WO R K I N G PA P E R S E R I E S
N O 1 3 4 4 / M AY 2 011
HAVE EURO AREA
AND EU ECONOMIC
GOVERNANCE
WORKED?
JUST THE FACTS
by Demosthenes Ioannou
and Livio Stracca
WO R K I N G PA P E R S E R I E S
N O 13 4 4 / M AY 2011
HAVE EURO AREA
AND EU ECONOMIC
GOVERNANCE WORKED?
JUST THE FACTS 1
by Demosthenes Ioannou
and Livio Stracca 2
NOTE: This Working Paper should not be reported as representing
the views of the European Central Bank (ECB).
The views expressed are those of the authors
and do not necessarily reflect those of the ECB.
In 2011 all ECB
publications
feature a motif
taken from
the €100 banknote.
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=1846328.
1 We thank participants at a seminar at the ECB, in particular G. Gloeckler, F. Mongelli, P. Mohl and P. Rother; Z. Truchlewski,
D. Skotkova, A. Ghione, J.-P. Vidal and an anonymous referee for useful suggestions and input.
2 Both authors: European Central Bank, DG International and European Relations,
Kaiserstrasse 29, D-60311 Frankfurt am Main, Germany;
e-mails: [email protected]
and [email protected]
© European Central Bank, 2011
Address
Kaiserstrasse 29
60311 Frankfurt am Main, Germany
Postal address
Postfach 16 03 19
60066 Frankfurt am Main, Germany
Telephone
+49 69 1344 0
Internet
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 authors.
Information on all of the papers published
in the ECB Working Paper Series can be
found on the ECB’s website, http://www.
ecb.europa.eu/pub/scientific/wps/date/
html/index.en.html
ISSN 1725-2806 (online)
CONTENTS
Abstract
4
Non-technical summary
5
1 Introduction
6
2 Literature review: the stability
and growth pact
8
3 Data
9
4 A first look at the evidence
11
5 The stability and growth pact
and fiscal behaviour
5.1 Empirical model
5.2 Results
11
11
13
6 The Lisbon strategy
and economic performance
6.1 Empirical model
6.2 Results
15
17
18
7 Conclusions
20
References
21
Tables
26
ECB
Working Paper Series No 1344
May 2011
3
Abstract
We test whether two key elements of the EU and euro area economic governance framework, the Stability and Growth Pact and the Lisbon Strategy, have
had any impact on macroeconomic outcomes. We test this proposition using a
di¤erence-in-di¤erence approach on a panel of over 30 countries, some of which
are non-EU (control group). Hence, the impact of the EU economic governance
pillars is evaluated based on both the performance before and after their application as well as against the control group. We …nd strong and robust evidence
that neither the Stability and Growth Pact nor the Lisbon Strategy have had
a signi…cant bene…cial impact on …scal and economic performance outcomes.
We conclude that a profound reform of these pillars is needed to make them
work in the next decade.
Keywords: Stability and Growth Pact, Lisbon Strategy, euro area, European Union, governance, institutions.
JEL:
4
ECB
Working Paper Series No 1344
May 2011
Non-technical summary
This paper presents an ex post analysis of whether two key elements of the EU and euro
area economic governance framework, the Stability and Growth Pact (SGP) and the
Lisbon Strategy, have achieved their goals. Using a difference-in-difference approach in
a panel setting, we look at a wide range of annual data coming from 36 countries, over a
sample period from 1980 to 2010. We consider a large set of possible control variables,
as well as variables which could interact with the success (or lack thereof) of the two EU
governance pillars. Their performance EU is assessed both against the countries’ own
past performance as well as a control group of non-euro area or non-EU countries,
controlling for different national economy characteristics.
Overall, our results indicate that so far economic governance in the EU and the euro area
has had limited or no success. In particular, we find that
1) The Stability and Growth Pact has had no overall effect on the behaviour of the
primary balance. While it has increased the counter-cyclicality of fiscal policy it has also
increased its sensitivity to the political business cycle;
2) The Lisbon Strategy has had at best no impact on the behaviour of real per capita GDP
growth, employment growth and labour productivity growth.
For the SGP, an optimistic reading of our results is that finding no effect of the SGP on
fiscal outcomes is an indicator of success since it implies that this institutional framework
has prevented the establishment of Economic and Monetary Union to adversely affect
fiscal behaviour, a risk that was emphasised widely in the run-up to the euro. However,
we consider that an appropriate set of fiscal rules in a monetary union should go beyond
the no-change outcome and impose greater fiscal discipline than otherwise, on account of
the possible negative spillovers that fiscal profligacy in individual countries may have on
other EMU participants whether via the single monetary policy or otherwise, as well as
the possible reduction in market discipline for individual Member States brought about
by a stable and solid single currency. Our results indicate that the SGP has not delivered
according to this stricter benchmark.
The main policy implication stemming from our analysis is that substantial progress, or a
“quantum leap”, must be made in EU and euro area economic governance to ensure and
enhance the gains of Economic and Monetary Union for the benefit of European citizens.
ECB
Working Paper Series No 1344
May 2011
5
1
Introduction
In the aftermath of the global …nancial crisis euro area Member States have experienced unprecedented challenges to their public …nances in particular, and their
economic policies more generally. These challenges have exposed the weaknesses of
the two pillars of EU and euro area economic governance, namely the …scal framework of the Maastricht Treaty and the Stability and Growth Pact (SGP) on the one
hand, and the Lisbon Strategy (LS) on the other. Looking forward, it is essential to
strengthen both pillars of EU economic governance. Before we look forward, however,
we must set the record straight on the performance of these economic governance pillars looking backward, in their …rst decade. This paper tries to answer this question:
have the SGP and the LS worked against the objectives that they were created for?
One main contribution of this paper is to extend beyond the EU and use non-EU
OECD countries as a control group in an econometric investigation of the e¤ectiveness of the SGP, applying a di¤erence-in-di¤erence approach where the outcomes are
compared both in terms of their own past, for the group of countries which have been
subject to the "treatment" (the EU economic governance pillars), as well as the performance of the countries in the control group, also taking into account the in‡uence
of a set of control variables. This approach goes beyond most contributions so far on
the SGP which typically focus on EMU or EU countries alone (e.g. Gali and Perotti
2003, Annett 2006, Golinelli and Momigliano 2009, Bernoth et al. 2009). Moreover,
existing studies usually consider signi…cantly shorter samples (among those that are
most related to our paper, Annett ends in 2004, and Gali and Perotti in 2002). Instead, we use annual data spanning from 1980 and up to 2009, covering 36 countries.
Finally, we test not only for the e¤ect of the SGP on average …scal behaviour, but
also extend our analysis to other possible dimensions of its in‡uence, such as the
degree of pro-cyclicality of …scal policy, its being subject to the political cycle, and
its responsiveness to market discipline.1
We also apply the same di¤erence-in-di¤erence approach vis-à-vis the non-EU
OECD control group for the LS. In the case of the LS, our assessment comes at the
time of conclusion of the original reform agenda and when the European Union is
now ironing out the details of its successor strategy, Europe 2020. In this analysis, we
also include a number of control variables that have come up in the literature on the
political economy of structural reform. Unlike previous studies, we do not focus on
reform e¤orts but rather on outcomes in terms of long term economic performance
measures. To our knowledge, this type of exercise has not been performed before, at
least in a systematic manner.
Our paper is related to several di¤erent strands of the …scal policy literature,
in particular to three of them. One is the political economy of …scal policy. For
example, it has been emphasised that …scal pro‡igacy may depend on the size and
1
However, we don’t consider the complementarities between …scal policy and structural reforms
(as, for example, in Buti et al. 2009).
6
ECB
Working Paper Series No 1344
May 2011
=FID F= >FM<IED<EK LK@ 8E; *<E:? ,FL9@E@ 8E; -8:?J K?< ;<J@>E F=
<C<:KFI8C JPJK<DJ C<J@E8 8E; .89<CC@E@ 8E; *<IJJFE 8E; -M<EJJFE K?<
@DG8:K F= G8IK@J8E GFC@K@:J FE RJ:8C GFC@:P &8D9<IK@E@ K?< =I8>D<EK8K@FE F=
K?< 9L;><K8IP GIF:<JJ "8CC<I9<I> 8E; K?< GFC@K@:J F= 9L;><K D8O@D@J8K@FE 8E;
RJ:8C @CCLJ@FE I8Q<E (LD<IFLJ I<8JFEJ =FI K?< T;<R:@K 8E; ;<9K 9@8J<JU ?8M<
8CJF 9<<E @;<EK@R<; =FI 8E FM<IM@<N J<< -:?L:BE<:?K I8Q<E 8E; C<J@E8
8E; *<IFKK@ N?@:? I8E>< =IFD K?< <Z<:KJ F= K?< <C<:KFI8C :P:C< KF <O:<JJ@M<CP
FGK@D@JK@: =FI<:8JKJ $FELE> 8E; &8I:? 8E; 9L;><K :8GKLI< 9P GI<JJLI< >IFLGJ
C<J@E8 8E; I8Q<E 1< @E:CL;< J<M<I8C GFC@K@:8C <:FEFDP M8I@89C<J @E FLI
8E8CPJ@J
J<:FE; JKI8E; F= C@K<I8KLI< CFFBJ 8K ;4%#. 36.'4 8E; RJ:8C :FLE:@CJ @J:8C ILC<J
;<RE<; 8J =# 2'3/#0'05 %10453#+05 10 ;4%#. 21.+%: '923'44'& +0 5'3/4 1( # 46//#3:
+0&+%#513 1( ;4%#. 2'3(13/#0%'> %FG@KJ 8E; -PD8EJBP 8I< E<<;<; KF ;<8C N@K?
K?< GIF9C<D F= K@D< @E:FEJ@JK<E:P .?< 98J@: GI@E:@GC< @J VKP@E> FE<J<C= KF K?< D8JKV
-:?<CBC< @E FI;<I KF 8MF@; FM<IJG<E;@E> FM<I K?< 9LJ@E<JJ :P:C< #E K?@J G8G<I
N< CFFB 8K K?< -!* 8J 8 JLGI8
E8K@FE8C ILC< 8E; ;F EFK =F:LJ 8CJF =FI C8:B F= ;8K8
=FI FLI C8I><I J8DGC< F= :FLEKI@<J FE 0#5+10#. RJ:8C @EJK@KLK@FEJ JL:? 8J @E <9ILE
8E; %LD8I 8E; K?< GFJJ@9C< @EK<IGC8P N@K? K?< JLGI8
E8K@FE8C ILC<
.?@I; K?<I< @J 8 J@>E@R:8EK C@K<I8KLI< FE K?< -!* 8J JL:? 1< I<GFIK FE K?@J
C@K<I8KLI< 8E; @KJ I<C8K@FE N@K? FLI NFIB @E 8 J<G8I8K< J<:K@FE -<:K@FE )LI D8@E I<JLCK @J K?8K K?< -!* 8E; K?< &- ?8M< EFK J@>E@R:8EKCP @DGIFM<; K?<
G<I=FID8E:< F= / '<D9<I -K8K<J @E K<IDJ F= RJ:8C GFC@:P 8E; CFE> K<ID <:FEFD@:
G<I=FID8E:< )LI I<JLCK =FI K?< -!* EF J@>E@R:8EK @DG8:K F= K?< -!* FE K?< 8M<I8><
9<?8M@FLI F= K?< GI@D8IP 98C8E:< C<E;J @KJ<C= KF 8E FGK@D@JK@: @EK<IGI<K8K@FE J@E:<
K?< <JK89C@J?D<EK F= 8 DFE<K8IP LE@FE :FLC; ?8M< 2'3 4' @< N@K?FLK K?< -K89@C@KP
8E; !IFNK? *8:K C<; KF 8 8'#-'0+0) F= RJ:8C ;@J:@GC@E< 8E; KF C8I><I ;<R:@KJ @E
89JFCLK< K<IDJ .?<I<=FI< FE< GFJJ@9C< I<8;@E> F= FLI RE;@E> @J K?8K K?< -!* N8J
8K C<8JK 89C< KF GI<M<EK K?@J =IFD ?8GG<E@E> 8CK?FL>? FLI 8E8CPJ@J @J EFK 98J<; FE
8 JKIL:KLI8C DF;<C 8E; N< :8E K?<I<=FI< ILE EF :FLEK<I
=8:KL8C "FN<M<I N< RE;
K?@J GFJJ@9C< @EK<IGI<K8K@FE EFK M<IP GC8LJ@9C< 8E; KFF :FDGC8:<EK J@E:< 8EFK?<I B<P
F9A<:K@M< F= K?< -!* @J KF GI<M<EK RJ:8C @II<JGFEJ@9@C@KP @E / '<D9<I -K8K<J =IFD
JG@CC@E> FM<I KF K?< J@E>C< DFE<K8IP GFC@:P .?@J NFLC; :8CC =FI /13' 453+0)'05 C@D@KJ
FE RJ:8C GFC@:P @E 8 DFE<K8IP LE@FE K?8E FK?<IN@J< 8E; FE K?@J 8::FLEK FLI I<JLCKJ
@E;@:8K< K?8K K?< -!* ?8J EFK ;<C@M<I<; )M<I8CC FLI I<JLCKJ JKI<E>K?<E K?< :8J< =FI
8 UHL8EKLD C<8GU @E K?< I<=FID F= K?< / 8E; <LIF 8I<8 <:FEFD@: >FM<IE8E:<
.?< G8G<I @J FI>8E@J<; 8J =FCCFNJ -<:K@FE I<M@<NJ JFD< C@K<I8KLI< FE K?< -!*
-<:K@FE ;<J:I@9<J K?< ;8K8 N?@C< -<:K@FE >@M<J 8 GI<C@D@E8IP CFFB 8K K?< <M@;<E:<
-<:K@FE CFFBJ 8K K?< IFC< F= K?< -!* 8E; K?< '88JKI@:?K .I<8KP KF :FEJKI8@E 8E;
@ESL<E:< RJ:8C 9<?8M@FLI -<:K@FE ILEJ 8 J@D@C8I 8E8CPJ@J =FI K?< &- -<:K@FE :FE:CL;<J 8E; GI<J<EKJ JFD< GFC@:P @DGC@:8K@FEJ
ECB
Working Paper Series No 1344
May 2011
7
2
Literature review: the Stability and Growth Pact
There is already a copious literature on the SGP, to which for reasons of space we
cannot do full justice. Here we brie‡y review a few of the issues that are most relevant
for our analysis.
