MSc Public Administration: International and European Governance
Master Thesis
Supervisor: Dr. Alexandre Afonso
Second reader: Prof.dr. Bernard Steunenberg
FINAL THESIS
Macroeconomic Imbalance Procedure: Clashing Preferences
Between Commission and Council
Alessandro Gasparotti s1792229
Submitted on: 10/01/2018
Abstract: The introduction of the Macroeconomic Imbalance Procedure, as part of the reformed EU economic governance, poses questions on which institution is in control of the process. When the Commission drafts country specific assessments and recommendations at the end of the monitoring cycle, how much and with what intensity do member states in Council modify these prescriptions? Using textual analysis on the official documents of both institution we try to explain the extent of changes and their direction, through the employment of institutional, political, and economic predictors. While changes decreased in comparison to similar procedures before the crisis, suggesting the empowerment of the Com-mission, changes are still partially explained by well known Council’s dynamics.
Foreword
The writing of this dissertation is the culmination of a journey started
at Leiden University and continued at SciencesPo in Paris. Embarking in
writing an original piece of research is a daunting but rewarding
activ-ity. I would like to thank my supervisor, who accepted my request and
agreed on supervising my progress from abroad, thus allowing me to
en-rich my economics’ knowledge in Paris. His patience and support have
been key in completing this thesis. My gratitude goes especially to my
for-mer colleagues at the Clingendael Institute, whose knowledge and
kind-ness guided me toward understanding what I really hope to obtain
pro-fessionally, this whole research was inspired by my time working there.
Gratitude also goes to friends and family, people met along the road
dur-ing this experience: people who inspired me and supported me all along.
A final thanks goes to Alvise, a long-time friend, who invested his time
List of Abbreviations
AMR: Alert Mechanism Report CAP: Correction Action Plan
COREPER: Permanent
Representa-tives Committee
CSR: Country Specific
Recommen-dation
ECB: European Central Bank
ECFIN (DG): Directorate-General
for Economic and Financial Affairs
EDP: Excessive Deficit Procedure EFC: Economic and Financial
Com-mittee
EIP: Excessive Imbalance Proce-dure
EMCO: Employment Committee EMU: Economic and Monetary Union
EU: European Union
EZ: Euro Zone
GDP: Gross Domestic Product IDR: In Depth Review
IMF: International Monetary Fund MIP: Macroeconomic Imbalance Procedure
OMC: Open Method of Coordination PMR: Poisson Model Regression QMV: Qualified Majority Voting RQMV: Reverse Qualified Majority
Voting
SGP: Stability and Growth Pact SPC: Social Protection Committee TEU: Treaty on the European Union TFEU: Treaty on the Functioning of
the European Union
Country Abbreviations AT: Austria (e) BE: Belgium (e) BG: Bulgaria CY: Cyprus (e) CZ: Czech Republic DE: Germany (e) DK: Denmark EE: Estonia (e) EL: Greece (e) ES: Spain (e) FI: Finland (e) FR: France (e) HR: Croatia HU: Hungary IE: Ireland (e) IT: Italy (e) LT: Lithuania (e) LU: Luxembourg (e) LV: Latvia (e) MT: Malta (e) NL: Netherlands (e) PL: Poland PT: Portugal (e) RO: Romania SE: Sweden SI: Slovenia (e) SK: Slovakia (e) UK: United Kingdom
Contents
Introduction 7
Economic governance reform . . . 8
Macroeconomic Imbalance Procedure: legalistic introduction . . . . 10
Reforms and critiques . . . 14
Research question . . . 18
Theory 20 International contracts and MIP . . . 20
European Commission: policy entrepreneur and watchdog . . . 23
Member states’ preferences determinants . . . 28
Logic of action in the Council . . . 30
Hypotheses . . . 31
Research Design 34 Member states’ behaviour . . . 36
Variables and Operationalization . . . 38
Population and case selections . . . 38
Dependent variables . . . 38
Explanatory variables . . . 43
Analysis 46 Descriptive Statistics . . . 46
Weakening Commission’s Assessments . . . 49
Editing Commission’s Assessments . . . 52
Discussion . . . 56
Academic Implications . . . 62
Practical Implications . . . 64
Study Limitations . . . 65
Future Research Directions . . . 67
