical policy. On the other hand, integrated international financial markets may offer more scope to absorb shock through risk sharing, suggesting there is less need for government to step in. In the same way output volatility (defined as the standard deviation of output business cycle) has been found to negatively affect growth (e.g. Ramey and Ramey, 1995) and is usually correlated with government expenditure volatility (e.g. Zimmerman, 1995; Lane, 2003; Fatás and Mihov, 2006). Moreover, since volatility of government expenditure is a combination of output volatility and policy measures, we need to consider output volatility in order to identify the effect of government expenditure volatility on growth.

Thus, in order to take into account robustly the determinants of growth, and to control simultaneity between growth and government size, and growth and fiscal volatility, in our empirical model we include in addition to the fiscal variables the initial log of GDP per
capita
0Y, the average total investment share of GDP ( I/Y), initial human capital h,
the average growth rate of the population n, openness OPEN , and output

14ECB
Working Paper Series No 849
January 2008
volatilityYV.11 We also include a country dummy for Germany and Finland to account
for breaks, and a year dummy for EMU and the EU single market ( SM), to see if EMU and
the creation of the single market were important determinants for growth among the EMU
and EU countries, respectively.12
Another relevant issue to be discussed using cross-country growth regression is the time span to be used. Usually studies in the growth literature make use of large time spans (30-40 years) and consider the average value of growth determinants over this time period. However, when fiscal and other policy variables are included, this could raise several problems, such as endogeneity and significant simultaneity. In particular, regarding fiscal policy, over long time spans the level of government spending and income are likely to be influenced by demographics, in particular an increasing share of elderly people. Thus, errors in the growth variable will affect GDP, demographics, and taxes or government spending. As a result, the independent variables, government revenue or government
spending as a share of GDP, are correlated with the error term in the growth regression,
and this will produce biased estimates. A second problem is that such cross-section studies, using long observation periods, give rise to an endogenous selection of fiscal policy
13. For instance, over a long time span
growth is likely to influence tax revenues. Countries that initially raise taxes and experience lower growth during the observation period are more likely to reduce taxes. In a similar way, countries that raise taxes without experiencing a negative growth effect are

11 See Appendix 1 for a detailed descripition on how we construct output and p ublic spending and revenue
volatility.
12 Eventually, events such as EMU could also be considered endogenous outcomes of growth.
13 In order to check for robustnwess of our results,i n section 4.3 we address the problem of simultaneity
using different empirical specifications.

15ECB
Working Paper Series No 849
January 2008
more likely to continue having high taxes. This means that cross-section studies over long
time spans may fail to capture growth causality effects of taxation. A third important issue is that growth regressions over long time spans may be inefficient since they discard all information on within-country variation both in terms of growth and production, and in terms of government size. Moreover, the inclusion of period dummies
help to control for the fact that most countries experienced a reduction in the growth rate in
the 1970s and 1980s. Country dummies take into account specific country-effects. For all those reasons we will focus mainly on combined cross-section time-series regressions using seven five-year periods from 1970 to 2004 (e.g. 1970-1974, …, 2000-2004), and we use pooled country and time fixed effects, and robust standard errors to control for heteroscedasticity between countries. Therefore, we estimate the two following growth equations, respectively for general government revenue and expenditure:
14

2
, 1 1, 1, 1, 1 , 1 1 ,R
it it it it it t i it gR R X T SDEJ V -I X H    w     (1)
2
, 2 2, 2, 2, 2 , 2 2 ,E
it it it it it t i it gE E X T SDEJ V -I X [    w     (2)
where the index i (i=1, …, 28) denotes the country, the index t (t= 1970-1974, 1975-1979,
…, 1999-2004) indicates the period, and
D1 and D2 stand for the individual effects to be
estimated for each country i. g is the growth rate of real GDP per capita, R is the set of

14 We analyse two separate sets of equations since reve nue and expenditure are usually quite correlated (the
correlation is 0.91 for total revenue and total spending , but lower for sub-components), and this would create
serious problems of multicollinaerity and would not allo w to identify which ex penditure and revenue
variables (both in terms of share and in terms of volatility) matter for growth. For the complete set of
correlations between spending and revenue items see Table A3 in Appendix 2. Moreover, at least in terms of
total expenditure and revenue, one shou ld be aware that symmetric effect s for revenue and expenditure could
be expected if both are linked to economic growth.

16ECB
Working Paper Series No 849
January 2008
general government revenue variables as percentage of GDP, E is the set of general
government expenditure variables as percentage of GDP, RV is the vector of revenue
volatility variables, and EVis the vector of expenditure volatility variables. X includes a
set of control variables (initial level of output per capita, output volatility, investment
share, human capital, population growth, and openness), and T and S are year and country
dummies. Additionally, regressions (1) and (2) also include square terms for R and E in
order to test the possible effect on economic growth of different government sizes.
4. Empirical Analysis
4.1 Data
In this paper we focus on OECD and EU countries. The countries included in the analysis
are the EU15 members, thereafter indicated as EU (Austria, Belgium, Denmark, France, Finland, Germany, Greece, Italy, Ireland, Luxembourg, the Netherlands, Portugal, Spain, Sweden, and the UK), and for the rest of OECD: Australia, Canada, Czech Republic,
Hungary, Iceland, Japan, Korea, New Zealand, Norway, Poland, Slovakia, Switzerland and
the United States. Studying just these two samples of countries offers several advantages. Firstly, a longer span of data is available for the OECD and the EU than for a broader set of countries, including for instance, developing countries. Secondly, data quality and cross-country comparability are also likely to be of a higher standard for the OECD, and this is extremely important when we consider the measure of fiscal volatility, since volatility would increase in presence of measurement errors. Thirdly, as argued by Grier and Tullock (1989), data from the OECD and the rest of the world do not share a common set of coefficients in cross-country growth regressions and thus should not be pooled.

17ECB
Working Paper Series No 849
January 2008

Fiscal data for the general government are obtained from the OECD Economic Outlook database (see Appendix 2 for details). For the computation of the volatility measures, all fiscal variables are converted into constant prices using the GDP deflator, since we do not want to eliminate any growth in government spending that takes the form of an increase in
the relative price of public sector outputs, and since for most aggregates there is not a well
defined deflator. In terms of public spending, we examine the following variables: Total Expenditure ( E)
and its breakdown in Transfers (TRA), Subsidies (SUB ), Government Investment ( GINV ),
and Government Consumption ( GCON ). For the latter variable we differentiate also
between Wage ( GWAGE) and Non-Wage components (GNONWAGE ). In terms of revenue
variables we analyse Total Revenue ( R), and its main components: Direct Taxes ( DIR),
Indirect Taxes ( IND) and Social Contributions ( SOC).
For all these variables we construct two measures. The first is the relative size of each variable as a percentage of GDP. The second considers cyclical volatility, which is computed as the standard deviation of the cyclical component of each variable.
15 In Table
2 we report the results for the average volatility.
[Insert Table 2]
Human capital ( h) is taken from the Barro-Lee (2001) data set. The rest of the control
variables described in the next section are taken from the from the Heston-Summers-Aten (2006) dataset (Penn World Table 6.2).
16

15 See Appendix 1 for a more detailed discussion about the filtering methods.
16 See Appendix 2 for a detailed description of all the variables and their availability.