Buti, Eij¢ nger and Franco note (2003) that the Maastricht convergence criteria
and, later, the SGP’s 3% de…cit limit appeared acceptable to enforce …scal discipline
at a time when public …nances in a number of EU countries appeared to be on an
unsustainable path. In 1992, the EU’s average debt ratio was almost 60% of GDP;
by 1997, it had climbed to almost 75%. While this ratio fell to 63% in 2003, von
Hagen (2003) argues that this result cannot be directly attributed to the SGP and
is subject to two quali…cations. First, the increase in the average debt ratio from
1992-1997 was driven mainly by debt expansion in only …ve countries. Second, the
decrease in the average debt ratio from 1997 to 2001 again saw small states outperform
large states, achieving a reduction in their debt ratios of almost 20 percentage points
(as against the average 5.3% reduction in large states’ debt ratios). Annett (2006)
uses these …ndings to re…ne the argument that the SGP is inherently more suited to
small countries, suggesting that the SGP “could be suited to a subgroup of countries
that (i) are small and more likely to accept an external constraint; (ii) have the
potential for macroeconomic volatility and so appreciate an external anchor; and (iii)
rely on the commitment form of …scal governance.” Buiter (2005), writing on the
same period, states more bluntly that as regards sustainability “the SGP has made
a contribution. . . only where its prescriptions were incentive-compatible for the target
country, that is, aligned with that country’s domestic policy objectives.”
As noted by von Hagen (2003), since most European countries had sizeable …scal
expansions during the 1970s and 1980s, a period of consolidation could be expected
in the 1990s irrespective of the Maastricht criteria or the SGP’s strictures. Thus,
we can interpret what Fatás and Mihov (2003) described as countries’"consolidation
fatigue" as an example of diminished incentive-compatibility. The ECB (2005) (see
also Morris et al. 2006) noted that, from 1999, …scal consolidation stalled or went into
reverse in most euro-area countries. Economic downturn in 2001 led to deterioration
in public …nances, putting an increasing number of Member States at risk of, or
…rmly in, excessive de…cit positions. Troubles in France and Germany led to the
"suspension" of the SGP and its eventual reform in 2005, a development viewed with
alarm by both the European Commission and the ECB.
More generally, Filipek and Schreiber (2010) state that until the onset of the
global …nancial crisis, the SGP appeared to be successful: countries were meeting their
Medium-Term Objectives (MTOs) and most had balanced budgets or even surpluses.
However, it would seem wrong to attribute this to the 2005 reform which in fact
weakened the Pact. Indeed, Buiter (2005) and Filipek and Schreiber agree that the
SGP did not provide the incentives for necessary restraint during upswings to create
room for expansionary measures during downturns. Furthermore, von Hagen (2003)
8
ECB
Working Paper Series No 1344
May 2011
noted that the narrow focus of the SGP on annual de…cits may keep governments
from adopting reform policies that might result in larger de…cits initially before the
desired growth and employment e¤ects kick in.
The implementation of the SGP has obviously been followed and evaluated also
by international institutions. The IMF (2010) emphasised how the crisis had exposed
three long-standing weaknesses in the euro area’s …scal framework. First, the SGP
had failed to encourage the buildup of su¢ cient bu¤ers in good times and lower debt
to prudent levels, limiting room for manoeuvre in bad times. Second, …scal surveillance’s narrow focus on procedural aspects and formal de…cit limits, twinned with
Council’s reluctance to use binding legal instruments to mandate policy corrections in
EDP enforcement, aggravated structural ‡aws. Third, the euro area …scal framework
lacked centralized crisis management and resolution capacities. The IMF (2010) thus
advocated a strengthening of economic governance in EMU with a focus on enforcing
budgetary discipline.
Finally, other research has considered more closely the connection between the
SGP and national …scal institutions and rules. Although the econometric analysis in
this paper does not delve into national …scal frameworks (as does for example Debrun
and Kumar 2007), it is worth considering this aspect of the SGP in terms of relevance
for our conclusions.
From a ’…scal institutionalist’perspective, the SGP should be more successfully
combined with “commitment”member states because it strengthens their rules-based
frameworks; while the legitimacy of delegation states’ministers of …nance is undermined by the SGP because they no longer have room for manoeuvre (Hallerberg
2004). However, country speci…c empirical evidence does not always …t this model
(Hodson 2009). Moreover, factors such as the role of veto players in the budgetary
process (e.g. German Bundesrat), or the degree of public spending decentralisation
(Afonso and Hautpmeier 2009) also seem to a¤ect …scal outcomes.
In terms of the existing empirical research on the e¤ects of the Stability and
Growth Pact, Table 1 presents an overview of existing studies and key results. The
general message arising from this literature (see e.g. Gali and Perotti 2003) is that
the SGP does not seem to have had a major impact on …scal behaviour in the euro
area. One important di¤erence with our approach, as noted, is that these studies are
not based on di¤erence-in-di¤erence estimates and cover signi…cantly smaller sample
periods, typically up to the early to mid-2000s.
3
Data
The empirical analysis in the paper is based on annual data from 36, mostly advanced,
countries, a list of which is contained in Table 2. The sample includes 25 EU countries,
15 of which are in the euro area now, and 11 non-EU countries. Since most variables
are not available for all countries, in the regression analyses the country sample
numbers between 30 and 34. The sample period is 1980 to 2009.
ECB
Working Paper Series No 1344
May 2011
9
ELD9<I F= D8:IF<:FEFD@: M8I@89C<J ?8M< 9<<E F9K8@E<; =IFD FZ K?< J?<C=
JFLI:<J K?< >FM<IED<EK 9L;><K 98C8E:< 8E; ;<9K 8J 8 J?8I< F= !* @E:FD< G<I
:8G@K8 <DGCFPD<EK K?< C89FLI J?8I< F= @E:FD< 8E; C89FLI GIF;L:K@M@KP =IFD K?<
LIFG<8E FDD@JJ@FE ') ;8K898J< K?< FLKGLK >8G K?< GI@D8IP 8E; :P:C@
:8CCP 8;ALJK<; GI@D8IP 98C8E:< K?< >FM<IED<EK J?8I< F= @E:FD< KFK8C >FM<IED<EK
;@J9LIJ<D<EKJ 8J 8 J?8I< F= !* 8E; KI8;< FG<EE<JJ =IFD K?< ) :FEFD@:
)LKCFFB ;8K898J< GFGLC8K@FE 8E; K?< *** J?8I< F= NFIC; !* =IFD K?< #' 1FIC;
:FEFD@: )LKCFFB ;8K898J< ,<J<8I:? 8E; <M<CFGD<EK <OG<E;@KLI< 8J 8 J?8I< F=
!* =IFD K?< ) -.( ;8K898J< K?< ,LC< F= &8N @E;@:8KFI =IFD K?< 1FIC;
8EB !FM<IE8E:< #E;@:8KFIJ ;8K898J< 8M8@C89C< =IFD FEN8I;J 8E; LG;8K<; KF
J<< %8L=D8EE <K 8C 8E; 8 ELD9<I F= M8I@89C<J :8GKLI@E> K?< GFC@K@
:8C 9LJ@E<JJ :P:C< 8E; GFC@K@:8C @EJK@KLK@FEJ =IFD K?< LG;8K< F= K?< 8K898J<
F= *FC@K@:8C #EJK@KLK@FEJ J<< <:B <K 8C @E8CCP N< 8CJF LJ< ) ;8K8
FE DGCFPD<EK *IFK<:K@FE &<>@JC8K@FE 8E; *IF;L:K '8IB<K ,<>LC8K@FE K?< C8KK<I
@EK<IGFC8K<; KF 8E 8EEL8C =I<HL<E:P
1< 8CJF ;<RE< ;LDDP M8I@89C<J KF :8GKLI< K?< / <:FEFD@: >FM<IE8E:< GIF:<JJ<J
K?8K N< 8I< @EK<I<JK<; @E J =8I 8J K?< -K89@C@KP 8E; !IFNK? *8:K @J :FE:<IE<; FLI
98J<C@E< D<8JLI< @J 8 ;LDDP M8I@89C< N?@:? K8B<J M8CL<J @= 8 :FLEKIP @J @E K?<
<LIF 8I<8 8E; FK?<IN@J< .?< I<8JFE =FI K?@J :?F@:< @J K?8K K?< -K89@C@KP 8E; !IFNK?
*8:K @J 8E <JJ<EK@8C =<8KLI< F= K?< DFE<K8IP LE@FE 8E; K?8K 8 JLGI8
E8K@FE8C RJ:8C ILC<
@J 8E @DGFIK8EK LE;<IG@EE@E> F= K?< DFE<K8IP LE@FE #K D8P 8CJF 9< 8I>L<; K?8K K?<
RJ:8C C@D@KJ J<K @E K?< '88JKI@:?K .I<8KP 8E; I<@E=FI:<; @E K?< -K89@C@KP 8E; !IFNK?
*8:K @ESL<E:<; :FLEKIP 9<?8M@FLI <M<E 9<=FI< AF@E@E> K?< <LIF @E K?< :FEM<I><E:<
GIF:<JJ ;LI@E> K?< ILE
LG KF K?< J@E>C< :LII<E:P 1< K?<I<=FI< 8CJF :FEJ@;<I 8 M8I@8EK
F= K?< ;LDDP 7 N?@:? K8B<J M8CL< 8CJF @E K?< KNF P<8IJ GI<:<;@E>
K?< 8;FGK@FE F= K?< <LIF 'FI<FM<I N< 8CJF :FDGLK< 8 ;LDDP
K8B@E> M8CL< =IFD FEN8I;J =FI 8CC / :FLEKI@<J FI =IFD K?< DFD<EK K?<P
AF@E K?< / @E8CCP N< 8CJF :FEJ@;<I K?< GFJJ@9@C@KP K?8K K?< E8KLI< F= K?< *8:K ?8J
:?8E><; 8=K<I 8 P<8I N?@:? 8::FI;@E> KF DFJK F9J<IM<IJ D8IB<; 8 J@>E@R:8EK
N<8B<E@E> F= K?< <Z<:K@M< ;@J:@GC@E< K?8K K?< *8:K <E=FI:<J 1< K?<I<=FI< 8CJF ;<RE<;
7 8J 8 ;LDDP M8I@89C< ;<RE<; 8J K?< 98J<C@E< -!* 9LK FECP LEK@C 8E; Q<IF 8=K<IN8I;J
J =8I 8J K?< &- @J :FE:<IE<; K?< ;<RE@K@FE @J DFI< JKI8@>?K=FIN8I; ?<I< J@E:<
@K 8GGC@<J KF 8CC / :FLEKI@<J 8=K<I 1< :FDGLK< 8 ;LDDP M8I@89C< K8B@E> M8CL<J =FI 8CC / :FLEKI@<J <@K?<I =IFD FI =IFD K?< P<8I N?<E K?<P
AF@E K?< / 1< 8CJF :FEJ@;<I K?< GFJJ@9@C@KP K?8K K?< &- @J 8 :LDLC8K@M< GIF:<JJ
K?< <Z<:KJ F= N?@:? DLCK@GCP 8J K@D< <C8GJ<J =FCCFN@E> @KJ RIJK @DGC<D<EK8K@FE .?<
M8I@89C< 7 K?<I<=FI< :FDGLK<J =FI <8:? P<8I K?< ELD9<I F= P<8IJ
@E N?@:? 8 :<IK8@E :FLEKIP ?8J 9<<E JL9A<:K KF K?< &- K?@J ELD9<I @J F9M@FLJCP Q<IF
=FI EFE
/ :FLEKI@<J
#$.' I<GFIKJ K?< JLDD8IP JK8K@JK@:J =FI 8CC M8I@89C<J LJ<; @E K?@J G8G<I
#EJ<IK .89C<J ?<I<
10
ECB
Working Paper Series No 1344
May 2011
4
A …rst look at the evidence
We …rst look at basic summary statistics for the key variables that are relevant depending on whether countries are, or are not, subject to the SGP and to the LS.
The data are reported in Table 4. Of course, this only represents preliminary, unconditional evidence as it treats all countries in the same way, irrespective of possible
determinants, the only di¤erence being whether they are subject to the modalities of
EU and euro area economic governance or not.
For …scal variables, the data look clearly better for countries which (and when
they) have been subject to the SGP. The average primary balance is +0.3% within
the SGP, and -1.2% without it. Countries under the SGP have also experienced
less variation, with a standard deviation of 3.3% (including both cross section and
time series variation) against 4.4%. Results are similar for the other de…nitions of
the EU …scal rules that we propose (M AAST RICHT , SGP _P RE, SGP _P RE03).