Bibliography 70 Appendix 88 List of variables . . . 88 Correlation Table . . . 91 Further Tests . . . 92 List of Tables 1 Macroeconomic Imbalance Procedure findings for EU mem-ber states . . . 38
2 Example of coding weakening variable . . . 42
3 Summary statistics . . . 48
4 Generalized logistic regression model for the probability of weakening Commission’s texts . . . 51
5 Negative binomial regression model for the probability of edit-ing Commission’s texts . . . 53
6 Pearson Correlation . . . 91
7 Weakening odd ratios . . . 92
8 Generalized logistic regression model for the probability of weakening Commission’s texts - euro membership . . . 93
9 Generalized logistic regression model for the probability of
weakening Commission’s texts - euro out . . . 94
10 Negative binomial regression model for the probability of
edit-ing Commission’s texts - euro membership . . . 95
11 Negative binomial regression model for the probability of
edit-ing Commission’s texts - euro out . . . 96
List of Figures
1 European Semester Timetable . . . 11
2 Longitudinal data on the dependent variable edit . . . 39
3 Longitudinal data on the dependent variable weakening . . . . 40
4 Determinants of weakening and editing Commission’s
recom-mendations . . . 56
5 Number of edits and textual changes leading to a weakening
Introduction
The financial, economic and sovereign debt crisis that hit the European
Union (EU) in 2008 had, among various consequences, the rethinking of its
economic governance architecture, which proved flawed and ill-prepared
to anticipate, avoid, and manage such turmoil (P. Lane, 2012). Among the
shortcomings of the Economic and Monetary Union (EMU) exposed by the
crisis, the narrow focus on budgetary indicators such as public deficits
and debts did not prevented the piling up of dangerous imbalances (De
Grauwe, 2013), and the whole monitoring of these indicators, inscribed in
the Stability and Growth Pact (SGP), lacked compliance by member states
(Savage & Verdun, 2007). As a consequence, some countries respecting
the SGP criteria ended up caught in turbulent times, like Spain and
Ire-land, while others - notably Greece - not having sound finances when the
crisis hit, had no room of manoeuvre for countercyclical fiscal policies and
were forced to implement austerity measures that contracted its GDP by
25% (Constâncio, 2014).
To adjust the institutional architecture, a bold set of reforms was
intro-duced - leading to the most rapid deepening of integration in EU history
(Jones, Kelemen, & Meunier, 2016) - reinforcing existing mechanisms and
establishing new ones to overcome the limitations previously neglected.
Other than establishing ad hoc mechanisms for financial assistance to
dis-tressed member states, with the adoption of the six-pack1 and two-pack2
-1Entered into force December the 13th 2011. Consists of five regulations and one di-rective. Regulation No 1175/2011, Regulation No 1177/2011, Regulation No 1173/2011, Directive No 2011/85/EU refer to fiscal policy, while Regulation No 1176/2011 and Reg-ulation No 1174/2011 address macroeconomic imbalances.
2Entered into force May the 30th2013. Consists of two regulations ensuring additional coordination: Regulation No 472/2013 and Regulation No 473/2013.
respectively agreed in 2011 and 2013 - the Euro Zone (EZ) tightened fiscal
surveillance and established a new procedure, the Macroeconomic
Imbal-ance Procedure (MIP), for the prevention and correction of macroeconomic
imbalances, inscribed in the new streamlined annual cycle of policy
co-ordination for economic governance, the European Semester (Stylized in
Figure 1).
Since 2012 then, other than a reinforced SGP to which it is a complement,
the MIP assesses macroeconomic developments within member states, in
order to correct excessive imbalances before they can threat the smooth
functioning of the single market and the common currency (European
Commission, 2016c, p. 18).
Economic governance reform
The Macroeconomic Imbalance Procedure is considered by some
au-thors as a true break-though to overcome the short-comings of EMU
archi-tecture if applied symmetrically (De Grauwe, 2012), it is at times presented
as a new layer in already existing mechanisms, expanding their reach but
reproducing flawed governance structure (Verdun, 2015).