18ECB
Working Paper Series No 849
January 2008

4.2. Results and Discussion
In Table 3 we report the estimates of the effect of general government total revenue, both
in terms of GDP’s share and volatility, on output growth (using the HP filter for annual
data with the smoothness parameter, O, equal to 6.25). The results suggest that revenues-
to-GDP seem to impinge negatively on the real growth of per capita GDP both for the
OECD and the EU countries. In particular, a percentage point increase in the share of total
revenue in GDP would decrease output growth by 0.12 percentage points both for the
OECD and the EU countries. In contrast, revenue volatility does not seem to significantly affect growth.
17 In addition, it seems that effect of government revenues ratios squared
does not depend on the relative size of government.
[Insert Table 3]
Similar implications emerge from analysing the effect of general government total spending. In Table 4 we report the estimates of the effect of total spending, both in terms of GDP’s share and volatility, on output growth. As in the case for total revenue, countries with a higher share of expenditure in GDP tend to grow more slowly. In particular, for the OECD (EU) countries an increase of one percentage point in the share of total expenditure to GDP would decrease growth by 0.13 (0.09) percentage points. The magnitude of the coefficient is almost identical in absolute terms to the effect of total investment on growth. Again, the effect of government spending ratios squared does not seem to depend on the relative size of government.
[Insert Table 4 here]

17 The effect of the control variables is generally significant (except for the dummies EMU and SM), and
their sign is consisten t with the one expected.

19ECB
Working Paper Series No 849
January 2008
A different pattern emerges in terms of spending volatility. While it has no effect on
growth for the OECD countries, it has a negative and significant effect for the EU countries. For these countries in particular, an increase of one percent in spending volatility (on average) would decrease growth by 0.76 percentage points.
18

Analysing the effect of each component of general government revenue on growth (Table
5), we can see that both indirect taxes and social contributions as a percentage of GDP have a negative effect on growth for both the OECD and the EU countries. In particular, while an increase of one percentage point in indirect taxes ratio lowers growth by 0.30 (0.40) percentage points for the OECD (EU) countries, an increase of the same magnitude in social contributions ratio decreases growth by 0.34 (0.38) percentage points for the OECD (EU) countries. Thus, it seems that while for the OECD countries social contributions are more detrimental to growth, for the EU countries indirect taxes are more harmful. In contrast, direct taxes and size do not seem to affect growth significantly for either set of countries.
19 This could suggest that direct taxes (such as income taxes) are less
distortionary than indirect taxes (such as VAT, sales taxes, goods and services taxes) and social contributions.
[Insert Table 5]
In terms of volatility it emerges that while the volatility of indirect taxes negatively affects growth in the OECD sample, the volatility of social contributions has a negative impact on growth for the EU country sample. In particular, a one percent increase in indirect taxes

18 The effect on growth is computed by multiplyi ng the estimated coefficient for the average volatility with
the average volatility (see Table 2). Va lues in the text and those obtained by multiplication may differ due to
rounding .
19 One has to notice that we are examining aver age tax burden and not margin al tax burden.

20ECB
Working Paper Series No 849
January 2008
volatility (social contribution) lowers growth for the OECD (EU) countries by 0.07 (0.27)
percentage points. Repeating the same analysis for the main components of general government expenditure, we can observe that for both sets of countries, subsidies and government consumption as a
percentage of GDP have a significant negative impact on growth.
20 This negative effect
could be explained by the fact that subsidies could provide in some cases did-incentives
and be distortionary. From columns one and three in Table 6, it is possible to see that while an increase of one percentage point in subsidies lowers growth by 0.44 (0.71) percentage points in the OECD (EU) countries, an increase of one percentage point in government consumption decreases growth by 0.23 (0.31). Thus, although the effect is more pronounced in the EU countries, subsidies seem to be more detrimental to growth than government consumption. Moreover, in the OECD countries, although government consumption hampers growth, it seems to have a less negative effect for countries with relatively bigger governments.
21 However, this improvement is quite negligible in terms of
magnitude, with the result that even for these countries government consumption is detrimental to growth. In contrast, while government investment does not seem to affect growth significantly for both sets of countries, transfers have a positive and significant effect for the EU countries.
[Insert Table 6]
In terms of volatility, while all the measures of spending volatility generally do not have a significant effect in the OECD countries, both government consumption and investment

20 One should be aware that there could be reverse cau sality between growth and subisidies, in the sense that
in a situation of lower growth governments may choose to increase the amount of subsidies given.
21 This could be partially explained by the fact that among these countries there are transition economies
such as the Czech Republic and Hungary where government consumption could have helped in the initial
restructuring process of the economy.

21ECB
Working Paper Series No 849
January 2008
volatility have a negative and statistically signif icant effect on growth in the EU countries.
For this set of countries in particular, an increase of one percent in government
consumption (investment) volatility lowers growth by 0.41 (0.46) percentage points.This could be explained by the fact, that higher government consumption and investment volatility will enchance macroeconomic uncertainty, reducing thereby growth and private
investment, also through an upward effect on the interest rate (Ayagari et al. 1992).
Splitting government consumption into wage and non-wage components, both variables as a percentage of GDP turn out to affect growth negatively for both sets of countries (Table 7). However, while in the OECD countries the magnitude of these effects is quite similar, in the EU countries non-wage components of government consumption seem to be more detrimental, and this result persists when we consider volatility measures. Moreover, in the EU countries the negative effect of wages is less relevant for those countries where the government wage share is higher than 18 percent.
22 However, this size effect is very
modest, with the result that also for these countries government wage is detrimental to growth.
[Insert Table 7]

4.3. Robustness Analysis
A relevant issue when specifying a panel growth equation is whether or not to include
country dummies. While the inclusion of country specific effects has the advantage of controlling for unobserved country heterogeneity, it could lead to misleading conclusion in the analysis of the results. In fact, it has to be kept in mind that the fixed effect estimator is equivalent to the OLS estimator after applying the “within” transformation. Therefore,

22 The variable GWAGE2 is not significant when we drop all the observations with government wage share
higher than 18%. These countries are: Belgium (1970-1974), Denmark (1980-1984) and Sweden (1975-
1979, 1980-1984).

22ECB
Working Paper Series No 849
January 2008
what is measured empirically is the deviation from country averages and not the long-run
growth effect. In order to control for possible misspecification due to the inclusion of country dummies, we re-estimate equations (1) and (2) using only time dummies. Analyzing the results
reported in Table 8, it clearly emerges that they are also robust when we exclude country
specific fixed effects. In particular, as we have mentioned in the previous section, the effect of general government total revenue and expenditure in term’s of GDP’s share and volatility (for the expenditure variable) are detrimental for growth in both the OECD and in the EU set of countries.
[Insert Table 8]
A second concern regarding our econometric specification is the possible reverse causality between growth and government size and growth and fiscal volatility. Regarding the first issue, one should bear in mind that if the elasticities of government spending and taxation with respect to output variations are less than 1 (which seems reasonable), then a five-year period of fast growth will determine a decrease of the ratios E/Y and R/Y . Thus, in order to
control for a possible endogeneity problem in our regression, we have re-estimated equations (1) and (2) using only the initial level of government spending and revenue-to-GDP ratios, in other words, the shares for the first year of each of the 5-year periods (see Table 9).
[Insert Table 9]
Similarly to this point, the direction of causality between fiscal volatility and growth in the simple OLS estimation is not very clear. Even if control variables as output volatility,