Prima facie, therefore, one is tempted to conclude that the SGP has been a success
in increasing the average level as well as in reducing the standard deviation of the
primary balance. We will see, however, that this conclusion does not survive in the
conditional analysis, where we include other possible determinants of the primary
balance.
(insert Table 4 here)
Concerning the LS, the unconditional results are less promising. Average per
capita income growth, perhaps the best single yardstick of the LS (more discussion on
this later), has been 1.2% on average in the countries subject to the Lisbon Strategy,
against 2.4% in other countries (including EU countries before 2000), and with a
higher standard deviation.
5
5.1
The Stability and Growth Pact and …scal behaviour
Empirical model
We address the problem of evaluating the impact of the EU …scal institutions (the
SGP and Maastricht) in three sequential steps. First, we estimate a model of ex post
…scal behaviour over the full set of countries (both EU and non-EU), similar to, for
example, Gali and Perotti (2003). The model is speci…ed on the primary balance,
that is, not the cyclically adjusted budget balance. The choice of this left hand side
variable - rather than the more common cyclically adjusted one - is motivated by our
desire to study how the …scal institutions have shaped not only the average behaviour
of governments but also their responsiveness to the business cycle, that is, the degree
of pro-cyclicality of …scal policy.
ECB
Working Paper Series No 1344
May 2011
11
.?@J ><E<I8C DF;<C D8P 9< JG<:@R<; 8J 8 ;@Z<I<E:<
@E
;@Z<I<E:< G8E<C ;8K8 DF;<C
" " ! N?<I< @J K?< :FLEKIP K?< P<8I " K?< GI@D8IP 98C8E:< 8E; K?< ! M<:KFI
@E:CL;@E> 8 ELD9<I F= GFJJ@9C< ;<K<ID@E8EKJ F= RJ:8C 9<?8M@FLI 9FK? F= 8E <:FEFD@:
8E; 8 GFC@K@:8C@EJK@KLK@FE8C E8KLI< (FK< K?8K @E K?< ! M<:KFI N< @E:CL;< K?< FLKGLK
>8G N?@:? D8P >@M< I@J< KF 8 I<M<IJ< :8LJ8C@KP GIF9C<D =FI K?@J I<8JFE K?< <HL8K@FE
@J <JK@D8K<; 9P @EJKILD<EK8C M8I@89C<J LJ@E> !'' .?< DF;<C 8CJF @E:CL;<J :FLEKIP
RO<; <Z<:KJ 8E; K@D< ;LDD@<J 8E; 8 :FII<:K@FE =FI K?< JD8CC J8DGC< @J 8GGC@<;
<M<IPN?<I<
#E 8 J<:FE; JK<G N< N8EK KF J<< N?<K?<I K?< -!* ?<I< 8 J?FIK
:LK =FI K?< J<K F=
GFJJ@9C< / RJ:8C ILC<J K?8K N< :FEJ@;<I D8KK<IJ =FI K?< #7'3#)' 9<?8M@FLI F= RJ:8C
GFC@:@<J
" " ! #= K?< -!* ;LDDP @J J@>E@R:8EK K?<E K?< 9<?8M@FLI F= :FLEKI@<J JL9A<:K KF K?<
JLGI8
E8K@FE8C ILC< @J JPJK<D8K@:8CCP ;@Z<I<EK =IFD K?8K F= J@D@C8I :FLEKI@<J K?8K 8I<
EFK JL9A<:K KF K?< ILC< #E GIF:<<;@E> @E K?@J N8P N< 8I< :FEJ@JK<EK N@K? K?< C@K<I8KLI<
FE K?< <:FEFD<KI@:J F= GIF>I8D <M8CL8K@FE J<< #D9<EJ 8E; 1FFC;I@;>< =FI 8
JLIM<P
.?< K?@I; 8E; RE8C JK<G @J KF 8E8CPJ< N?<K?<I K?< ILC< 8Z<:KJ K?< <C8JK@:@KP F=
RJ:8C GFC@:@<J KF :?8E><J @E K?< <:FEFD@: 8E; @EJK@KLK@FE8C <EM@IFED<EK 1< K?<I<=FI<
<OK<E; K?< DF;<C @E KF @E:CL;< @EK<I8:K@FE K<IDJ N@K? 8 JL9J<K " F= K?< ! M<:KFI
" " ! " )E< B<P HL<JK@FE @J N?<K?<I K?< -!* ?8J D8;< RJ:8C GFC@:@<J DFI< FI C<JJ GIF
:P:C@:8C .?8K HL<JK@FE D8P 9< 8;;I<JJ<; 9P CFFB@E> 8K K?< @EK<I8:K@FE K<ID 9<KN<<E
K?< FLKGLK >8G 8E; K?< -!* ;LDD@<J 8E; @J K8:BC<; 8J JL:? C8K<I (FK< K?8K J@E:<
@J 8 9@E8IP ;LDDP K8B@E> M8CL<J Q<IF 8E; K?< G8I8D<K<I :8E 9< @EK<IGI<K<;
8J 8 D8I>@E8C <Z<:K K?8K @J K?< D8I>@E8C 9<E<RK F= DFM@E> =IFD 8 J@KL8K@FE F= EF
KI<8KD<EK KF KI<8KD<EK 1< J?FLC; <DG?8J@J< 8K K?@J GF@EK K?8K K?@J 8E8CPJ@J @J <O GFJK 8E; ;F<J EFK
K?<I<=FI< 8@D 8K <JK@D8K@E> 8E <O 8EK< RJ:8C GFC@:P ILC< @E K?< J8D< N8P 8J G8G<IJ
98J<; FE I<8C K@D< M8I@89C<J ;F <<KJD8 <K 8C !FC@E<CC@ 8E; 'FD@>C@8EF
@D8;FDF EFK?<I :FE:<IE K?8K FE< D8P ?8M< FE FLI 8GGIF8:? @J K?8K
K?< ;<:@J@FE KF AF@E K?< <LIF 8E; K?< / FI <LIF 8I<8 RJ:8C @EJK@KLK@FEJ D8P 9< 8E
<E;F><EFLJ FE< N?@:? D8P >@M< I@J< KF I<M<IJ< :8LJ8C@KP J@E:< K?< KI<8KD<EK @J EFK
)E< C8> F= K?< FLKGLK >8G @J LJ<; 8J K?< @EJKILD<EK N?@:? @DGC@<J K?8K K?< <HL8K@FE @J <O8:KCP
@;<EK@R<; .?<I< @J EF J@>E F= N<8B @EJKILD<EKJ @E K?@J <JK@D8K@FE (FK< K?8K @E FLI :8J< @J
8GGIFO@D8K<CP K?< J8D< J@Q< 8J N?@:? JL>><JKJ K?8K K?< (@:B<CC 9@8J J?FLC; 9< @= 8EPK?@E>
JD8CC J<< $L;JFE 8E; )N<E 12
ECB
Working Paper Series No 1344
May 2011
given exogenously. There is indeed some evidence that countries joining the euro have
had (in the year preceding entry) a higher public debt to GDP ratio than countries
which have not adopted the euro. For example, the 11 countries which formed the
euro area in 1999 had a combined debt to GDP ratio of 64.3%, against 47.1% in
the remaining countries (a situation which has reversed since then). Since having a
higher debt may increase incentives towards …scal consolidation especially beyond a
certain level, our empirical approach may entail a small bias towards …nding a positive
e¤ect for the …scal rule variables. Note that this is only relevant for the conditions
prevailing at the time of euro area (or EU) entry. The e¤ect of higher or lower debt
over the whole sample is already captured in our model by the country …xed e¤ect
(see Imbens and Wooldridge 2009, in particular page 70); we also include the lagged
debt to GDP ratio in the xit vector.3
5.2
Results
The results of the estimation of equation (1) are reported in Table 5. A high debt
to GDP ratio in the previous year, higher trade openness and economic size all contribute to a better primary balance. By contrast, income per capita and the size of
government are statistically insigni…cant. The result for economic size is interesting in
view of the consideration that larger countries typically have larger …scal multipliers
(Buti and Pench 2004); this may imply that they have less need for expansionary …scal policies. The output gap is positive and signi…cant, at around 0.3, indicating that
average …scal behaviour is counter-cyclical.4 We also test (second column) whether
there is any indication of asymmetry between a positive and a negative output gap,
and we …nd that the source of counter-cyclicality only comes from times in which the
output gap is positive (good times). Due to the relatively large size of the standard
errors, however, we are not able to conclude that the di¤erence in the coe¢ cients
associated to positive and negative output gaps is statistically signi…cant.5
In the third column, we add political variables taken from the Database of Political
Institutions. We consider several variables capturing (i) the political cycle, (ii) the
strength and cohesion of the government, (iii) political stability and (iv) a measure of
…scal centralization. We …nd that years in which legislative elections take place are
strongly associated to a worse primary budget balance, by about 0.5% and statistically
highly signi…cant. This con…rms the existing widespread evidence in the literature
of a political business cycle in our sample of countries. We also …nd that the vote
share of the government parties is associated to better budgetary outcomes, indicating
that stronger governments are better able to keep the …scal house in order. Finally, a
3
To deal with the problem of EMU entry endogeneity, Alesina, Ardagna and Galasso (2008) use
instead an instrumental variable approach, where the instruments are the estimated probabilities of
joining a monetary union.
4
This result is very robust to changes in the instruments list.
5
The Wald test is not reported for brevity.
ECB
Working Paper Series No 1344
May 2011
13
M8I@89C< @;<EK@=P@E> :FLEKI@<J N?<I< C<>@JC8KFIJ 8I< <C<:K<; LJ@E> 8 N@EE<I
K8B<
8CC ILC<
GCLI8C@KP @J 8CJF J@>E@R:8EK 8E; GFJ@K@M< GIF989CP 8>8@E @E;@:8K@E> K?8K :FLEKI@<J
N@K? JKIFE><I >FM<IED<EKJ 8I< 9<KK<I 89C< KF D8E8>< GL9C@: RE8E:<J
#E K?< E<OK KNF :FCLDEJ N< 8E8CPJ< 8 J?FIK<I J8DGC< 8E; K8B< FLK
8E; =IFD K?< 98J<C@E< J8DGC< @< K?< JG<:@8C :FE;@K@FEJ F= K?< >CF98C
RE8E:@8C :I@J@J )M<I8CC I<JLCKJ 8I< J@D@C8I KF K?< 98J<C@E< <O<I:@J<
.?< J@OK? :FCLDE I<GFIKJ I<JLCKJ =FI K?< :P:C@:8CCP 8;ALJK<; 9L;><K 98C8E:< (FK
JLIGI@J@E>CP K?< FLKGLK >8G @J EFN @EJ@>E@R:8EK 'FI<FM<I KNF F= K?< K?I<< GFC@K@:8C
M8I@89C<J EFN 8CJF 9<:FD< @EJ@>E@R:8EK
#EJ<IK .89C< ?<I<
.LIE@E> KF <HL8K@FE .89C< I<GFIKJ I<JLCKJ =FI K?< ;@Z<I<EK M<IJ@FEJ F= K?<
-K89@C@KP 8E; !IFNK? *8:K ;LDDP J :8E 9< J<<E K?<P 8I< @EM8I@89CP @EJ@>E@R:8EK
@DGCP@E> K?8K K?< / <LIF 8I<8 RJ:8C ILC< @J @II<C<M8EK KF <OGC8@E K?< 8M<I8>< 9<
?8M@FLI F= K?< GI@D8IP 98C8E:< FE:< K?< :FEKIFC M8I@89C<J 8I< @E:CL;<; 1< ;@J:LJJ
K?< GFJJ@9C< @EK<IGI<K8K@FEJ F= K?@J I<JLCK C8K<I K K?@J JK8>< N< FECP I<:8CC K?8K FLI
GIF:<;LI< ?8J @= 8EPK?@E> 8 JD8CC 9@8J KFN8I;J RE;@E> 8 J@>E@R:8EK <Z<:K N?@:?
@DGC@<J K?8K K?< EF <Z<:K I<JLCK @J IF9LJK 8E; GFJJ@9CP <M<E :FEJ<IM8K@M< 1< 8CJF
K<JK N?<K?<I K?< -!* D8KK<IJ 8K C<8JK =FI :FLEKI@<J N?@:? ?8M< 8 ;<R:@K KF !*
I8K@F 89FM< )E< NFLC; <OG<:K K?8K 8K C<8JK :FLEKI@<J @E K?@J GFJ@K@FE J?FLC; 9<
GLJ?<; ?8I;<I KFN8I;J RJ:8C :FEJFC@;8K@FE LE;<I K?< -!* K?8E FK?<IN@J< 1?8K N<
RE; @J @EJK<8; K?< FGGFJ@K< K?< -!* 8GG<8IJ KF ?8M< 3'&6%'& :FLEKI@<JU @E:<EK@M<J
KF GLIJL< 8 :FII<:K@FE @E K?< GI@D8IP 98C8E:< 8=K<I 9<@E> @E 8E <O:<JJ@M< ;<R:@K GIF
:<;LI< GIF989CP FE 8::FLEK F= K?< 9FELJ I<GI<J<EK<; 9P K?< CFN C<M<C F= @EK<I<JK
I8K<J LE;<I '/
#EJ<IK .89C< ?<I<
LIK?<IDFI< N< I<GFIK I<JLCKJ =FI K?< <JK@D8K@FE F= <HL8K@FE @E #$.' -K8IK@E> =IFD FLI 98J<C@E< D<8JLI< F= K?< -!* RIJK :FCLDE @K :8E 9< J<<E K?8K K?<
RJ:8C ILC< 8GG<8I) KF ?8M< D8;< :FLEKI@<J DFI< :FLEK<I
:P:C@:8C 9LK 8CJF DFI< JL9A<:K
KF K?< GFC@K@:8C :P:C< K?8E FK?<IN@J< .?@J I<JLCK @J ?FN<M<I EFK M<IP IF9LJK KF K?<
;<RE@K@FE F= K?< -!* ;LDDP 8E; @K :FLC; K?<I<=FI< 9< 8E FLK:FD< K?8K @J 8JJF:@8K<;
DFI< KF K?< <LIF 8I<8 G<I J< I8K?<I K?8E KF @KJ RJ:8C =I8D<NFIB K?FL>? @K @J M<IP
;@Y:LCK KF ;@JK@E>L@J? K?< KNF @EK<IGI<K8K@FEJ 98J<; FE FLI ;8K8 )LI I<JLCK JK8E;J @E
:FEKI8JK N@K? !8C@ 8E; *<IFKK@ N?F =FLE; K?8K GIF
:P:C@:8C@KP N8J DFI< DLK<;
8=K<I K?< 8;FGK@FE F= K?< '88JKI@:?K .I<8KP .?<@I I<JLCKJ D8P ?8M< 9<<E 8Z<:K<;
?FN<M<I 9P K?< '88JKI@:?K :I@K<I@8 K?<DJ<CM<J 8E; K?< ILE
LG KF K?< @EKIF;L:K@FE
1< ;F EFK @E:CL;< K?< ,LC< F= &8N @E;@:8KFI 9<:8LJ< @K NFLC; I<;L:< K?< J8DGC< J@Q< KFF DL:?