Having recognised the inadequacy of the pre crisis institutional
architec-ture of the EMU, in September 2010 the Commission presented the
Six-Pack legislative package (European Commission, 2010a, 2010b), aimed at
reinforcing surveillance and discipline in the economic governance of the
EU. Other than strengthening the SGP, the novelty of the package was the
Macroeco-nomic Imbalance Procedure (MIP). After a quick discussion,3 the Six-Pack
was approved and entered into force in December 2011. It’s legalistic basis
is enshrined in the EU Regulation No 1176/2011 on the prevention and
corrections of macroeconomic imbalances, which applies to all member
states, and the EU Regulation No 1174/2011 on enforcement measures to
correct excessive macroeconomic imbalances, which only applies to Euro
Zone members and entails procedures that can lead to sanctions.
Before discussing the MIP and the framework in which it operates, a
definition of what is considered an imbalance is presented:4
‘Imbalances means any trend giving rise to macroeconomic developments
which are adversely affecting, or have the potential adversely to affect, the
proper functioning of the economy of a Member State or of the economic and
monetary union, or of the Union as a whole.’ Regulation No 1176/2011
(article 2(1))
And furthermore:
‘Excessive imbalances means severe imbalances, including imbalances
that jeopardise or risks jeopardising the proper functioning of the economic
and monetary union.’ Regulation No 1176/2011 (article 2(2))
While apparently the definition of imbalance is clear in the legislation,
when it comes to assessing when one imbalance has to be considered
ex-3Interestingly, as noted in an interview given by one ECB’s top official to Moschella (2014, p. 1274) “The lack of a debate on IMF (that has an analogous mechanism in place and decades of experience on such monitoring) was conspicuous by its absence”, hinting that the unusual urgency to respond to the issue might have induced policy-makers to focus on delivering quickly instead of the quality of the response.
4A more precise definition was given by Thomas Wieser, former President of the EU Economic and Financial Committee (EFC): ‘A macroeconomic imbalance is the (negative or
positive) position of a domestic, external or financial variable... [which] may - if uncorrected over time - make the national savings/investment balance so untenable that it self-corrects
cessive is a controversial topic. Political scientists noted that the
vague-ness of the definition can result in excessive discretionary powers to EU
authorities (Chalmers, 2012) and consequent illegitimate constraints to
national decision-making (Schmidt, 2015).
Moreover, from an economic perspective, an imbalance is not inherently
dangerous (Blanchard & Giavazzi, 2002). Imbalances, most notably in the
current account, depend on many factors, among which comparative
ad-vantages and country specialization (Hunt & Morgan, 1995), political and
institutional preferences described by the scholarship of varieties of
cap-italism (Hall & Soskice, 2001), the Balassa (1964) Samuelson (1964)
hy-pothesis in time of economic catching-up (Tressel et al., 2014), and many
more (Wyplosz, 2016). Notwithstanding their origin, the crisis has made
clear their impact on the stability of the EMU. As Buti and Carnot (2012)
-two officials at the Commission - highlighted, the crisis cannot be simply
explained by fiscal profligacy in some member states, but it is part of a
wide set of financial and macroeconomic developments that were
under-estimated and now are monitored in the MIP.
Macroeconomic Imbalance Procedure: legalistic introduction
As Figure 1 shows, the starting point of the new surveillance
proce-dure is the Alert Mechanism Report (AMR), prepared by the Commission,
which on a country-by-country basis identifies issues that necessitate a
review. Although the Commission examines the economic indicators
pre-This subsection is mainly based on actual legislation, a Compendium produced by the Commission (European Commission, 2016c), and the Commission’s website: https: //ec.europa.eu/info/business-economy-euro/economic-and-fiscal-policy-coordination/ eu-economic-governance-monitoring-prevention-correction/european-semester/
Figure 1. European Semester Timetable6
6Retrieved from: (last access 09/01/2018)
sented in a scoreboard encompassing both external competitiveness and
internal imbalances - with upper and lower bounds - there is no automatic
interpretation of the results, meaning that qualitative analyses are
under-taken, giving the Commission discretion and flexibility in its assessments.
The conclusions AMR are discussed in the EU’s Council of Economic and
Finance Ministers and in the Eurogroup for matters related to EZ
mem-ber states,5 after which, taking into account the feedbacks from member
states, the Commission decides which countries needs further
investiga-tion and prepares country-specific in depth reviews (IDR), as stated in
article 5 of Regulation No 1176/2011.
IDRs are far-reaching and account for country peculiarities. Surveillance
missions are held in member states, opinion by the European Central
Bank (ECB) officials can be requested for EZ members, and country
spe-cific situation are assessed against spillover effects in the rest of the Union.