23ECB
Working Paper Series No 849
January 2008
initial level of GDP, and time and country effects can account for simultaneity, in the case
of both fiscal volatility and growth they are determined by the same set of explanatory variables. Therefore, in addition to the initial level of government and expenditure ratios,
we also used the one lag value for our measure of fiscal volatility (
1RV and1EV). Again, the
results are robust to these different specifications. Moreover, using the one lagged
measures of fiscal volatility, it comes out that also government total revenue volatility (in addition to expenditure volatility) is detrimental for growth.
5. Conclusion
Over the last three decades until the mid-1990s there has been a significant expansion of
government participation in the economy in all industrialized countries. Regarding this point, many studies have claimed that restrictions on fiscal policy could be favourable to growth. In fact, from a theoretical point of view, government size is likely to be detrimental to economic growth, for instance, due to inefficiency of government activities, while government volatility is likely to increase macroeconomic uncertainty. Thus, in order to understand how to restrict fiscal policy volatility and limit government size, it is quite relevant to assess which components of general government spending and revenue (both in terms of size and volatility) have a negative effect on growth. Therefore, in this paper we examined the effect of government size and fiscal volatility on growth for a set of OECD and EU countries. The overall results suggest that both dimensions tend to hamper growth in both country samples. Total revenue and total
expenditure seem to impinge negatively on the real growth of per capita GDP both for the
OECD and the EU countries. In particular, a percentage point increase in the share of total revenue (total expenditure) would decrease output by 0.12 and 0.13 percentage points

24ECB
Working Paper Series No 849
January 2008
respectively for the OECD and for the EU countries. It is worthwhile mentioning that the
magnitude of the effect is almost identical in absolute terms to the effect of total investment (private and public) share on growth. Moreover, total expenditure volatility also has a negative and statistically significant effect on growth, at least for the EU countries.

Breaking up total revenue into direct taxes, indirect taxes and social contributions, our
results suggest that among total revenue the variables that are most detrimental to growth, both in terms of size and volatility, are indirect taxes and social contributions. At the same time, analysing the components of total spending (transfers, subsidies, government consumption and government investment) the results suggest that, while for both set of countries both subsidies and government consumption have a significantly negative impact on growth, government investment does not significantly affect growth, and transfers have a positive and significant effect only for the EU countries. Moreover, for the EU countries, public consumption and investment volatility have a sizeable, negative and statistically significant effect on growth. These results are also in line with some available related empirical evidence pointing to the negative effects on growth of public spending, particularly in the case of developed countries.
23
There are relevant policy implications to be drawn from these results. It seems that revenue reductions that occur mainly in terms of indirect taxes and social contributions, and cuts in government consumption and subsidies may contribute positively to fostering economic growth in the country samples analysed. Moreover, public capital formation may indeed turn out to be less productive if devoted to in efficient projects, or if it crowds out private

23 For instance, de Ávila and Strauch (2003), in a pa nel framework, report that government consumption and
transfers are detrimental for growth in the EU, while Föls ter and Henrekson (2002) al so conclude that there
is a negative relationship between government expenditure and growth.

25ECB
Working Paper Series No 849
January 2008investment.24 These conclusions also provide useful indications to policy makers when
deciding which components of public finances to adjust (namely by redirecting spending
towards more growth enhancing activities, in a context of limited public resources and fiscal constraints).
Finally, regarding possible further empirical research, one could envisage looking more
closely at the optimal size and the nature of the relationship between the role of the various
components of government spending and revenue and growth.

24 Afonso and St. Aubyn (2007) report that crowding-i n effects of public investme nt on private investment
can vary considerably across countries.

26ECB
Working Paper Series No 849
January 2008
References

Afonso, A and Furceri D. (2007). “Business Cycle Synchronization and Insurance Mechanisms in the EU”. Manuscript. Afonso, A. and St. Aubyn, M. (2007). “Macroeconomic Rates of Return of Public and Private Investment: Crowding-in and Crowding-out Effects”. Manuscript. Afonso A, Schuknecht, L. and Tanzi, V. (2005). “Public Sector Efficiency: An
International Comparison”, Public Choice , 123 (3-4), 321-347.

Afonso A., Ebert, W., Schuknecht, L. and Thöne, M. (2005). “Quality of public finances and growth,” European Central Bank Working Paper n. 438. Aghion, P., Angeletos, G., Banerjee, A. and Manova, K. (2005). “Volatility and Growth: Credit Constraints and Productivity-Enhancing Investment”, NBER Working Papers,
11349. Alesina, A. and Bayoumi, T. (1996). “The Costs and the Benefits of Fiscal Rules: Evidence from U.S. States”, NBER Working Papers, 5614.
Alt, J. and Lowry, R. (1994). “Divided Government, Fiscal Institutions, and Budget Evidence from the States”, American Political Science Review 88 (4), 811-828.
Altaman, M. (1992). “Business Cycle Volatility in Developed Market Economies, 1870-1986: Revisions and Conjectures”, Eastern Economic Journal , 18, 259-275.
Altaman, M. (1995). “Business Cycle Volatility and Economic Growth: the Historical Record, 1870-1986”, Journal of Post Keynesian Economics , 17, 561-577.
Asdrubali, P., Sorensen, B. and Yosha, O. (1996). “Channels of Interstate Risk-sharing: US 1963-1990”, Quarterly Journal of Economics , 144, 1081-1110.
Ayagari, S, Christiano, L. and Eichenbaum, M. (1992). “The Output, Employment, and Interest Rate of Government Consumption”, Journal of Monetary Economics , 30, 73-86.
Barro, R. (1990). “Government Spending in a Simple Model of Endogenous Growth,” Journal of Political Economy 98 (5).
Barro, R. and Lee, J. (2001). “International Data on Educational Attainment: Updates and Implications”, Oxford Economic Papers 53(3), 2001.

Barro, R. and Sala-i-Martin, X. (1995). Economic Growth . McGraw-Hill, New York.
Bayoumi, T. and Masson, P. (1995). “Fiscal Flows in the United States and in Canada:
Lessons from Monetary Union in Europe”, European Economic Review , 39 (2), 253-74.

Baxter, M. and King, R. (1999). “Measuring Business Cycles: Approximate Band-Pass Filters for Economic Time Series”, Review of Economics and Statistics , 81 (4), 575-593.

27ECB
Working Paper Series No 849
January 2008
Beck, M. (1982). “Towards a theory of public sector growth,” Public Finance , 37, 163-77.

Bernanke, B. (1983). “Irreversibility, Uncertainty ad Cyclical Investments”, Quarterly
Journal of Economics, 85-106.
Blackburn, K. and Varvarigos, D. (2005). “Growth, Uncertainty and Finance”, Centre for Growth and Business Cycle Research Discussion Papers, 48. Bird, R. (1971). “‘Wagner’s law: a pooled time-series, cross section comparisons,”
National Tax Journal , 38, 209-218.