8E; 9<:8LJ< @K @J GIF989CP EFK M<IP I<C<M8EK KF <OGC8@E RJ:8C 9<?8M@FLI @E I@:? :FLEKI@<J
LK@ <K 8C RE; J@D@C8I <8ICP <M@;<E:< =FI K?< @DG8:K F= K?< GFC@K@:8C 9LJ@E<JJ :P:C< LE;<I
'/
14
ECB
Working Paper Series No 1344
May 2011
of the euro in the 1990s. From the interaction term between economic size and the
SGP, we also …nd no evidence that smaller economies have systematically bene…ted
more from the SGP contrary to what is suggested, for example, by von Hagen (2003)
and Annett (2006).
Finally, we want to establish whether the Stability and Growth Pact has a¤ected
the role of market discipline which, as noted by many observers, seems to have operated in quite an ine¢ cient way in EMU, with government bond spreads being …rst
very small and not reactive to …scal conditions, and then (from 2008 onwards) very
large and exceptionally responsive. What we want to test here is not whether government bond spreads react to …scal imbalances, as commonly acknowledged in the
literature, but rather the other way round, namely, if for given spreads the primary
balance reacts to market signals. In the last column of Table 7, therefore, we include
an interaction term between the SGP and the previous year’s long-term government
bond spread versus the United States Treasury bond yield, taken to represent a global
safe-asset benchmark. There is surprisingly little literature on the disciplining role
of the bond market on government behaviour. Lane (1993) sets out some general
conditions for market discipline to be e¤ective. De Haan and Sturm (2000) study
government bond spreads in Europe and come to the conclusion that market discipline is not very e¤ective.8 Our results indicate that market discipline - as measured
by the in‡uence of government bond spreads in the previous year on the current year’s
primary balance - does not seem to matter much in the determination of primary balances more generally, and this has been so also under the SGP. We also add squared
terms of these variables to capture possible non-linearities but these are again statistically insigni…cant. Therefore, we take this as an indication that if the SGP has
had any e¤ect, it is not through its in‡uence on the working of market discipline on
government behaviour (although it may still have in‡uenced market discipline meant
as the responsiveness of yield spreads to …scal imbalances).
(Insert Table 7 here)
6
The Lisbon Strategy and economic performance
Moving beyond the EU/euro area …scal rules, we consider the possible impact of the
Lisbon Strategy (LS) on structural economic performance. We look at outcomes and
do not consider reform e¤orts. This distinguishes our paper from other papers such
as Duval and Elmeskov (2006) and Alesina, Ardagna and Galasso (2008), which have
tried inter alia to establish a link between the intensity of structural reforms and the
introduction of the euro. An important characteristic of the LS was precisely that it
did not focus on a particular set of structural reforms to be implemented, leaving individual EU countries much freedom under the central coordinating procedures of the
8
Bulut (2009) also …nds little evidence of market discipline for sovereign borrowers of developing
countries.
ECB
Working Paper Series No 1344
May 2011
15
IF8; :FEFD@: *FC@:P !L@;<C@E<J 8E; K?< DGCFPD<EK !L@;<C@E<J K?< V#EK<>I8K<;
!L@;<C@E<JW GFJK
KF GLIJL< K?<@I FNE N8PJ .?< &- ;@; ?FN<M<I :FEK8@E 8
ELD9<I F= HL8EK@K8K@M< F9A<:K@M<J @E K<IDJ F= FLK:FD<J #F8EEFL <K 8C IFD
8 D<K?F;FCF>@:8C G<IJG<:K@M< @E KIP@E> KF 8J:<IK8@E K?< D8I>@E8C :FEKI@9LK@FE F= K?<
&- N< E<<; KF :FEKIFC =FI 8 J<I@<J F= ;<K<ID@E8EKJ K?8K ?8M< 9<<E @;<EK@R<; @E K?<
C@K<I8KLI< 8J GFK<EK@8C ;<K<ID@E8EKJ F= I<=FID <ZFIK 8E; <:FEFD@: G<I=FID8E:< FI
K?@J N< ?8M< CFFB<; 8K K?< C@K<I8KLI< FE K?< GFC@K@:8C <:FEFDP F= JKIL:KLI8C I<=FID
.?@J C@K<I8KLI< ?8J <DG?8J@J<; @E G8IK@:LC8I K?< ;<>I<< F= FG<EE<JJ F= K?< <:FE
FDP N?@:? D8B<J @K <8J@<I KF :FEM@E:< JK8B<?FC;<IJ >@M<E K?< ?@>?<I ;<>I<< F= <OK<I
E8C :FDG<K@K@FE #' K?< J@Q< F= K?< <:FEFDP N@K? JD8CC<I :FLEKI@<J 9<@E>
DFI< FG<E C<J@E8 8E; 18:Q@8I> 8E; DFI< :LCKLI8CCP ?FDF><EFLJ K?<I<9P 8C
CFN@E> >I<8K<I <Z<:K@M<E<JJ F= ;<:@J@FE D8B@E> K?< ;<>I<< F= :<EKI8C@J8K@FE;<MFCLK@FE
F= ;<:@J@FE D8B@E> GIF:<JJ<J N@K? K?< =FID<I @DGCP@E> C<JJ I<J@JK8E:< .FDGJFE
K?< =FID F= <C<:KFI8C ILC<J K?8K @J D8AFI@K8I@8E M<IJLJ GIFGFIK@FE8C N@K? K?<
C8KK<I @DGCP@E> C<JJ GFN<I J?8I@E> 8E; DFI< :FEJ<EJLJ F<I@ <K 8C 9LK GFJ
J@9CP 8CJF @DGCP@E> DFI< JLJK8@E89C< JFCLK@FEJ N?@:? I<S<:K K?< @EK<I<JKJ F= 9IF8;
D8AFI@K@<J #' K?< E8KLI< F= K?< GFC@K@:8C JPJK<D N@K? G8IC@8D<EK8IP 98J<;
JPJK<DJ GFJJ@9CP 9<@E> DFI< S<O@9C< K?8E GI<J@;<EK@8C
J =FI K?< GFJJ@9C< I<8JFEJ 9<?@E; K?< J?FIK:FD@E>J F= K?< &- @E G8IK@:LC8I K?<
C8K<JK C@K<I8KLI< ?@>?C@>?KJ J<M<I8C F= K?< =8:KFIJ 8CI<8;P @;<EK@R<; 9P K?< %FB ,<GFIK
8E; K?< LIFG<8E FDD@JJ@FE 8CI<8;P 98:B @E K?< D@;
J N?<E
K?< &@J9FE -KI8K<>P LE;<IN<EK 8 I<=FID @E G8I8CC<C KF K?< -!* I<=FID #E G8IK@:LC8I
8J 8CJF <OGC8@E<; 9P #F8EEFL <K 8C K?< I<=FID F= K?< &- ;@; EFK 8:?@<M< 8
J?8IG<I =F:LJ F= K?< 8@DJ F= K?< -KI8K<>P EFI F= K?< <E?8E:<D<EK F= K?< >FM<IE8E:<
KFFCJ LJ<; KF @DGC<D<EK @K <> 9<E:?D8IB@E> 8>8@EJK HL8EK@K8K@M< F9A<:K@M<J
-<8I:?@E> =FI 8 ;<<G<I :8LJ< ,LK8 FCC@>EFE 8E; -:?FLK 8E;
$FI;8E 8CC K<E; KF 8KKI@9LK< K?< =8@CLI< F= K?< &@J9FE -KI8K<>P KF E8K@FE8C
GFC@K@:8C :FEJKI8@EKJ ,LK8 @E G8IK@:LC8I <DG?8J@J<J K?8K 9<:8LJ< E8K@FE8C
>FM<IED<EKJ @E K?< / ?8M< I<K8@E<; GI8:K@:8CCP JFC< :FDG<K<E:< @E <:FEFD@: GFC@:P
JG@CCFM<I <Z<:KJ F= I<=FID 8I< EFK =LCCP @EK<IE8C@Q<; N?@:? 8CCFNJ M<JK<; @EK<I<JKJ KF
CF99P JL::<JJ=LCCP @E E8K@FE8C :8G@K8CJ *8IK@:LC8ICP I<C<M8EK @E K?@J I<>8I; ?8M< 9<<E
K?< RE8E:@8C 8E; DFI< 9IF8;CP J<IM@:<J J<:KFI N?@:? ?8M< 9<<E K8I><KJ =FI C@9<I
8C@J8K@FE LE;<I K?< &@J9FE -KI8K<>P FCC@>EFE RE;J K?< /UJ ;@J8GGF@EK@E>
G<I=FID8E:< KF 9< K?< I<JLCK F= 8 :FCC<:K@M< 8:K@FE GIF9C<D N?@:? <D<I><J VN?<E 8L
KFEFDFLJ >FM<IED<EKJ J<<B KF D8O@D@J< :FCC<:K@M< LK@C@K@<J @E @JFC8K<; :FEJK@KL<E:@<JW
G -:?FLK 8E; $FI;8E 8I>L< K?8K DL:? F= K?< &@J9FE -KI8K<>P I<C@<;
FE DF;<J F= VE<KNFIB<; >FM<IE8E:<W N?<I<9P :<EKI8C 9F;@<J ;<G<E;<; LGFE K?< :F
FG<I8K@FE 8E; AF@EK I<JFLI:< DF9@C@Q8K@FE F= GFC@:P 8:KFIJ FLKJ@;< K?<@I ?@<I8I:?@:8C
:FEKIFC .?< 8LK?FIJ HL<JK@FE N?<K?<I E8K@FE8C 8E; JL9
E8K@FE8C 8;D@E@JKI8K@FEJ
?8M< LG>I8;<; K?<@I :F
FI;@E8K@E> :8G8:@K@<J KF D8B< E<KNFIB
98J<; DF;<J =LE:K@FE
<Z<:K@M<CP %F:QFI @;<EK@R<J F9A<:K@M< 8E; JL9A<:K@M< =8:KFIJ @ESL<E:@E> JK8K<JU
@DGC<D<EK8K@FE F= K?< &@J9FE -KI8K<>P /E;<I K?< =FID<I ?<8;@E> ?< @E:CL;<J K?<
16
ECB
Working Paper Series No 1344
May 2011
general level of development of a country and society’s adaptability to the globalisation process. Under subjective factors, Koczor notes the importance of e¢ ciency
of governmental action, political will and the consensus for reform, the extent to
which non-governmental entities (employer and employee organisations) work with
the government to draft and implement the strategy, as well as the social acceptance
of reform.
Wyplosz (2010), in line with the call by Ioannou et al. (2008) for a more explicit
benchmarking, argues that the shift from pointed criticism to diplomatic peer pressure from the Commission undermined the process. Further, he notes that “political
leaders are not raised to encourage critical comments from each other. More importantly, perhaps, while even polished exercises of apparent mutual admiration could
still exercise some pressure, political leaders never forget that they are accountable to
domestic voters.”
Finally, Padoan (2009) sees an incomplete policy mix and a delay in capitalising on
a changing international environment as the main reasons for failure. The persistent
European de…cit in R&D is, in his opinion, the result of a failure to promote the
emergence and growth of innovative businesses in new sectors.
6.1
Empirical model
For the empirical model we follow a similar approach as in Section 4. Let zit be an
indicator of economic performance that is relevant for the LS. We …rst estimate a
model
zit = zi;t 1 + xit + i + t + "it
(4)
where the performance indicator is regressed on a vector of possible fundamental
determinants x (possibly also timed t 1 where reverse causality is a potential concern). Once we obtain a satisfactory parsimonious model for equation (4), we add
the dummy variables capturing the LS:
zit = zi;t
1
+ xit +
i
+
t
+ LISBONit + "it
(5)
The coe¢ cient captures the additional e¤ect, coming on top of all other control
variables, stemming from the fact that a given country is subject to a supra-national
process, the LS. To simplify things, …nding
> 0 would imply that the LS has
"worked" and that its success is visible in the data. Also in this case, as for the
analysis of ex post …scal behaviour, we emphasise the risk of selection bias, as it
could well be that countries with structural weaknesses are precisely those which
undertake a stronger reform e¤ort, in the same way as patients who are more ill are
more likely to take a certain medicine. In the case of the LS, this may be less of a
concern since it applies indistinctly to the whole EU and the decision to join the EU
(unlike, at least in part, the decision to join the euro area) largely re‡ects geographical
ECB
Working Paper Series No 1344
May 2011
17
and political determinants, not economic policy objectives.9 Furthermore, this is if
anything a source of bias in the direction of …nding > 0; rather than the other way
round.