The conclusion of IDRs, depending on the result of the assessment, can
lead to three different outcomes.
1. If no imbalance is detected - or in the case the member states was not
subjected to IDR - the Commission will avoid further action, for the
rest of the cycle of surveillance.
2. If imbalances are detected - even in the case of excessive ones that do
not require enhanced surveillance - the Council agrees, by Qualified
Majority Voting (QMV), on preventive country specific
recommenda-tions (CSR) drafted by the Commission to correct imbalances (article
5As noted in Vanhercke, Zeitlin, and Zwinkels (2015) and Zeitlin (2014), the Committee of Permanent Representatives (COREPER) is more and more involved in mediating debates so to provide a broader discussion, with the aim of including experts of different policy-fields, due to the overreaching nature of macroeconomic surveillance.
6(4)).
MIP related CSRs are issued in the larger framework of the European
Semester, together with recommendations related to the SGP and the
integrated guidelines.
3. If imbalances are deemed excessive and requiring corrective action,
Chapter III of Regulation No 1176/2011 details the functioning of the
Excessive Imbalance Procedure (EIP). In case of non compliance, EZ
member states can receive pecuniary sanctions, as described in
Reg-ulation No 1174/2011.
In a nutshell, the Council can adopt by QMV a recommendation
drafted by the Commission in which it recognises the imbalances as
excessive, and specifies a set of policies to be followed with a deadline
upon which the concerned member state has to submit a Correction
Action Plan (CAP). The full execution of policy recommendations puts
the procedure in abeyance, but its conclusion is dependent on the
correction of the imbalance, meaning that not implementing the
poli-cies in the recommendation but achieving a satisfying result is the
key objective. 7
The Commission is tasked with monitoring implementation of the
CAP. If the Council considers that the member states has not taken
corrective measures, it can adopt a Commission recommendation
of non-compliance setting new deadlines. Such recommendation is
adopted within 10 days unless rejected by reversed qualified majority
voting (RQMV) (article 10(4)).
7Member state ownership it thus encouraged. Even the CAP is drafted by the con-cerned member state and approved or rejected by the Council on a Commission report. It the CAP is rejected twice, fines can be imposed for EZ members.
For EZ member states, Regulation No 1174/2011 envisages an
an-nual fine when two consecutive CAPs are rejected (article 3(2)),
in-terest bearing deposit in case of non-compliance with the CAP
(ar-ticle 3(1)), and annual fine in case of two Council decision on
non-compliance (article 3(2)). The yearly amount of sanctions and deposit
recommended by the Commission are 0.1% of the country’s gross
do-mestic product (GDP) in the preceding year (article 5). Sanctions are
adopted by the newly introduced RQMV: only EZ member states vote,
minus the concerned one (article 5). If after 10 days the
recommen-dation is not rejected, it is automatically adopted.
Reforms and critiques
Since its inception, the MIP has been adjusted and modified several
times, suggesting that the Commission takes advantage of the vagueness
characterising Regulation No 1176/2011 to overcome rising shortfalls,
possibly due to the limited investment in time during the drafting phase.
Such changes involved (i) the set of indicators used to produce the AMR,
(ii) the categorisation of imbalances in the IDRs’ conclusions, and (iii) the
terminology and policy priorities set out in CSRs.
Indicators referring to social conditions were added in order to better grasp
what these imbalances mean for affected citizens (European Commission,
2015c); such a turn indicates both increased attention by the Commission
toward transparency (part of throughput according to Schmidt (2013)),
supporting output legitimacy, and the successful lobbying of the
Employ-ment Committee (EMCO) and Social Protection Committee (SPC) within
by the Directorate General for Economic and Financial Affairs (ECFIN) in
analysing social matters and producing CSRs (Zeitlin & Vanhercke, 2014;
Vanhercke et al., 2015).
The categorization of imbalance, and consequently their severity, was
mod-ified almost every year, giving importance either to gravity or the urgency
of policy response, until the process was streamlined in 2016 to enhance
transparency in the MIP implementation (European Commission, 2015f).
Policy priorities and the very terminology adopted in CSRs evolved during
the years, with a revitalised attention to flexicurity and poverty reduction
(Bekker, 2015, 2017), disproving the assumption of a Commission
inher-ently neoliberal, operating to undermine the welfare state (Joerges, 2014).