Brunetti, A. (1998). “Policy volatility and economic growth: a comparative empirical analysis”, European Journal of Political Economy , 14 (1), 35-52.
Canova, F. and Pappa, E. (2005). “Does it cost to be virtuous? The macroeconomic effects of fiscal constraints”, NBER Working Papers, 11065. Christiano, L. and Fitzgerald, T. (2003). “The Band Pass Filter,” International Economic
Review , 4 (2), 435-465.
de Ávila, D. and Strauch, R. (2003). “Public finances and long-term growth in Europe – evidence from panel data analysis” ECB Working Paper N. 246. Easterly, W. and Rebelo, S. (1993), “Fiscal Policy and Economic Growth: An Empirical Investigation,” Journal of Monetary Economics 32 (3), 417-458. EC (2007). Quarterly Report on the Euro Area , vol. 6, N. 1, March, European
Commission. Fatás A. (2002). “The Effects of Business Cycles on Growth” in Loayza, N. and Soto, R. (eds.) Economic Growth: Sources, Trends and Cycles , Central Bank of Chile.
Fatás, A. and Mihov, I. (2003). “The Case for Restricting Fiscal Policy Discretion”, Quarterly Journal of Economics , 118, 1419-1447.
Fatás, A. and Mihov, I. (2006). “The Macroeconomics Effects of Fiscal Rules in the US States”, Journal of Public Economics , 90, 101-117.
Fölster, S. and Henrekson, M. (2002). “Growth effects of government expenditures and taxation in rich countries”, European Economic Review 45 (8), 1501-1520.
Furceri, D. (2007a). “Long-Run Growth and Volatility: Which Source Really Matters?” Manuscript, University of Illinois at Chicago. Furceri, D. (2007b). “Is Government Expenditure Volatility Harmful for Growth? A Cross-Country Analysis,” Fiscal Studies 28 (1), 103-120.
Gong, L. and Zou, H. (2002). “Effect of growth and volatility in public expenditures on economic growth: theory and evidence”, Annals of Economics and Finance , 3, 379-406.

28ECB
Working Paper Series No 849
January 2008
Grier, K., and Tullock, G. (1989). “An empirical analysis of cross-national economic
growth: 1951-1980,” Journal of Monetary Economics 24, 259-276.
Hemming, R.; Kell, M. and Mahfouz, S. (2002). “The Effectiveness of Fiscal Policy in Stimulating Economic Activity – A Review of the Literature,” IMF Working Paper No. 02/208.
Heston, A., Summers, R. and Aten, B. (2006). Penn World Table Version 6.2, Center for
International Comparisons of Production, Income and Prices at the University of
Pennsylvania, September.
Hodrick, R. and Prescott, E. (1997). “Postwar U.S. Business Cycles: An Empirical
Investigations”, Journal of Money, Credit and Banking , 29(1), 1-16.
Lane, P. (2003). “The Cyclical Behavior of Fiscal Policy: Evidence from the OECD”, Journal of Public Economics, 87, 2261-2275.
Leamer, E. (1983). “Let’s Take the Con Out of Econometrics”, American Economic
Review , 73, 31-43.
Levine, R. and Renelt, D. (1992). “A Sensitivity Analysis of Cross-Country Growth Regressions”, American Economic Review , 82, 942-963.
Levinson, A. (1998), “Balanced Budgets and Business Cycles: Evidence from the States”, National Tax Journal , Vol. 51, No. 4, December.
Lucas, R. (1988). “On the mechanism of economic development’, Journal of Monetary
Economics , 22, 3-42.
Martinez-Mongay, C. (2002). “Fiscal policy and the size of governments’, in Buti, M., von Hagen, J. and Martinez-Mongay, C. (eds.), The behaviour of fiscal authorities, Palgrave:
Basingstoke. Musgrave, R. (1969). Fiscal Systems , Yale University Press, New Haven.
Peacock, A. and Wiseman, J. (1961). The Growth of Public Expenditure in the United Kingdom, Princeton University Press. Pindyck, R. and Solimano, A. (1993). “Economic Instability and Aggregate Investments”, in Blanchard, O. and Fisher, S. (eds.) NBER Macroeconomics Annual . MIT Press,
Cambridge. Plosser, C. (1993). “The search for growth, in: Policies for Long-Run Growth”. The
Federal Reserve Bank of Kansas City symposium series .

Poterba, J. (1995). “State responses to Fiscal Crises: the Effects of Budgetary Institutions”,
Journal of Political Economy , 102, 799-821.
Ramey, G. and Ramey, V. (1995). “Cross-country evidence on the link-between volatility and growth”, American Economic Review, 85, 1138-1151.

29ECB
Working Paper Series No 849
January 2008

Ravn, M. and Uhlig, H. (2002). “On Adjusting the Hodrick-Prescott Filter for the Frequency of Observations”, Review of Economics and Statistics, 84, 371-380.
Rodrick, D. (1998). “Why Do More Open Economies Have Bigger Governments?”, Journal of Political Economy , 106(5), 997-1032,
Romer, P. (1990). “Endogenous Technological Change,” Journal of Political Economy , 98
(5), 71-102.

Roubini, N. and Sachs, J. (1989). “Government Spending and Budget deficits in the Industrialized Countries”, Economic Policy , 8, 99-132.
Sachs, J. and Sala-i-Martin, X. (1991).
“Fiscal Federalism and Optimum Currency Areas:
Evidence for Europe from The US”, NBER Working Paper 3885. Sala-i-Martin, X. (1997), “I Just Ran Two Million Regressions”, AEA Papers and
Proceedings 87, 178-183.
Stock, J. and Watson, M. (1999). “Business Cycle Fluctuations in U.S. Macroeconomic Time Series”, in Taylor, J. and Woodford, M. (eds.), Handbook of Macroeconomics , vol 1,
3-64, Elsevier.
von Hagen, J. (1998). “Fiscal Policy and Intranational Risk-sharing”, CEPR Discussion
Paper, 126.
Zagler, M. and Durnecker, G. (2003). “Fiscal policy and economic growth,” Journal of
Economic Surveys 17 (3), 397-418.
Zimmermann, C. (1995). “International Trade over the Business Cycle: Stylized Facts and Remaining Puzzles”. Working Paper 37, Centre de Recherche sur l’Emploi et les Fluctuations Économiques, Université du Québec à Montreal.

30ECB
Working Paper Series No 849
January 2008
Appendix 1 – Cyclical Volatility
Business cycle measures are obtained by detrending the series of real GDP, and of real
government expenditure and revenue items. Four different methods are used to detrend the
series of each country i and estimate its cyclical component. Letting t i t i Y y, ,ln , the first
measure is simple differencing (growth rate):
1 , , , t i t i t iy y c . (A1)
The second and the third method use the Hodrick-Prescott (HP) filter, proposed by
Hodrick and Prescott (1997). The filter decomposes the series into a cyclical t ic, and a
trend t ig, component, by minimizing with respect to t ig,, for the smoothness
parameter 0!O the following quantity:
  ¦¦

   T
tT
tt i t i t i t i g g g y
11
22
1 , 1 ,2
, ,O . (A2)
The methods differ because the second one consists of using the value recommended by
Hodrick and Prescott for annual data for the smoothness parameter ( O) equal to 100, while
the third method considers the smoothness parameter ( O) to be equal to 6.25. In this way,
as pointed out by Ravn and Uhlig (2002), the Hodrick-Prescott filter produces cyclical
components comparable to those obtained by the Band-Pass filter.
The fourth method makes use of the Band-Pass (BP) filter proposed by Baxter and King
(1999), and evaluated by Stock and Watson (1999) and Christiano and Fitzgerald (2003)
(who also compare its properties to those of the HP filter). The Low-Pass (LP) filter ) (LD ,
which forms the basis for the band pass filter, selects a finite number of moving average
weights hD to minimize:
³ S
SZ Z Gd Q2) ( , (A3)
where ¦ K
K hh
hL L D D) ( and ¦  K
K hh i
h K eZD Z D) ( .