It should be recalled that the LS was an overarching strategy entailing not only an
economic but also a social and environmental dimension. This was indeed identi…ed
as one of its weaknesses and may explain the poor performance in economic terms. We
focus on the economic dimension and take three indicators as best overall measures of
economic performance: (i) per capita income growth, (ii) labour productivity growth,
and (iii) employment growth.
6.2
Results
Table 8 reports the results for per capita income growth. In estimating equation (4)
and retaining the signi…cant variables, we …nd that a few variables are robustly associated to per capita income growth. First, initial conditions matter: the lagged per
capita income level has a negative sign, suggesting some catching up process. Second,
trade openness is associated to stronger per capita income growth, as is (surprisingly
from one angle but see below) a higher level of Employment Protection Legislation.
Turning to the political and institutional variables, we …nd that Political Stability
(note that a higher reading of this indicator implies less stability) and Proportional
Representation are associated to higher per capita growth. The results indicate that
countries with lower income per capita, higher trade openness, more employment
protection, more political stability and with a proportional political system tend to
experience higher per capita income growth. We also try a number of additional variables that turn out to be insigni…cant. We try the World Bank Rule of Law indicator,
as in Rodrik et al. (2004), but we …nd this variable insigni…cant, though correctly
signed. The insigni…cance is likely to be linked to two main di¤erences between the
analysis in this paper and Rodrik et al. (2004). First, we look at per capita income
growth rather than levels. Second, our panel includes mostly rich countries, while the
quality of institutions (such as the protection of property rights) are likely to explain
the di¤erence between poor and rich countries rather than the smaller di¤erences
among rich countries. Furthermore, economic size is also insigni…cant when included
together with openness (though it is signi…cant when included alone). Finally, we
also try several variables capturing the country’s political institutions, …nding all of
them insigni…cant (apart from those reported in Table 7).10
(Table 8 here)
#E 8;;@K@FE K?< GF@EK D8;< <8IC@<I =FI K?< :FLEKIP RO<; <Z<:KJ 8CJF 8GGC@<J KF K?@J G8IK F= K?<
8E8CPJ@J
.?< )UJ *IF;L:K '8IB<K ,<>LC8K@FE @J 8CJF @EJ@>E@R:8EK GIF989CP FE 8::FLEK F= K?< DFI<
C@D@K<; ;8K8 8M8@C89@C@KP
18
ECB
Working Paper Series No 1344
May 2011
.?< J<:FE; :FCLDE F= .89C< I<GFIKJ K?< J8D< <HL8K@FE N?<E 8;;@E> K?< &@J
9FE ;LDDP N?@:? @J @EJ@>E@R:8EK FCLDE K?<E I<GFIKJ K?< &#-)(73,;LDDP N?@:? :8K<IJ =FI K?< GFJJ@9@C@KP K?8K K?< &- ?8J 8 :LDLC8K@M< @EM<JKD<EK
C@B< E8KLI< CJF K?@J ;LDDP M8I@89C< @J @EJ@>E@R:8EK .?< =FLIK? KF J@OK? :FCLDEJ
GI<J<EK JFD< IF9LJKE<JJ 8E8CPJ@J @E G8IK@:LC8I @ I<JKI@:K@E> K?< J8DGC< G<I@F; KF
@@ K8B@E> FLK 8E; K?< <O:<GK@FE8C P<8IJ F= K?< >CF98C RE8E:@8C
:I@J@J 8E; @@@ <O:CL;@E> F9J<IM8K@FEJ =FI M<IP CFN @E:FD< G<I :8G@K8 .?< I<JLCKJ 8I<
GI8:K@:8CCP LE:?8E><; N@K? K?< FECP <O:<GK@FE F= *IFGFIK@FE8C ,<GI<J<EK8K@FE K?8K
@J EFK J@>E@R:8EK 8EPDFI< @E K?< J8DGC< G<I@F; 1< :FE:CL;< K?<I<=FI<
K?8K K?<I< @J EF <M@;<E:< F= 8E @DG8:K F= K?< &- FE G<I :8G@K8 @E:FD< >IFNK? =FI K?<
/ :FLEKI@<J
.89C< I<G<8KJ K?< J8D< <O<I:@J< =FI C89FLI GIF;L:K@M@KP >IFNK? 8EFK?<I D8@E
JK8P F= K?< &- >8@E N< RE; K?8K C8>><; G<I :8G@K8 @E:FD< C<M<CJ K<E; KF ?8M< 8
E<>8K@M< @DG8:K FE GIF;L:K@M@KP >IFNK? N?@C< 8>8@E KI8;< FG<EE<JJ ?8J 8 GFJ@K@M<
@DG8:K K?FL>? C<JJ IF9LJKCP JK8K@JK@:8CCP J@>E@R:8EK K?8E @E K?< :8J< F= G<I :8G@K8
@E:FD< >IFNK? 1< 8CJF RE; K?8K 8 ?@>?<I C8>><; N8>< J?8I< F= @E:FD< @J J@>E@R:8EK
N@K? 8 GFJ@K@M< J@>E @E;@:8K@E> K?8K 8 ?@>?<I N8>< J?8I< F= @E:FD< =FJK<IJ GIF;L:K@M
@KP >IFNK? .?@J D8P 9< ;L< KF K?< =8:K K?8K 8 ?@>?<I N8>< J?8I< F= @E:FD< D8B<J
:8G@K8C DFI< J:8I:< 8E; ?<E:< I8@J<J K?< D8I>@E8C <Y:@<E:P F= :8G@K8C :FEKI@9LK@E>
KF ?@>?<I GIF;L:K@M@KP >IFNK? )E:< 8>8@E N< KIP 8 ELD9<I F= 8;;@K@FE8C M8I@89C<J
N?@:? KLIE FLK KF 9< @EJ@>E@R:8EK @E K?< <JK@D8K<; <HL8K@FE .?< &@J9FE ;LDD@<J
8I< ?<I< DFJKCP J@>E@R:8EK 9LK E<>8K@M<CP J@>E<; @E;@:8K@E> K?8K 9<@E> G8IK F= K?< &?8J I<;L:<; I8K?<I K?8E @E:I<8J<; C89FLI GIF;L:K@M@KP >IFNK? )E< @EK<IGI<K8K@FE
F= K?@J I<JLCK EFK<; @E JFD< C@K<I8KLI< @J K?8K K?< &- ?8; KF GLIJL< KNF F9A<:K@M<J
K?8K 8I< ;@Y:LCK KF I<:FE:@C< J@DLCK8E<FLJCP K?8K @J I8@J@E> C89FLI GIF;L:K@M@KP 8E;
<OG8E;@E> <DGCFPD<EK ZFIKJ ;@I<:K<; 8K K?< C8KK<I F9A<:K@M< D8P ?8M< N<8B<E<;
K?< RIJK F9A<:K@M< .?< I<JLCKJ 8I< IF9LJK KF :?8E>@E> K?< J8DGC< G<I@F; 8E; N?<E
<O:CL;@E> K?< >CF98C RE8E:@8C :I@J@J
.89C< ?<I<
@E8CCP #$.' I<GFIKJ I<JLCKJ =FI <DGCFPD<EK >IFNK? #E K?@J :8J< N< RE;
C89FLI =FI:< >IFNK? 8 D8@ECP ;<DF>I8G?@: M8I@89C< JKIFE>CP J@>E@R:8EK 8E; GFJ
@K@M< N?@C< K?< C8>><; <DGCFPD<EK J?8I< @J E<>8K@M< JL>><JK@E> 8E <IIFI :FII<:K@FE
9<?8M@FLI :FLEKI@<J N@K? 8 ?@>?<I <DGCFPD<EK J?8I< <OG<I@<E:@E> CFN<I <DGCFPD<EK
>IFNK? 8E; K?< FK?<I N8P IFLE; .?<I< @J 8CJF JFD< <M@;<E:< K?8K K?< 0FK< -?8I< F=
!FM<IED<EK *8IK@<J 8E @E;@:8KFI F= K?< JKI<E>K? 8E; JK89@C@KP F= >FM<IED<EKJ <O<IKJ
8 GFJ@K@M< @DG8:K FEKI8;@:K@E> K?< <8IC@<I I<JLCK N@K? G<I :8G@K8 @E:FD< >IFNK? 8J
K?< @E;<G<E;<EK M8I@89C< N< RE; K?< ) D<8JLI< F= DGCFPD<EK *IFK<:K@FE &<>
@JC8K@FE KF 9< @EJ@>E@R:8EK GFJJ@9CP 9<:8LJ< @K @J :FII<C8K<; KF K?< <DGCFPD<EK J?8I<
@E C<M<CJ 9LK EFK E<:<JJ8I@CP KF K?< >IFNK? I8K< F= <DGCFPD<EK (FK JLIGI@J@E>CP 8K
K?@J JK8>< N< RE; K?< &- ;LDD@<J @EJ@>E@R:8EK <M<IPN?<I< <O:<GK @E :FCLDE N?<I< K?< 7 ;LDDP @J 0')#5+7'.: J@>E<; )M<I8CC FLI I<JLCKJ :FE
M<I>< KF K?< JKIFE> :FE:CLJ@FE K?8K K?< &- ?8; 8K 9<JK EF @DG8:K FE K?< M8I@89C<J
K?8K @K N8J J<K KF GFJ@K@M<CP :FEKI@9LK< KF
.89C< ?<I<
ECB
Working Paper Series No 1344
May 2011
19
7
Conclusions
.?@J G8G<I @J 8E '9 2145 8E8CPJ@J F= N?<K?<I KNF B<P <C<D<EKJ F= K?< / 8E; <LIF
8I<8 <:FEFD@: >FM<IE8E:< K?< -K89@C@KP 8E; !IFNK? *8:K 8E; K?< &@J9FE -KI8K<>P
8:?@<M<; K?<@I >F8CJ 8 ;<:8;< FI DFI< 8=K<I K?<P N<I< <JK89C@J?<; 1< ?8M< CFFB<; 8K
8 N@;< I8E>< F= 8EEL8C ;8K8 :FD@E> =IFD :FLEKI@<J FM<I K?< J8DGC< G<I@F; =IFD
KF N< ?8M< :FEJ@;<I<; 8 C8I>< J<K F= GFJJ@9C< :FEKIFC M8I@89C<J 8J N<CC 8J
M8I@89C<J N?@:? :FLC; @EK<I8:K N@K? K?< JL::<JJ FI C8:B K?<I<F= F= K?< / <:FEFD@:
>FM<IE8E:< G@CC8IJ )M<I8CC FLI I<JLCKJ @E;@:8K< K?8K JF =8I <:FEFD@: >FM<IE8E:< @E
K?< / 8E; K?< <LIF 8I<8 ?8J ?8; C@D@K<; FI EF JL::<JJ FI K?< &@J9FE -KI8K<>P FLI
I<JLCKJ 8I< EFK C@B<CP KF 9< =FLE; JLIGI@J@E> 1?@C< K?< I<JLCKJ FE K?< <Z<:KJ F= K?<
-K89@C@KP 8E; !IFNK? *8:K D8P 9< =FLE; DFI< JLIGI@J@E> K?<P 8I< :FEJ@JK<EK N@K?
K?< FM<I8CC N<8B<E@E> F= K?< *8:K <JG<:@8CCP J@E:< E FGK@D@JK@: I<8;@E> F= FLI I<JLCKJ @J K?8K RE;@E> EF <Z<:K F= K?< -!* FE RJ:8C
FLK:FD<J @J 8E @E;@:8KFI F= JL::<JJ J@E:< @K @DGC@<J K?8K K?@J J<K F= ILC<J ?8J GI<
M<EK<; K?< <JK89C@J?D<EK F= K?< DFE<K8IP LE@FE @E LIFG< KF 8;M<IJ<CP 8Z<:K RJ:8C
9<?8M@FLI 8 I@JB K?8K N8J <DG?8J@J<; N@;<CP @E K?< ILE
LG KF K?< @EKIF;L:K@FE F=
K?< J@E>C< DFE<K8IP GFC@:P #E K?@J @EK<IGI<K8K@FE FLI RE;@E> F= EF <Z<:K F= K?< -!*
FE K?< 8M<I8>< GI@D8IP 98C8E:< @J C8I><CP 8 I<JLCK F= 8 KNF <Z<:KJ :FLEK<I8:K@E> <8:?