31ECB
Working Paper Series No 849
January 2008

The LP filter uses ) (ZDK to approximate the infinite MA filter ) (ZE . Defining
) ( ) ( ) (ZDZEZG  { , and then minimizing Q, we minimize the discrepancy between the
ideal LP filter ) (ZE and its finite representation ) (ZDK at frequency Z. The main
objective of the BP filter as implemented by Baxter and King (1999) is to remove both the
high frequency and low frequency components of a series, leaving the business-cycle frequencies. This is obtained by subtracting the weights of two low pass filters. We define
LZ and HZ, the lower and upper frequencies of two low pass filters, as eight and two
respectively for annual data. We therefore remove all fluctuations shorter than two or longer than eight years. The frequency representation of the band pass weights becomes
) ( ) (L K H KZDZD , and forms the basis of the Baxter-King filter, which provides an
alternative estimate of the trend and the cyclical component. The four filters yield substantially similar results, with only minor differences (differencing generally produces the most volatile series, while the HP6.25 the smoothest). For example, analysing the effect of total expenditure volatility on growth for the EU countries (results of the third column in Table 3), we can see that the effect is negative and
significant for all filtering methods. Moreover, as suggested also by Ravn and Uhlig
(2002), the choice of the HP6.25 seems to be the most sensible for annual data. In fact, and for the EU country sample for instance, the associated effect is very close to the average of the effects obtained with the other filtering methods (see Table A1).

32ECB
Working Paper Series No 849
January 2008
Table A1 – Robustness Check (EU)
HP6.25 HP100 Diff BP
VE-46.337*** -50.119*** -20.426** -31.683***
(-2.74) (-3.31) (-2.13) (-2.31)
Average volatility 0.017 0.023 0.0 269 0.018
Effect -0.765 -1.134 -0.550 -0.556

Notes: t-statistics are in parenthesis.
**, *** – Statistically significant at the 5, and 1 percent level respectively.

Appendix 2 – Data Sources
Initial Output (Yo) – The log of Heston-Summers-Aten (2006) variable “Real GDP per
capita at the beginning of each time period, 2000 international prices; Laspeyres index”. Growth-rate (g) – The five-year average of the Heston-Summers-Aten (2006) annual
growth rate variable. Population Growth (n) – The average of the annual log difference of Heston-Summers-
Aten (2006) population variable. Investment Share of GDP (I/Y) – Heston-Summers-Aten (2006) variable “Real Gross
Domestic Investment, private and public, % Real GDP per capita, 2000 international prices”. Human Capital (h) – Barro-Lee (2001) variable “Average schooling years in the total
population over age 25 at the beginning of each time period”. Openness (OPEN) – Heston-Summers-Aten (2006) variable “Exports plus Imports divided by Real GDP”.
Total Revenue (R),
Direct Taxes (DIR), Indirect Taxes (IND), Social Contributions (SOC),
Total Expenditure (E), Government Consumption (GCON), Government Investment
(GINV), Transfers (TRA), Subsidies (SUB), Government Wages (W), GDP, GDP Deflator .
Source: OECD Economic Outlook, current prices.
Non-Wage Components of Government Consumption (NW) . Source: OECD Economic
Outlook and author calculations, current prices.
See the OECD Economic Outlook: Sources and Manual, for details on the definition and
construction of the variables.

33ECB
Working Paper Series No 849
January 2008
Table A2 – Data Availability
Y n h I/Y OPEN R DIR IND SOC E GCON GINV TRA SUB W NW
Australia 70 70 70 70 70 70 70 70 70 70 70 70 70 70 – –
Austria 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Belgium 70 70 70 70 70 70 70 70 70 70 70 70 85 70 70 70 Canada 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Czech Republic 90 70 90 90 90 92 93 93 93 92 92 92 92 92 92 92
Denmark 70 70 70 70 70 71 71 71 71 70 70 70 70 70 70 70
Finland 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
France 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 Germany 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 Greece 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Hungary 70 70 70 70 70 91 91 91 91 91 91 91 91 91 91 91
Iceland 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Ireland 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Italy 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Japan 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Korea 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Luxembourg 70 70 70 70 70 90 90 90 90 90 90 90 90 90 90 90
Netherlands 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
New Zealand 70 70 70 70 70 86 86 86 86 86 86 86 86 86 86 86
Norway 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Poland 70 70 70 70 70 95 95 95 95 95 95 95 95 95 95 95
Portugal70 70 70 70 70 77 77 77 77 77 77 77 77 77 77 77
Slovakia86 70 90 86 87 94 94 94 94 94 94 94 94 94 94 94
Spain 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Sweden 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 Switzerland 70 70 70 70 70 90 90 90 90 70 70 70 70 70 70 70
United Kingdom 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
United States 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 70
Note: In the table we report the first year of data availability for each variable.
Table A3 – Correlation Matrix for Fiscal Variables
R DIR IND SOC E GCON GINV TRA SUB W NW
R 1
DIR 0.67 1
IND 0.51 0.25 1
SOC 0.40 -0.23 -0.15 1
E 0.91 0.54 0.43 0.48 1
GCON 0.84 0.59 0.50 0.29 0.86 1
GINV -0.30 -0.34 -0.06 -0.11 -0.34 -0.24 1
TRA 0.72 0.40 0.41 0.25 0.60 0.47 -0.13 1
SUB 0.36 0.08 0.41 0.14 0.35 0.22 0.09 0.31 1
W 0.75 0.67 0.47 0.04 0.70 0.82 -0.22 0.44 0.29 1
NW 0.50 0.17 0.26 0.45 0.58 0.68 -0.13 0.24 0.02 0.13 1

34ECB
Working Paper Series No 849
January 2008
Tables

Table 1 – Total public expenditure as a % of GDP
(General government)

Change in pp
1970 1975 1980 1985 1990 1995 2000 2005 05-70 05-95
Australia 25.9 31.7 33.2 39.7 35.7 38.3 34.8 34.9 9.0 -3.4
Austria 39.7 46.2 49.4 53.0 51.5 56.0 51.4 49.9 10.2 -6.1
Belgium 41.0 48.9 54.7 58.3 52.2 51.9 49.0 49.8 8.8 -2.1 Canada 36.0 41.1 41.6 48.3 48.8 48.5 41.1 39.3 3.2 -9.2 Czech Republic .. .. .. .. .. 54.0 41.7 43.6 .. -10.3
Denmark .. 46.5 53.6 55.8 55.9 59.5 53.9 52.8 .. -6.7
Finland 30.9 38.5 40.1 46.4 48.0 61.5 48.3 50.1 19.2 -11.4 France 39.2 43.8 45.6 51.0 49.4 54.4 51.6 53.9 14.7 -0.6
Germany 38.4 48.7 46.9 45.1 43.6 48.3 45.1 46.8 8.4 -1.5
Greece 26.5 30.2 32.1 45.1 49.2 50.1 51.2 46.7 20.1 -3.5
Hungary .. .. .. .. .. 55.4 46.6 49.9 .. -5.5
Iceland 31.1 37.5 35.7 38.0 41.5 42.7 42.1 43.4 12.3 0.7
Ireland 44.8 49.0 54.6 54.2 43.1 41.3 31.6 34.1 -10.7 -7.1 Italy 32.5 39.3 40.8 49.8 52.9 52.5 46.1 48.2 15.8 -4.2 Japan 20.8 27.4 32.1 32.1 31.8 36.5 39.2 37.0 16.2 0.5
Korea . .. 18.7 21.2 20.0 20.0 20.8 23.9 29.1 .. 8.2
Luxembourg .. .. .. .. 37.7 39.7 37.6 43.2 .. 3.5
Netherlands 43.7 50.5 54.2 55.7 52.9 50.0 44.0 45.5 1.7 -4.6
New Zealand .. .. .. .. 53.2 42.0 39.6 38.2 .. -3.8
Norway 39.1 43.3 46.1 44.0 54.0 51.5 42.7 42.8 3.7 -8.7
Poland .. .. .. .. .. 47.7 41.0 42.7 .. -5.0
Portugal .. .. 34.2 39.3 40.3 43.1 43.1 47.7 .. 4.7
Slovakia .. .. .. .. .. 47.0 51.7 37.1 .. -9.9
Spain 23.0 25.7 33.5 42.3 42.6 44.2 39.0 38.2 15.2 -6.0
Sweden 43.9 47.4 64.1 64.8 61.3 67.1 56.8 56.3 12.4 -10.8
Switzerland . . . . 30.0 34.5 33.9 36.2 .. 1.6
United Kingdom 42.0 49.8 46.4 47.4 42.4 44.9 37.5 44.9 2.8 -0.1
United States 32.3 33.3 34.3 35.3 36.3 37.3 38.3 39.3 7.0 2.0
Minimum 20.8 18.7 21.2 20.0 20.0 20.8 23.9 29.1 8.3 8.2
Maximum 44.8 50.5 64.1 64.8 61.3 67.1 56.8 56.3 11.5 -10.8
Simple average 35.0 39.9 42.6 46.0 44.8 47.3 43.1 43.8 9.1 -3.5
Source: OECD (complemented with AMECO and National Bank of Belgium data).

Table 2 – Average volatility for Output and Fiscal Variables (HP 6.25)
EU OECD
Output 0.014 0.015
Total spending 0.017 0.054
Total revenue 0.017 0.054
Transfers 0.030 0.062
Subsidies 0.071 0.132
Government consumption 0.016 0.078
Government investment 0.057 0.106
Gov. cons. wages 0.016 0.050
Gov. cons. non-wage 0.026 0.055
Direct taxes 0.037 0.070
Indirect taxes 0.027 0.084
Social contributi ons 0.026 0.070

35ECB
Working Paper Series No 849
January 2008
Table 3 – Total General Government Revenue and Growth
(including country and period dummies)

OECD EU
0Y-4.541*** -4.196*** -3.002 -2.669
(-4.22) (-3.73) (-1.33) (-1.15)
VY-3.302 -8.434 29.987 21.393
(-0.23) (-0.56) (1.02) (0.67)
I/Y 0.130*** 0.136*** 0.102 0.105
(3.14) (3.26) (1.24) (1.28) OPEN 3.098*** 3.254*** 4.543** 4.767**
(3.09) (3.21) (2.42) (2.55) N -49.082 -68.211 -62.885 -81.559
(-1.16) (-1.48) (-0.91) (-1.07) H 0.581** 0.539** 0.451* 0.425*
(2.62) (2.39) (1.8) (1.62) FIN -3.987*** -4.058*** -4.402*** -4.440***
(-3.25) (-3.30) (-4.87) (-4.92)
GER 0.965 1.063 1.067* 1.137*
(0.82) (0.90) (1.80) (1.87)
SM 0.459 0.406 -0.878 -0.985
(1.23) (1.08) (-0.95) (-1.06)
EMU -0.404 -0.382 -0.432 -0.434 Control
variables

(-0.83) (-0.79) (-0.63) (-0.63)
R/Y -0.122*** -0.269* -0.117*** -0.259
(-3.43) (-1.88) (-2.97) (-1.51)
(R/Y)2 0.002 0.002
(1.06) (0.92)
VR 0.177 0.182 -21.669 -21.318 Revenue
variables

(0.63) (0.65) (-1.44) (-1.42)
No. obs. 159 159 95 95
R-square 0.74 0.74 0.72 0.72
Adj. R-square 0.64 0.64 0.59 0.59

Notes: t-statistics are in parenthesis. Robust standard errors to control for heteroscedasticity.
*, **, *** – Statistically significant at the 10, 5, and 1 percent level respectively.
FIN – dummy variable that assu mes the value 1 for the period 1990.
GER – dummy variable that assumes the value 1 for the period 1991.
SM – dummy variable that assumes the value 1 after year 1991 for the EU15 countries.
EMU– dummy variable that assumes the value 1 after year 1998 for the EMU countries.
The HP6.25 filter was used to decompose the series.

36ECB
Working Paper Series No 849
January 2008Table 4 – Total General Government Expenditure and Growth
(including country and period dummies)
OECD EU
0Y -5.399*** -5.103*** -4.210** -4.017**
(-5.17) (-4.85) (-2.20) (-2.16)
VY -5.028 -10.572 29.422 17.701
(-0.27) (-0.77) (1.04) (0.54)
I/Y 0.115*** 0.113*** 0.056 0.056
(2.91) (2.89) (0.78) (0.79) OPEN 2.988*** 3.109*** 4.348*** 4.735**
(3.16) (3.30) (2.61) (2.86) N -68.675* -88.000** -38.436 -61.258
(-1.70) (-2.10) (-0.61) (-0.83) H 0.672*** 0.658** 0.293 0.302
(3.17) (3.12) (1.19) (1.20) FIN -3.398 *** -3.505*** -4.303*** -4.289***
(-2.89) (-3.00) (-5.02) (-4.82)
GER 0.830 0.962 0.903* 1.017*
(0.74) (0.86) (1.80) (1.91)
SM 0.486 0.407 1.626 -0.581
(1.38) (1.15) (1.47) (-0.38)
EMU -0.431 -0.437 -0.390 -0.466
Control
variables

(-0.94) (-0.96) (-0.46) (-0.57)
E/Y -0.130*** -0.304*** -0.085** -0.270**
(-5.12) (-2.78) (-2.39) (-1.96)
(E/Y)2 0.002* 0.002
(1.63) (1.46)
VE 0.139 0.142 -46.337*** -43.372*** Expenditure
variables

(0.53) (0. 54) (-2.74) (-2.63)
No. obs. 159 159 95 95
R-square 0.77 0.77 0.77 0.77
Adj. R-square 0.68 0.68 0.66 0.66
Notes: t-statistics are in parenthesis. Robust standard errors to control for heteroscedasticity.
*, **, *** – Statistically significant at the 10, 5, and 1 percent level respectively.
FIN – dummy variable that assu mes the value 1 for the period 1990.
GER – dummy variable that assumes the value 1 for the period 1991.
SM – dummy variable that assumes the value 1 after year 1991 for the EU15 countries.
EMU– dummy variable that assumes the value 1 after year 1999 for the EMU countries.
The HP6.25 filter was used to decompose the series.