FK?<I K?8K @J K?< GFJ@K@M< <Z<:KJ F= K?< -!* ?8M< :FLEK<I98C8E:<; K?< E<>8K@M< <O
K<IE8C@K@<J F= K?< :FDDFE :LII<E:P FE :FLEKI@<JU @E:<EK@M<J KFN8I;J RJ:8C GIL;<E:<
CK?FL>? K?<I< D@>?K 9< 8E <C<D<EK F= KILK? @E K?@J @EK<IGI<K8K@FE N< ;F EFK RE; @K
<EK@I<CP GC8LJ@9C< #E 8EP :8J< FLI 8GGIF8:? ;F<J EFK 8CCFN LJ KF <OGC@:@KCP :FEJ@;<I
:FLEK<I=8:KL8CJ 1< 8CJF :FEJ@;<I K?8K 8E 8GGIFGI@8K< J<K F= RJ:8C ILC<J @E 8 DFE<K8IP
LE@FE J?FLC; >F 9<PFE; K?< EF
:?8E>< FLK:FD< 8E; @DGFJ< >I<8K<I RJ:8C ;@J:@GC@E<
K?8E FK?<IN@J< FE 8::FLEK F= K?< GFJJ@9C< E<>8K@M< JG@CCFM<IJ K?8K RJ:8C GIFS@>8:P @E
@E;@M@;L8C :FLEKI@<J D8P ?8M< FE FK?<I D<D9<IJ 8CJF K?IFL>? K?< J@E>C< DFE<K8IP
GFC@:P 8J N<CC 8J K?< GFJJ@9C< I<;L:K@FE @E D8IB<K ;@J:@GC@E< =FI @E;@M@;L8C '<D9<I
-K8K<J 9IFL>?K 89FLK 9P K?< :FDDFE :LII<E:P )LI I<JLCKJ @E;@:8K< K?8K K?< -!*
?8J EFK ;<C@M<I<; 8::FI;@E> KF K?@J JKI@:K<I 9<E:?D8IB .?< D8@E GFC@:P @DGC@:8K@FE
JK<DD@E> =IFD FLI 8E8CPJ@J @J K?8K JL9JK8EK@8C GIF>I<JJ FI 8 HL8EKLD C<8G @J E<:
<JJ8IP @E I<=FID@E> / 8E; <LIF 8I<8 <:FEFD@: >FM<IE8E:< >F@E> =FIN8I; @E FI;<I KF
<EJLI< K?< :FEK@EL@E> 9<E<RKJ F= '/ =FI LIFG<8E :@K@Q<EJ .?@J :FE:CLJ@FE :FD<J
8K 8 K@D< N?<E @DGFIK8EK I<=FIDJ KF <:FEFD@: >FM<IE8E:< 8I< 9<@E> :FEK<DGC8K<; @E
K?< / =FCCFN@E> K?< GIFGFJ8CJ F= K?< FDD@JJ@FE FE -<GK<D9<I #E K<IDJ F= HL<JK@FEJ =FI =LKLI< I<J<8I:? N< RE; K?8K K?< @EK<I8:K@FE 9<KN<<E
JLGI8
E8K@FE8C ILC<J JL:? 8J K?< -K89@C@KP 8E; !IFNK? *8:K 8E; K?< &@J9FE -KI8K<>P
8E; E8K@FE8C RJ:8C 8E; <:FEFD@: @EJK@KLK@FEJ @J 8 GIFD@J@E> R<C; F= @EM<JK@>8K@FE
N?@:? N< ?8M< EFK GLIJL<; ?<I< FN@E> KF ;8K8 C@D@K8K@FEJ 9LK N?@:? D<I@KJ =LIK?<I
8KK<EK@FE
FI <O8DGC< -N<;@J? GI@D< D@E@JK<I I<;I@B ,<@E=<C;K NIFK< @E $LE< K?8K 7'0 +( 231)3'44
*#4 $''0 /#&' +5 /645 $' 4#+& 5*#5 5*' +4$10 )'0&# 8+5* 10.: # :'#3 3'/#+0+0) $'(13' +5 +4 51 $'
'7#.6#5'& *#4 $''0 # (#+.63'
20
ECB
Working Paper Series No 1344
May 2011
References
[1] Afonso, A. and S. Hauptmeier (2009): "Fiscal behaviour in the European Union:
rules, …scal decentralization and government indebtedness", ECB Working Paper
No. 1054.
[2] Alesina, A. and G. Tabellini (1990): "Voting on the budget de…cit", American
Economic Review, 80, 1, pp. 37-49.
[3] Alesina, A. and A. Drazen (1991): "Why are …scal stabilisations delayed", American Economic Review, 81, pp. 1170-1180.
[4] Alesina, A. and R. Perotti (1999): "Budget de…cits and budget institutions", in
Poterba J. and J. von Hagen, Fiscal Institutions and Fiscal Performance, NBER,
Chicago: University of Chicago Press.
[5] Alesina, A., and R. Wacziarg (1998): "Openness, country size, and government",
Journal of Pulic Economics, 69, pp. 305-321.
[6] Alesina, A., Ardagna, S. and V. Galasso (2008): "The euro and structural reforms", NBER Working Paper No. 14479.
[7] Annett, A. (2006): "Enforcement and the Stability and Growth Pact: How …scal
policy did and did not change under Europe’s …scal framework", IMF Working
Paper No. 06/116.
[8] Beck, T., Clarke, G., Gro¤, A., Keefer, P. and P. Walsh (2001): "New tools
in comparative political economy: The Database of Political Institutions." 15:1,
165-176 (September), World Bank Economic Review.
[9] Bernoth, K., Hughes Hallet, A. and J. Lewis (2009): "Did …scal policy makers
know what they were doing? Reassessing …scal policy with real time data",
CEPR Discussion Paper No. 6758.
[10] Beetsma, R., Giuliodori, M. and P. Wierts (2009): "Budgeting versus implementing …scal policy in the EU", Economic Policy, forthcoming.
[11] Boeri, T., Castanheira, M., Faini, R. and V. Galasso (2006): Structural reforms
Without Prejudices, Oxford University Press.
[12] Buiter, Willem H. (2005): "The “Sense and Nonsense of Maastricht”Revisited:
What Have We Learnt About Stabilization in EMU?", Discussion Paper Series
–No. 5405, Centre for Economic Policy Research.
[13] Bulut, L. (2009): "Market Disciplining of the Developing Countries’Sovereign
Governments", available at SSRN: http://ssrn.com/abstract=1323440.
ECB
Working Paper Series No 1344
May 2011
21
56 LK@ ' @AYE><I - 8E; I8E:F ,<M@J@K@E> '/UJ -K89@C@KP *8:K
GI8>D8K@: N8P =FIN8I; )O=FI; ,<M@<N F= :FEFD@: *FC@:P 0FCLD< (LD
9<I 56 LK@ ' 8E; * (FFI; @J:8C @J:I<K@FE 8E; C<:K@FEJ @E K?< 8ICP
3<8IJ F= '/ $FLIE8C F= FDDFE '8IB<K -KL;@<J GG 56 LK@ ' 8E; & *<E:? V1?P ;F &8I>< FLEKI@<J CFLK K?< -K89@C@KP
*8:K E; 1?8K 8E 9< FE< 9FLK @KW $FLIE8C F= FDDFE '8IB<K -KL;@<J
GG 56 LK@ ' ,F><I 1 8E; .LII@E@ #J &@J9FE =8I =IFD '88J
KI@:?K .I8;<
FZJ 8E; :FDGC<D<EK8I@K@<J 9<KN<<E RJ:8C ;@J:@GC@E< 8E; JKIL:KLI8C
I<=FIDJ -@=F :FEFD@: -KL;@<J GG 56 @D8;FDF $ "8J K?< -K89@C@KP 8E; !IFNK? *8:K '8;< @J:8C *FC@:P
DFI< *IF
P:C@:8C &8 &<KKI< ;L *## *## I<J<8I:? :<EK<I @JJL< 56 @D8;FDF $ @J:8C GFC@:@<J @E I<8C K@D< 1FIB@E> *8G<I (F
56 FCC@>EFE -K<=8E .?< &@J9FE -KI8K<>P D8:IF<:FEFD@: JK89@C@KP 8E;
K?< ;@C<DD8 F= >FM<IE8E:< N@K? >FM<IED<EKJ FI N?P LIFG< @J EFK 9<:FD@E>
K?< NFIC;UJ DFJK ;PE8D@: <:FEFDP #EK<IE8K@FE8C $FLIE8C F= *L9C@: *FC@:P MFC
EFJ GG 56 ;< "88E $ 8E; $
-KLID F RE8E:@8C D8IB<KJ 8E; K?< '88JKI@:?K
.I<8KP ;@J:@GC@E< >FM<IED<EKJ (<N <M@;<E:< GGC@<; @E8E:@8C :FEFD@:J
GG 56 <9ILE 2 8E; ' - %LD8I .?< ;@J:@GC@E<
<E?8E:@E> IFC< F= RJ:8C
@EJK@KLK@FEJ K?<FIP 8E; <DG@I@:8C <M@;<E:< #' 1FIB@E> *8G<I (F 56 I8Q<E .?< GFC@K@:8C 9LJ@E<JJ :P:C< 8=K<I P<8IJ @E <IE8EB<
8E; % IF>FZ <;J (, '8:IF<:FEFD@:J EEL8C 8D9I@;>< '#. *I<JJ
56 LM8C , 8E; $ CD<JBFM .?< <Z<:KJ F= '/ FE JKIL:KLI8C I<=FIDJ
@E C89FLI 8E; GIF;L:K D8IB<KJ 1FIB@E> *8G<I (F 56 LIFG<8E <EKI8C 8EB 'FEK?CP LCC<K@E I8EB=LIK
*' 3'(13/ 1( 5*' 5#$+.+5: #0& 3185* #%5
56 X *' 631 #5 5'0 ? .'44104 #0& %*#..'0)'4 @=K? :<EKI8C 98EB@E>
:FE=<I<E:< (FM<D9<I
56 LIFG<8E FDD@JJ@FE "13-+0) 51)'5*'3 (13 )3185* #0& ,1$4 ? 0'8
45#35 (13 5*' +4$10 53#5'): ILJJ<CJ
22
ECB
Working Paper Series No 1344
May 2011
56 LIFG<8E FDD@JJ@FE ! 6%%'44'4 #0& %*#..'0)'4 #(5'3 5'0
:'#34 1( %101/+% #0& 10'5#3: !0+10 &LO<D9FLI> )Y:< =FI )Y:@8C *L9C@
:8K@FEJ F= K?< LIFG<8E FDDLE@K@<J
56
8K\J EKFE@F 8E; #C@8E '@?FM )E :FEJKI8@E@E> RJ:8C GFC@:P ;@J:I<K@FE
@E '/ )O=FI; ,<M@<N F= :FEFD@: *FC@:P 0FCLD< (LD9<I 56
@C@G<B % 8E; . -:?I<@9<I .?< -K89@C@KP 8E; !IFNK? *8:K *8JK
*<I=FID8E:< 8E; LKLI< ,<=FIDJ FCC<>< F= 1@CC@8D 8E; '8IP <G8IKD<EK F=
:FEFD@:J X 1FIB@E> *8G<I 56 !8C@ $ 8E; , *<IFKK@ @J:8C GFC@:P 8E; DFE<K8IP @EK<>I8K@FE @E L
IFG< :FEFD@: *FC@:P GG 56 !FC@E<CC@ , 8E; - 'FD@>C@8EF .?< :P:C@:8C I<8:K@FE F= RJ:8C GFC@:@<J
@E K?< <LIF 8I<8 K?< IFC< F= DF;<CC@E> :?F@:<J 8E; ;8K8 M@EK8><J @J:8C -KL;@<J
GG 56 "8CC<I9<I> ' 1/'45+% 6&)'54 +0 # !0+5'& 6312' +4%#. 17'30#0%'
(31/ 5*' '0& 1( 3'5510 "11&4 51 ! FIE<CC /E@M<IJ@KP *I<JJ #K?8:8 8E;
&FE;FE
56 "F;JFE (8K@FE8C @J:8C #EJK@KLK@FEJ 8E; K?< -K89@C@KP 8E; !IFNK?
*8:K I< T<C<>8K@FEU -K8K<J 8K 8 @J8;M8EK8>< @E 1.+%: 04536/'054 (13
160& +4%#. 1.+%+'4 +4%#. 6.'4 #0& 045+565+104 PLJF
@
8J8CJ $ <IFFJ<
- CFI<J 8E; 'FLC@E & <;J *8C>I8M< '8:D@CC8E
56 #D9<EJ ! 1 8E; $ ' 1FFC;I@;>< ,<:<EK ;<M<CFGD<EKJ @E K?<
<:FEFD<KI@:J F= GIF>I8D <M8CL8K@FE $FLIE8C F= :FEFD@: &@K<I8KLI< GG
56 #EK<IE8K@FE8C 'FE<K8IP LE; V FJK<I@E> -KIL:KLI8C ,<=FIDJ @E #E;LJKI@8C
FLEKI@<JW 1FIC; :FEFD@: )LKCFFB
56 #EK<IE8K@FE8C 'FE<K8IP LE; LIF I<8 *FC@:@<J #' FLEKIP ,<GFIK
(F $LE< 56 #EK<IE8K@FE8C 'FE<K8IP LE; LIF I<8 *FC@:@<J #' FLEKIP ,<GFIK
(F $LCP 56 #F8EEFL <I;@E8E;LJJ< ' &F L:8 ' FLJJ<EJ 1 @GFCCFE< *
,FJFC@8 8E; " -K8?C <E:?D8IB@E> K?< &@J9FE -KI8K<>P )::8J@FE8C *8G<I (F 56 $FELE> & 8E; ' &8I:? #DGIFM@E> RJ:8C GFC@:P @E K?< / K?< :8J<
=FI @E;<G<E;<EK =FI<:8JKJ :FEFD@: *FC@:P GG ECB
Working Paper Series No 1344
May 2011
23
56 $L;JFE , 8E; & )N<E JK@D8K@E> ;PE8D@: G8E<C DF;<CJ 8
>L@;< =FI D8:IF<:FEFD@JKJ :FEFD@:J &<KK<IJ GG 56 %8L=D8EE %I88P 8E; ' '8JKILQQ@ !FM<IE8E:< '8KK<IJ 0## !FM
<IE8E:< #E;@:8KFIJ =FI 1FIC; 8EB *FC@:P ,<J<8I:? 1FIB@E> *8G<I
(F 56 %F:QFI '8I:@E &@J9FE -KI8K<>P @DGC<D<EK8K@FE @E E E8CPJ@J
.?< *FC@J? #EJK@KLK< F= #EK<IE8K@FE8C Z8@IJ 18IJ8N '8I:?