37ECB
Working Paper Series No 849
January 2008
Table 5 – General Government Revenue Composition and Growth
(including country and period dummies)
OECD EU
Controls n.r. n.r. n.r. n.r.
DIR -0.039 -0.101 0.095 0.124
(-0.80) (-0.08) (1.27) (0.68)
DIR2 0.002 -0.001
(0.52) (-0.16) IND -0.297*** -0.126 -0.404*** 0.117
(-3.80) (-0.35) (-3.13) (0.21) IND
2 -0.006 -0.022
(-0.46) (-1.08) SOC -0.338** -0.687*** -0.382*** -0.570*
(-3.77) (-2.71) (-3.20) (-1.88) SOC
2 0.014* 0.007
(1.66) (0.75)
VDIR 4.209 5.015 0.636 0.333
(1.25) (1.42) (0.12) (0.06)
VIND-0.821*** -0.795*** -10.816 -11.070
(-4.81) (-4.62) (-1.12) (-1.13)
VSOC-2.944 -3.782 -10.552** -8.918 Revenue
variables
(-0.86) (-1.05) (-2.14) (-1.53)
No. obs. 153 153 95 95
R-square 0.80 0.81 0.78 0.79
Adj. R-square 0.72 0.72 0.66 0.65

Notes: t-statistics are in parenthesis. Robust standard errors to control for heteroscedasticity. *, **, *** – Statistically significant at the 10, 5, and 1 percent level respectively.
n.r. – not reported. Full results regarding these variables are available upon request.
The HP6.25 filter was used to decompose the series.

38ECB
Working Paper Series No 849
January 2008
Table 6 – General Government Expenditure Composition and Growth
(including country and period dummies)

OECD EU
Controls n.r. n.r. n.r. n.r.
TRA 0.039 -0.831 0.386** -0.134
(0.20) (-1.14) (2.09) (-0.20)
TRA2 0.083 0.053
(1.12) (0.71)
SUB -0.438** -0.868* -0.708*** -1.062*
(-2.24) (-1.83) (-3.84) (-1.69) SUB
2 0.106 0.082*
(0.94) (0.50) GCON -0.227*** -0.943*** -0.313*** -0.719**
(-2.61) (-2.91) (-4.41) (-2.46)
GCON
2 0.020** 0.011
(2.47) (1.50)
GINV -0.019 -0.141 -0.164 0.277
(-0.10) (-024) (-0.74) (0.46)
GINV2 0.040 -0.048
(0.47) (-0.54)
TRAV 1.467* 1.059 -5.215 -6.334
(1.89) (1.55) (-1.11) (-1.22)
SUBV -0.370 0.706 -2.193 -1.946
(-0.17) (0.32) (-0.89) (-0.69)
GCONV 1.558 1.529 -26.343** -24.993*
(0.46) (0.36) (-2.14) (-1.89)
GINVV -2.605 -3.552 -8.021* -8.653** Expenditure variables

(-0.62) (-0.76) (-1.74) (-1.96)
No. obs. 148 148 85 85
R-square 0.77 0.79 0.87 0.89 Adj. R-square 0.66 0.67 0.79 0.79

Notes: t-statistics are in parenthesis. Robust standard errors to control for heteroscedasticity.
*, **, *** – Statistically significant at the 10, 5, and 1 percent level respectively.
n.r. – not reported. Full results regarding these variables are available upon request. The HP6.25 filter was used to decompose the series.

39ECB
Working Paper Series No 849
January 2008
Table 7 – General Government Expenditure Composition (wage and non-wage
disaggregation) and Growth (including country and period dummies)

OECD EU
Controls n.r. n.r. n.r. n.r.
TRA -0.192 -1.022 0.302 -0.060
(-0.09) (-1.37) (1.35) (-0.09)
TRA2 0.101 0.044
(1.33) (0.56) SUB -0.415** -0.971** -0.727*** -.0.686
(-2.07) (-2.15) (-3.73) (-0.96)
SUB
2 0.102 -0.022
(0.93) (-0.13) GWAGE -0.256** -1.679*** -0.318*** -1.534***
(-2.56) (-3.78) (-3.17) (-2.99) GWAGE
2 0.060*** 0.042**
(3.42) (2.44)
GNONWAGE -0.252* -0.211 -0.429** 0.307
(-1.87) (-0.42) (-2.42) (0.57)
GNONWAGE2 -0.001 -0.042
(-0.04) (-1.43)
GINV 0.048 0.436 -0.141 0.829
(0.25) (0.66) (-0.63) (1.13)
GINV2 -0.035 -0.116
(-0.36) (-1.16)
TRAV -0.655 -1.747 -7.740 -6.759
(-0.14) (-0.36) (-1.53) (-1.17)
SUBV -0.595 -0.329 -3.070 -3.172
(-0.28) (-0.16) (-1.25) (-1.26)
GWAGEV 1.851 2.624 -3.070 -5.209
(0.47) (0.66) (-0.22) (-0.39)
GNONWAGEV -0.010 0.019 -16.715* -11.705
(-0.09) (0.14) (-1.66) (-1.17)
GINVV -0.444 0.643 -7.47 -6.377 Expenditure
variables

(-0.24) (-0.39) (-1.52) (-1.22)
No. obs. 141 141 85 85
R-square 0.78 0.82 0.88 0.89
Adj. R-square 0.66 0.71 0.78 0.79

Notes: t-statistics are in parenthesis. Robust standard errors to control for heteroscedasticity.
*, **, *** – Statistically significant at the 10, 5, and 1 percent level respectively.
n.r. – not reported. Full results regarding these variables are available upon request.
The HP6.25 filter was used to decompose the series.

40ECB
Working Paper Series No 849
January 2008
Table 8 – Total General Government Expenditure and Revenue and Growth
(Including Only Period Dummies)

OECD EU
0Y-1.609*** -1.509*** -1.067 -1.895**
(-4.02) (-4.13) (-1.24) (-2.37)
VY5.931 1.767 49.467** 47.587
(0.42) (0.13) (2.02) (2.12)
I/Y 0.104*** 0.0655*** 0.044 0.005
(4.28) (2.62) (1.04) (0.12) OPEN 2.965*** 3.427*** 3.662** 4.357***
(3.06) (3.88) (2.51) (3.16) N 5.877 -16.657 -20.398 -22.869
(0.22) (-0.64) (-0.50) (-0.59) H 0.224*** 0.1.83*** 0.106 0.132
(3.06) (2.67) (0.81) (1.17) FIN -4.580*** -4.132*** -5.006*** -4.829***
(-3.45) (-3.27) (-3.77) (-3.90)
GER 0.364 0.384 0.954 0.754
(0.29) (0.32) (0.78) (0.67)
SM 1.239*** 1.203*** 1.888** 1.660**
(3.78) (3.87) (2.26) (2.15)
EMU -0.033 -0.160 -0.466 -0.453 Control
variables

(-0.07) (-0.33) (-0.61) (-0.64)
R/Y -0.053*** -0.055**
(-3.06) (-2.10)
VR 0.028 -10.043
(0.012) (-0.70)
E/Y -0.081*** -0.067***
(-5.03) (-2.94)
VE 0.034 -36.072** Fiscal
variables

(0.16) (-2.46)
No. obs. 159 159 95 95
R-square 0.57 0.61 0.60 0.66
Adj. R-square 0.51 0.56 0.51 0.58

Notes: t-statistics are in parenthesis. Robust standard errors to control for heteroscedasticity.
*, **, *** – Statistically significant at the 10, 5, and 1 percent level respectively.
FIN – dummy variable that assu mes the value 1 for the period 1990.
GER – dummy variable that assumes the value 1 for the period 1991.
SM – dummy variable that assumes the value 1 after year 1991 for the EU15 countries.
EMU– dummy variable that assumes the value 1 after year 1998 for the EMU countries
The HP6.25 filter was used to decompose the series.

41ECB
Working Paper Series No 849
January 2008

Table 9 – Total General Government Expenditure and Revenue and Growth, Robustness
Check for Endogeneity (including country and period dummies)

OECD EU
0Y-2.216*** -2.596*** -0.644 -5.477***
(-2.22) (-3.46) (-0.38) (-3.53)
VY-25.487 -20.957 14.482** 35.471
(-1.47) (-1.28) (0.46) (1.18)
I/Y 0.148*** 0.110** 0.143* 0.132
(2.88) (2.23) (1.80) (1.64) OPEN 4.174* 6.728*** 10.516** 0.554
(1.82) (4.24) (2.07) (0.16) N -50.103 -47.595 -79.333 -33.850
(-1.10) (-1.15) (-1.36) (-0.58) H 0.893*** 0.553*** 0.313 0.632***
(3.54) (4.22) (1.21) (2.72) FIN -4.497*** -4.422*** -4.682*** -5.073***
(-3.68) (-3.70) (-3.36) (-3.60)
GER 0.427 0.734 0.921 1.072
(0.36) (0.64) (0.78) (0.83)
SM 1.276*** 0.839** 1.023 2.085**
(3.34) (2.29) (1.03) (2.40)
EMU -0.330 -0.003 -0.026 0.405 Control
variables

(-0.67) (-0.01) (-0.03) (0.50)
R/Y
0 -0.071** -0.079**
(-2.14) (-2.31)
1RV -23.662** -23.556
(-2.02) (-1.44)
E/Y 0 -0.068*** -0.064**
(-3.16) (-1.99)
1EV
-35.752*** -35.072* Fiscal
variables

(-2.81) (-1.75)
No. obs. 159 159 95 95
R-square 0.57 0.61 0.60 0.66
Adj. R-square 0.51 0.56 0.51 0.58

Notes: t-statistics are in parenthesis. Robust standard errors to control for heteroscedasticity.
*, **, *** – Statistically significant at the 10, 5, and 1 percent level respectively.
FIN – dummy variable that assu mes the value 1 for the period 1990.
GER – dummy variable that assumes the value 1 for the period 1991.
SM – dummy variable that assumes the value 1 after year 1991 for the EU15 countries.
EMU– dummy variable that assumes the value 1 after year 1998 for the EMU countries
The HP6.25 filter was used to decompose the series.
R/Y 0, E/Y 0 – ratios for the first year of each of the 5-year periods.

42ECB
Working Paper Series No 849
January 2008European Central Bank Working Paper Series
For a complete list of Working Papers published by the ECB, please visit the ECB’s website
(http://www.ecb.europa.eu).
817 “Convergence and anchoring of yield curves in the euro area” by M. Ehrmann, M. Fratzscher, R. S. Gürkaynak
and E. T. Swanson, October 2007.
818 “Is time ripe for price level path stability?” by V. Gaspar, F. Smets and D. Vestin, October 2007.819 “Proximity and linkages among coalition participants: a new voting power measure applied to the International
Monetary Fund” by J. Reynaud, C. Thimann and L. Gatarek, October 2007.
820 “What do we really know about fi scal sustainability in the EU? A panel data diagnostic” by A. Afonso
and C. Rault, October 2007.
821 “Social value of public information: testing the limits to transparency” by M. Ehrmann and M. Fratzscher,
October 2007.
822 “Exchange rate pass-through to trade prices: the role of non-linearities and asymmetries” by M. Bussière,
October 2007.
823 “Modelling Ireland’s exchange rates: from EMS to EMU” by D. Bond and M. J. Harrison and E. J. O’Brien,
October 2007.
824 “Evolving U.S. monetary policy and the decline of in fl ation predictability” by L. Benati and P. Surico,
October 2007.
825 “What can probability forecasts tell us about in fl ation risks?” by J. A. García and A. Manzanares, October 2007.
826 “Risk sharing, fi nance and institutions in international portfolios” by M. Fratzscher and J. Imbs, October 2007.
827 “How is real convergence driving nominal convergence in the new EU Member States?” by S. M. Lein-Rupprecht,
M. A. León-Ledesma, and C. Nerlich, November 2007.
828 “Potential output growth in several industrialised countries: a comparison” by C. Cahn and A. Saint-Guilhem,
November 2007.
829 “Modelling in fl ation in China: a regional perspective” by A. Mehrotra, T. Peltonen and A. Santos Rivera,
November 2007.
830 “The term structure of euro area break-even in fl ation rates: the impact of seasonality” by J. Ejsing, J. A. García
and T. Werner, November 2007.
831 “Hierarchical Markov normal mixture models with applications to fi nancial asset returns” by J. Geweke
and G. Amisano, November 2007.
832 “The yield curve and macroeconomic dynamics” by P. Hördahl, O. Tristani and D. Vestin, November 2007.833 “Explaining and forecasting euro area exports: which competitiveness indicator performs best?” by M. Ca’ Zorzi
and B. Schnatz, November 2007.
834 “International frictions and optimal monetary policy cooperation: analytical solutions” by M. Darracq Pariès,
November 2007.
835 “US shocks and global exchange rate con fi gurations” by M. Fratzscher, November 2007.

43ECB
Working Paper Series No 849
January 2008836 “Reporting biases and survey results: evidence from European professional forecasters” by J. A. García
and A. Manzanares, December 2007.
837 “Monetary policy and core in fl ation” by M. Lenza, December 2007.
838 “Securitisation and the bank lending channel” by Y. Altunbas, L. Gambacorta and D. Marqués, December 2007.
839 “Are there oil currencies? The real exchange rate of oil exporting countries” by M. M. Habib
and M. Manolova Kalamova, December 2007.
840 “Downward wage rigidity for different workers and fi rms: an evaluation for Belgium using the IWFP procedure”
by P. Du Caju, C. Fuss and L. Wintr, December 2007.
841 “Should we take inside money seriously?” by L. Stracca, December 2007.842 “Saving behaviour and global imbalances: the role of emerging market economies” by G. Ferrucci and C. Miralles,
December 2007.
843 “Fiscal forecasting: lessons from the literature and challenges” by T. Leal, J. J. Pérez, M. Tujula and J.-P. Vidal,
December 2007.
844 “Business cycle synchronization and insurance mechanisms in the EU” by A. Afonso and D. Furceri,
December 2007.
845 “Run-prone banking and asset markets” by M. Hoerova, December 2007.846 “Information combination and forecast (st)ability. Evidence from vintages of time-series data” by C. Altavilla
and M. Ciccarelli, December 2007.
847 “Deeper, wider and more competitive? Monetary integration, Eastern enlargement and competitiveness in the
European Union” by G. Ottaviano, D. Taglioni and F. di Mauro, December 2007.
848 “Economic growth and budgetary components: a panel assessment for the EU” by A. Afonso
and J. González Alegre, January 2008.
849 “Government size, composition, volatility and economic growth” by A. Afonso and D. Furceri, January 2008.