56 %FB 1 <K 8C 8:@E> K?< ?8CC<E>< .?< &@J9FE -KI8K<>P =FI !IFNK?
8E; DGCFPD<EK ,<GFIK =IFD K?< "@>? &<M<C !IFLG :?8@I<; 9P 1@D %FB
56 %FG@KJ ! 8E; - -PD8EJBP @J:8C GFC@:P ILC<J #' )::8J@FE8C *8G<I
-<I@<J (F 56 &8D9<IK@E@ & V)E K?< I<;@JKI@9LK@M< GIFG<IKP F= 9L;><K ;<R:@KJW @E $
0FE "8><E 8E; , -KI8L:? <;J 045+565+104 1.+5+%4 #0& +4%#. 1.+%: %CLN<I
:8;<D@: *L9C@J?<I FJKFE FI;I<:?K &FE;FE
56 &8E< . '8IB<K ;@J:@GC@E< #' -K8Z *8G<IJ 56 'FII@J , )E><E8 " 8E; & -:?LBE<:?K .?< I<=FID 8E; @DGC<D<E
K8K@FE F= K?< -K89@C@KP 8E; !IFNK? *8:K )::8J@FE8C *8G<I (F 56 (@:B<CC - $ @8J<J @E PE8D@: 'F;<CJ N@K? @O<; Z<:KJ GG
56 ) :FEFD@: -LIM<P F= K?< LIF I<8 *LKK@E> K?< RJ:8C
?FLJ< @E FI;<I *8I@J $LCP
56 *8;F8E * ,<M@J@E> K?< &@J9FE -KI8K<>P K?IFL>? K?< ) #EEF
M8K@FE -KI8K<>P #E:FD 1FIBJ?FG FE #EEFM8K@FE 8E; !IFNK? *FC@:P @E K?< /
*I8>L< $8EL8IP
56 *<IJJFE . 8E; & -M<EJJFE V1?P 8 -KL99FIE FEJ<IM8K@M< 1FLC; ,LE
8 <R:@K *FC@:P N@K? .@D< #E:FEJ@JK<E:P *I<=<I<E:<JW +L8IK<ICP $FLIE8C F=
:FEFD@:J GG 56 ,F;I@B -L9I8D8E@8E 8E; .I<99@ #EJK@KLK@FEJ ILC< K?< GI@
D8:P F= @EJK@KLK@FEJ FM<I ><F>I8G?P 8E; @EK<>I8K@FE @E <:FEFD@: ;<M<CFGD<EK
$FLIE8C F= :FEFD@: !IFNK? GG 56 ,FL9@E@ ( 8E; $ -8:?J !FM<IED<EK -G<E;@E> 8E; L;><K <R:@KJ @E
K?< #E;LJKI@8C :FEFD@<J (, 1FIB@E> *8G<IJ (F 56 ,LK8 ' 1?P &@J9FE 8@CJ -@=F :FEFD@: -KL;@<J MFC 24
ECB
Working Paper Series No 1344
May 2011
56 -:?<CBC< 1 .?< *FC@K@:8C :FEFDP F= @J:8C *FC@:P F
FI;@E8K@FE
@E '/ IFD @J:@GC@E8I@8E <M@:< KF #EJLI8E:< II8E><D<EK $FLIE8C F=
FDDFE '8IB<K -KL;@<J GG 56 -:?FLK 8E; $FI;8E .?< LIFG<8E /E@FEUJ >FM<IE8E:< 8D9@K@FEJ
8E; @KJ 8;D@E@JKI8K@M< :8G8:@K@<J $FLIE8C F= LIFG<8E *L9C@: *FC@:P MFC EF GG 56 -:?LBE<:?K & / RJ:8C ILC<J @JJL<J 8E; C<JJFEJ =IFD GFC@K@:8C <:FE
FDP #EK<IE8K@FE8C :FEFD@:J 8E; :FEFD@: *FC@:P GG 56 .FDGJFE 1 1.+5+%#. '%101/: 1( 3'(13/ '44104 (31/ 2'04+104 231&
6%5 /#3-'54 #0& .#$163 /#3-'54 +0 5'0 %16053+'4 *8I@J )
56 MFE "8><E $ @J:8C @J:@GC@E< 8E; !IFNK? @E LIFC8E; OG<I@<E:<J
N@K? K?< -K89@C@KP 8E; !IFNK? *8:K 4# 1FIB@E> *8G<I 56 1PGCFJQ
.?< =8@CLI< F=
?KKGNNNMFO<LFI>@E;<OG?GHEF;<
K?<
&@J9FE
JKI8K<>P
ECB
Working Paper Series No 1344
May 2011
25
TABLE 1. Synoptic table: Empirical literature on the effects of the SGP
Author(s)
Estimation
approach /
method
Afonso
and Panel /
Hauptmeier (2010) LSDVc
Sample
Period
Main results
EU-27
19902005
x
x
Cimadomo (2005)
Panel /
Estimation of
linear and nonlinear
relationships
Euro
area
except
GRC
and
LUX
19812005
x
x
x
x
Forni
and Panel /
Momigliano (2005) OLS, FE-IV,
GMM
Gali and Perotti Panel /
FE IV
(2003)
Euro
area,
except
LUX
and
IRL
19932003
EA-11
19802002
x
x
x
x
Annett (2006)
Bernoth
(2010)
et
Panel /
pooled OLS, FE,
2SLS
al. Panel /
FD GMM
EA-11
19802004
x
EU-14
19952006
x
x
Positive impact of the SGP on the
primary balance, but not on primary
expenditure.
Negative impact of the electoral
cycle.
No significant evidence of a procyclical bias in downturns induced
by the SGP.
Tighter policies when public
indebtedness grows.
The policy is a-cyclical in case of
positive output gap.
Fiscal decentralisation contributes
to an increase in the total primary
spending-to-GDP ratio.
The differences between real-time
and ex post data are substantial and
tend to be systematic across time.
No major differences between
OECD and euro area countries
when it comes to fiscal reaction in
bad times.
Discretionary fiscal policy in EMU
countries has become more
countercyclical over time, as
observed in other countries, even if
OECD countries are more
countercyclical.
Decline in public investment
experienced over the last decade by
EMU countries is also part of a
global trend and is smaller than in
OECD countries.
Asymmetric
success
in
the
implementation of the SGP is
linked to the size, volatility as well
as commitment policy of each euro
area country.
The pro-cyclicality of fiscal policies
only arises in the ex post data.
Real time data suggests that
policymakers have tried to run
counter-cyclical discretionary
policy, but find it hard to do.
Note: LSDVc is the least square dummy variable estimator corrected for a dynamic panel data
setting; FE is the fixed-effect estimator; FE-IV is the fixed effect estimator using instrumenta
variables to deal with the endogeneity problem; 2SLS is the two-stage least square estimatorl
;
GMM is the generalised method of moments estimator; FD GMM is the first-difference
GMM (Blundell-Bond) estimator.
26
ECB
Working Paper Series No 1344
May 2011
TABLE 2. List of countries
EU
Austria
Belgium
Bulgaria
Cyprus
Czech Republic
Denmark
Estonia
Finland
France
Germany
Greece
Hungary
Ireland
Italy
Latvia
Lithuania
Luxembourg
Netherlands
Poland
Portugal
Romania
Slovak Rep.
Slovenia
Spain
UK
Non-EU
Australia
Canada
Iceland
Japan
Republic of Korea
Mexico
New Zealand
Norway
Switzerland
Turkey
USA
ECB
Working Paper Series No 1344
May 2011
27
28
ECB
Working Paper Series No 1344
May 2011
908
956
833
835
818
1070
964
857
907
907
1327
1047
1147
468
1229
1316
1194
1148
1147
1147
1147
1147
905
964
944
913
942
931
322
632
1147
1147
Obs
-0.85
-0.92
-0.54
52.84
1.75
1.89
72.07
44.37
-0.47
0.80
-0.87
63.81
0.07
1.14
0.27
45.36
0.50
0.86
0.13
0.16
0.26
0.05
0.01
0.02
0.02
0.01
0.55
57.88
1.69
2.11
0.17
0.80
Mean
4.08
4.31
3.14
31.85
6.01
3.96
37.45
8.76
2.72
1.40
1.63
28.29
0.26
0.70
0.44
22.26
0.50
0.35
0.34
0.36
0.44
0.22
0.03
0.04
0.03
0.02
0.11
7.15
0.56
1.01
0.37
2.11
Full sample
Std. Dev.
-12.25
-19.27
-14.51
3.79
-8.40
0.02
14.01
17.08
-12.56
0.00
-12.56
16.18
0.00
-0.64
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-0.26
-0.20
-0.31
-0.09
0.21
37.65
0.82
0.21
0.00
0.00
Min
20.40
16.13
7.73
193.49
94.55
23.79
222.08
70.93
12.09
12.09
0.00
152.71
1.00
2.12
1.00
100.00
1.00
1.00
1.00
1.00
1.00
1.00
0.12
0.15
0.16
0.10
0.78
84.59
3.97
4.19
1.00
10.00
Max
544
555
500
517
482
555
555
498
512
512
555
555
555
468
555
555
555
555
555
555
555
555
550
555
555
540
550
550
322
370
555
555
Obs
-0.05
-0.70
-0.40
51.36
1.42
1.71
80.05
43.33
-0.30
0.86
-1.07
56.53
0.08
1.14
0.26
46.03
0.48
0.89
0.27
0.32
0.46
0.10
0.01
0.02
0.02
0.01
0.56
55.72
1.69
1.99
0.34
1.65
Mean
3.87
4.78
3.32
33.62
6.33
3.78
38.45
7.22
2.78
1.48
1.83
26.23
0.26
0.70
0.44
13.38
0.50
0.31
0.44
0.47
0.50
0.30
0.03
0.04
0.03
0.02
0.11
7.00
0.56
0.90
0.47
2.80
1995-2010
Std. Dev.
-12.25
-19.27
-14.51
3.79
-4.28
0.02
16.69
20.60
-10.28
0.00
-10.28
16.18
0.00
-0.64
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-0.16
-0.20
-0.13
-0.09
0.21
39.46
0.82
0.21
0.00
0.00
Min
Sample period: annual data from 1980 to 2010. See text for further explanations on the sources and definitions of the data.
Public deficit
Primary balance
CAB
Public debt
Govt. bond spread vs. US
Real GDP weight
Trade openness
Govt. share of income
Output gap
Output gap>0
Output gap<0
Income per capita
Banking crisis dummy
Rule of Law
Legislative Election Held
Vote share of govt. parties
Plurality
Proportional representation
SGP
SGP_EXT
MAASTRICHT
SGP_PRE03
Employment growth
Per capita income growth
Labour productivity growth
Labour force growth
Employment share of population
Wage share of income
Product Market Regulation
Employment Protection Legislation
LISBON
LISBON_YEARS
TABLE 3. Summary statistics
20.40
16.13
7.22
193.49
94.55
23.79
222.08
62.91
12.09
12.09
0.00
152.71
1.00
2.12
1.00
73.90
1.00
1.00
1.00
1.00
1.00
1.00
0.08
0.15
0.16
0.10
0.78
76.24
3.97
3.76
1.00
10.00
Max
TABLE 4. Unconditional moments of the primary balance and per capita growth
Obs.
Full sample
No SGP
SGP
No SGP_PRE
SGP_PRE
No MAASTRICHT
MAASTRICT
No SGP_PRE03
SGP_PRE03
956
806
150
778
178
662
294
899
57
Full sample
No Lisbon
Lisbon
964
774
190
Mean
Std. dev.
Primary balance
-0.92
4.31
-1.17
4.42
0.39
3.39
-1.27
4.44
0.58
3.3
-1.41
4.56
0.18
3.46
-1.12
4.33
2.16
2.36
Per capita growth
2.2
3.7
2.4
3.5
1.2
4.4
Note: See text for the definition of the dummy variables. All data are in percentage points per year. The f
The full sample goes from 1980 to 2010 (annual data).
ECB
Working Paper Series No 1344
May 2011
29
30
ECB
Working Paper Series No 1344
May 2011
Cyclically adjusted primary balance (t-1)
Output gap
Banking crisis
Government share of income (t-1)
Income per capita (t-1)
Size (t-1)
Trade openness (t-1)
Debt to GDP (t-1)
Primary balance (t-1)
Output gap<0
Output gap>0
0.67***
(0.049)
0.022***
(0.0082)
0.037***
(0.013)
1.02***
(0.35)
-0.041*
(0.024)
0.059
(0.036)
-0.90
(0.59)
0.35***
(0.083)
(1)
0.62**
(0.25)
0.15
(0.14)
0.68***
(0.050)
0.020**
(0.0081)
0.035***
(0.013)
1.02***
(0.35)
-0.046*
(0.023)
0.068*
(0.037)
-0.90
(0.62)
(2)
(3)
0.017*
(0.0090)
0.23***
(0.072)
0.68***
(0.040)
0.029***
(0.0072)
0.031**
(0.012)
0.91**
(0.36)
Dependent variable: Primary balance (unless otherwise specified)
Political and institutional variables:
Vote Share of Government Parties
Economic variables:
TABLE 5. Average fiscal behaviour
0.020*
(0.011)
0.30***
(0.10)
0.63***
(0.051)
0.041***
(0.012)
0.019
(0.015)
1.14**
(0.49)
(4)
1992-2010
0.023**
(0.011)
0.13**
(0.065)
0.66***
(0.037)
0.026***
(0.0065)
0.038***
(0.012)
0.96**
(0.41)
(5)
No crisis:
1980-2007
0.0063
(0.0075)
0.061
(0.054)
0.66***
(0.047)
0.027***
(0.0062)
0.036***
(0.012)
0.87***
(0.32)
(6)
CAB (b)
ECB
Working Paper Series No 1344
May 2011
31
717
0.641
33
(a)
0
717
0.630
33
(a)
4.1e-10
731
0.642
34
(a)
0
-0.50***
(0.16)
1.03*
(0.57)
560
0.579
34
(a)
0
-0.46**
(0.19)
0.021
(0.80)
628
0.675
33
(a)
0
-0.46***
(0.15)
1.09**
(0.50)
692
0.625
32
(a)
0
-0.40***
(0.12)
0.43
(0.43)
Note: Robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1. The equation is estimated by instrumental variables (GMM), where the potentially endogenous
variable is the output gap. The instrument is one lag of the output gap, hence the equation is exactly identified. The sample period is 1980-2010 unless otherwise specified.
Time and country fixed effects are always included in the model.
(a) The equation is exactly identified.
(b) Cyclically adjusted primary balance.
Observations
R-squared
Number of country
J test (P value)
Kleibergen-Paap test for underidentification (P value)
Plurality
Legislative Election Held
32
ECB
Working Paper Series No 1344
May 2011
Deficit>2%(t-1)
Trade openness (t-1)
Size (t-1)
Debt to GDP (t-1)
Primary balance (t-1)
Output gap
SGP dummies:
SGP_PRE03
MAASTRICHT
SGP_PRE
SGP
Plurality
Vote Share of Government Parties
Political and institutional variables:
Legislative Election Held
Economic variables:
TABLE 6. Introducing the SGP dummies
-0.14
(0.30)
-0.50***
(0.15)
0.018*
(0.0090)
1.04*
(0.57)
0.24***
(0.072)
0.68***
(0.040)
0.029***
(0.0071)
0.91**
(0.36)
0.032**
(0.012)
(1)
Dependent variable: Primary balance
-0.020
(0.31)
-0.50***
(0.15)
0.017*
(0.0090)
1.03*
(0.57)
0.23***
(0.072)
0.68***
(0.040)
0.029***
(0.0072)
0.91**
(0.36)
0.032**
(0.013)
(2)
0.51*
(0.30)
-0.49***
(0.15)
0.014
(0.0090)
0.78
(0.59)
0.23***
(0.071)
0.68***
(0.040)
0.031***
(0.0073)
0.93***
(0.36)
0.030**
(0.012)
(3)
-0.16
-0.50***
(0.16)
0.017*
(0.0090)
1.04*
(0.57)
0.24***
(0.073)
0.68***
(0.040)
0.029***
(0.0072)
0.91**
(0.36)
0.032**
(0.012)
(4)
0.080
(0.33)
-0.50***
(0.15)
0.018**
(0.0090)
1.15**
(0.58)
0.24***
(0.072)
0.68***
(0.043)
0.029***
(0.0073)
0.94***
(0.36)
0.028**
(0.013)
0.18
(0.25)
(5)
ECB
Working Paper Series No 1344
May 2011
33
731
0.642
34
(a)
0
731
0.642
34
(a)
0
731
0.643
34
(a)
0
731
0.642
34
(a)
0
(0.30)
731
0.644
34
(a)
0
-0.87**
(0.40)
Note: Robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1. The equation is estimated by instrumental variables (GMM), where the potentially endogenous
variable is the output gap. The instrument is one lag of the output gap, hence the equation is exactly identified. The sample period is 1980-2010 unless otherwise specified.
Time and country fixed effects are always included in the model.
For the definition of the dummy variables, see text.
(a) The equation is exactly identified.
Observations
R-squared
Number of country
J test (P value)
Kleibergen-Paap test for underidentification (P value)
Deficit>2%(t-1)*SGP
34
ECB
Working Paper Series No 1344
May 2011
Trade openness (t-1)
Size (t-1)
Debt to GDP (t-1)
Primary balance (t-1)
Output gap
SGP
Debt to GDP(t-1)*SGP
Output gap*SGP
SGP dummies and interactions:
Plurality
Vote Share of Government Parties
Political and institutional variables:
Legislative Election Held
Government bond spread vs. the USA squared, t-1
Government bond spread vs. the USA, t-1
Economic variables:
TABLE 7. Introducing interaction terms
0.97
(0.75)
0.15*
(0.088)
-0.0042
-0.44***
(0.16)
0.017*
(0.0091)
1.11*
(0.61)
0.22***
(0.072)
0.68***
(0.040)
0.029***
(0.0071)
0.99***
(0.37)
0.034***
(0.013)
(1)
-0.45***
(0.16)
0.016*
(0.0091)
1.12*
(0.61)
0.22***
(0.073)
0.68***
(0.040)
0.030***
(0.0072)
1.01***
(0.37)
0.030**
(0.013)
(2)
Dependent variable: Primary balance
-0.47***
(0.16)
0.013
(0.0088)
0.90
(0.60)
0.28***
(0.097)
0.68***
(0.041)
0.031***
(0.0077)
0.89**
(0.37)
0.031**
(0.012)
(3)
-0.50***
(0.16)
0.018*
(0.0091)
1.03*
(0.58)
0.24***
(0.075)
0.68***
(0.040)
0.029***
(0.0073)
0.89**
(0.36)
0.032**
(0.013)
(4)
0.50
(0.73)
0.20*
(0.12)
0.0035
-0.31**
(0.14)
0.011
(0.0088)
0.53
(0.58)
0.24***
(0.073)
0.66***
(0.046)
0.025***
(0.0074)
0.77**
(0.37)
0.053***
(0.013)
-0.030
(0.042)
0.00056
(0.00042)
(5)
0.97
(0.78)
0.15*
(0.088)
-0.0042
-0.44***
(0.16)
0.017*
(0.0091)
1.11*
(0.62)
0.22***
(0.072)
0.68***
(0.040)
0.029***
(0.0071)
1.00***
(0.37)
0.034***
(0.013)
(6)
ECB
Working Paper Series No 1344
May 2011
35
Observations
R-squared
Size t-1*SGP
Government bond spread vs. the USA squared, t-1*SGP
Government bond spread vs. the USA, t-1*SGP
Legislative Election Held*SGP_PRE03
Debt to GDP(t-1)*SGP_PRE03
Output gap*SGP_PRE03
SGP_PRE03
Legislative Election Held*MAASTRICHT
Debt to GDP(t-1)*MAASTRICHT
Output gap*MAASTRICHT
MAASTRICHT
Legislative Election Held*SGP_PRE
Debt to GDP(t-1)*SGP_PRE
Output gap*SGP_PRE
SGP_PRE
Legislative Election Held*SGP
731
0.650
(0.0078)
-0.21***
(0.066)
731
0.647
0.75
(0.70)
0.11
(0.082)
-0.0042
(0.0076)
-0.13**
(0.053)
731
0.645
1.03
(0.64)
-0.13
(0.11)
-0.0061
(0.0081)
-0.059
(0.047)
731
0.642
0.55
(0.69)
-0.14
(0.16)
-0.0053
(0.0066)
-0.079
(0.088)
667
0.689
0.53
(0.43)
0.25
(0.25)
(0.0078)
-0.22***
(0.063)
731
0.650
0.0088
(0.17)
(0.0077)
-0.21***
(0.066)
36
ECB
Working Paper Series No 1344
May 2011
34
(a)
0
34
(a)
0
34
(a)
0
34
(a)
0
33
(a)
0
34
(a)
0
Note: Robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1. The equation is estimated by instrumental variables (GMM), where the potentially endogenous
variable is the output gap. The instrument is one lag of the output gap, hence the equation is exactly identified. The sample period is 1980-2010 unless otherwise specified.
Time and country fixed effects are always included in the model.
For the definition of the dummy variables see text, Section 3.
(a) The equation is exactly identified.
Number of country
J test (P value)
Kleibergen-Paap test for underidentification (P value)
ECB
Working Paper Series No 1344
May 2011
37
470
0.327
28
0.17***
(0.065)
-0.0013***
(0.00026)
0.00076***
(0.00014)
0.0026
(0.0034)
0.0060
(0.0085)
-0.010**
(0.0043)
-0.0035
(0.0033)
(4)
1992-2010
581
0.311
28
0.24***
(0.060)
-0.0011***
(0.00019)
0.00078***
(0.00013)
0.0051*
(0.0027)
0.030**
(0.015)
-0.0092**
(0.0041)
-0.00024
(0.0032)
(5)
No crisis: 19802007
538
0.453
26
0.34***
(0.067)
-0.0012***
(0.00019)
0.00072***
(0.00016)
0.0064**
(0.0026)
0.022
(0.016)
-0.0064**
(0.0029)
-0.00036
(0.0030)
(6)
Excluding
very low
income per
capita
Note: Pooled OLS with robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1. The sample period is 1980-2010 unless otherwise specified. Time and country
fixed effects are always included in the model. For the definition of the dummy variables see text, Section 3.
611
0.354
28
611
0.353
28
0.25***
(0.059)
-0.0012***
(0.00019)
0.00071***
(0.00012)
0.0048*
(0.0026)
0.029**
(0.015)
-0.0094**
(0.0039)
Observations
R-squared
Number of country
0.25***
(0.059)
-0.0012***
(0.00019)
0.00071***
(0.00012)
0.0055**
(0.0026)
0.029**
(0.014)
-0.0094**
(0.0039)
-0.00059
(0.0031)
(3)
-0.00061
(0.00064)
611
0.353
28
0.25***
(0.059)
-0.0012***
(0.00019)
0.00071***
(0.00012)
0.0056**
(0.0026)
0.029**
(0.014)
-0.0094**
(0.0039)
(2)
LISBON_YEARS
LISBON
Stability
Proportional Representation
Employment Protection Legislation
Trade openness (t-1)
Income per capita (t-1)
Per capita income growth (t-1)
(1)
Dependent variable: per capita income growth
TABLE 8. The Lisbon Strategy: per capita income growth
38
ECB
Working Paper Series No 1344
May 2011
657
0.349
37
0.20***
(0.056)
-0.00091***
(0.00026)
0.00023
(0.00014)
0.0014**
(0.00067)
-0.0064**
(0.0032)
(4)
1992-2010
757
0.162
37
0.20***
(0.057)
-0.00056***
(0.00015)
0.00026*
(0.00015)
0.0014***
(0.00050)
-0.0029
(0.0027)
(5)
No crisis: 19802007
Note: Pooled OLS with robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1. The sample period is 1980-2010 unless otherwise specified. Time and country
fixed effects are always included in the model.
For the definition of the dummy variables see text, Section 3.
865
0.332
37
865
0.332
37
0.19***
(0.052)
-0.00084***
(0.00016)
0.00022*
(0.00013)
0.0010**
(0.00047)
Observations
R-squared
Number of country
0.19***
(0.053)
-0.00080***
(0.00016)
0.00023*
(0.00013)
0.00099**
(0.00047)
-0.0052*
(0.0027)
(3)
-0.00097**
(0.00045)
865
0.329
37
0.20***
(0.052)
-0.00087***
(0.00016)
0.00021*
(0.00013)
0.00092*
(0.00047)
(2)
LISBON_YEARS
LISBON
Wage share of income (t-1)
Trade openness (t-1)
Income per capita (t-1)
Labour productivity growth (t-1)
(1)
Dependent variable: labour productivity growth
TABLE 9. The Lisbon Strategy: labour productivity growth
ECB
Working Paper Series No 1344
May 2011
39
636
0.492
36
0.25**
(0.11)
0.00025
(0.00017)
0.43***
(0.16)
-0.30***
(0.064)
0.00017**
(0.000086)
-0.00025
(0.0026)
(4)
1992-2010
726
0.439
36
0.27**
(0.11)
0.00020
(0.00016)
0.52***
(0.13)
-0.22***
(0.045)
0.00024**
(0.00011)
-0.000073
(0.0023)
(5)
No crisis: 1980-2007
Note: Pooled OLS with robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1. The sample period is 1980-2010 unless otherwise specified. Time and country
fixed effects are always included in the model.
For the definition of the dummy variables see text, Section 3.
831
0.513
36
831
0.511
36
0.26***
(0.092)
0.00027**
(0.00013)
0.54***
(0.13)
-0.23***
(0.042)
0.00013*
(0.000075)
Observations
R-squared
Number of country
0.27***
(0.091)
0.00027**
(0.00013)
0.54***
(0.13)
-0.23***
(0.042)
0.00012*
(0.000073)
-0.0022
(0.0022)
(3)
-0.00080**
(0.00034)
831
0.511
36
0.27***
(0.091)
0.00026**
(0.00013)
0.54***
(0.13)
-0.23***
(0.042)
0.00011
(0.000073)
(2)
LISBON_YEARS
LISBON
Vote Share of Government Parties
Employment share of income (t-1)
Labour force growth
Trade openness (t-1)
Employment growth (t-1)
(1)
Dependent variable: employment growth
TABLE 10. The Lisbon Strategy: employment growth
Wo r k i n g Pa p e r S e r i e s
N o 1 1 1 8 / n ov e m b e r 2 0 0 9
Discretionary
Fiscal Policies
over the Cycle
New Evidence
based on the ESCB
Disaggregated Approach
by Luca Agnello
and Jacopo Cimadomo