The Commission then, when empowered with new competencies, operated
in a rational and dynamic framework, through ideational change,
peer-pressure and learning (Lombardi & Woods, 2008), both within the
institu-tion and regarding its relainstitu-tions with member states.
While an EIP - where the Commission enjoys expanded power
vis-à-vis member states - has never been initiated, most countries experienced
imbalances of some sort, 8 leading the Council to adopt so far 296 MIP
related CSRs 9 within the European Semester, coupling these
recommen-dations with the ones related to the SGP, forming a coherent framework
for economic surveillance (European Commission, 2013a).
The MIP framework resembles the SGP’s one, at the point that in the
yearly cycle of the European Semester the two tools follow the same
time-line (Figure 1) and CSRs are released for both mechanisms, suggesting
8In 80 out of the 89 IDRs produced member states were found experiencing imbalances (58 times), or excessive imbalances (22 times)
that the flaws exposed in the fourteen year of SGP’s history before the
in-troduction of the Six-Pack - lot of research has been produced regarding
the shortcomings of the SGP, see for an introduction Baerg and Hallerberg
(2016), Heipertz and Verdun (2010) - might not have been properly
ad-dressed, with the risk of reproducing them, thus undermining the whole
rationale of enhanced surveillance.
The definition of excessive imbalances is broad and extremely vague,
giv-ing discretion to the Commission, and could imply overreach (Chalmers,
2012; Joerges & Weimer, 2012). CSRs - although not binding - are now
part of the MIP mechanism,10 meaning that sanctions could be triggered
in case of non compliance, expanding the power of the institution. For
that reason MIP has been seen as exemplary of the post-crisis
reasser-tion of the EU legal order’s harder edge (Dawson, 2015a, p. 1). Member
states on the other hand committed to rebalancing macroeconomic
indi-cators within the reinforced framework of economic governance. Hodson
(2017) considers however such framework as a limited source of binding
recommendations, describing it more as an instrument of peer-pressure,
where government face the precision of recommendations that they
ex-pected when signing it into law.
Others criticise the procedure as a move toward executive dominance
(Crum, 2013; Curtin, 2014), based on ‘governing by the rules and ruling by
the numbers’(Schmidt, 2015). On the other side of the debate we find some
10Previously recommendations were produced as an instrument of peer pressure within the Open Method of Coordination (OMC), based on article 121 TFEU, but they had no enforcement mechanism, being thus a mere suggestion. The difference is substantial, due to the fact the EU competences are not encompassing all the fields regarding which CSRs are published, posing questions of legitimacy once sanctions can be issued to non-complying countries.
think tanks arguing that peer pressure is failing, implementation rates of
CSRs are declining and there is little we can expect from the Semester or
the MIP (Gros & Alcidi, 2015). MIP is seen as a very weak mechanism
(Sapir & Wolff, 2015) or as ineffective (Darvas & Leandro, 2015).
Focusing on the legalistic perspective could however miss important
developments, for instance the desire of the Juncker’s Commission to
cre-ate political ownership and engaging in genuine dialogue with the
mem-ber states (European Commission, 2015d). Vagueness of the regulation
and consequent high level of discretion in the Commission’s assessments,
coupled with such renewed political turn pose questions on whether hard
compliance is indeed the reality of post-crisis EU, or if instead a new
bal-ance of power arose between the Commission and member states
regard-ing the Macroeconomic Imbalance Procedure.
Some scholars consider these new oversight powers as an expansion
of the Commission’s prerogatives in the face of member states, describing
them as a new wave of supranationalism (Dehousse, 2015; Schmidt, 2016;
Bauer & Becker, 2014; Laffan & Schlosser, 2016), while others concluded
that the new framework empowered member states in the Council, where
actual decision are taken (Puetter, 2012; Bickerton, Hodson, & Puetter,
2015; da Conceição-Heldt, 2016; Chang, 2013). Even though Bauer and
Becker (2014), drawing from Börzel (2005), show a medium expansion in
the breadth, and a strong expansion in the depth of Commission’s role in
economic policy surveillance since the crisis, and the faith in a more active
Commission, whose oversight powers are enhanced vis-à-vis national veto
player through the introduction of RQMV (Dawson, 2015b) at the time of
mea-sures under the MIP,11 although some member states have been under
scrutiny since its inception, in some case even on a deteriorating path.12
The preference for not exercising its powers at full is somehow puzzling,
and could be explained by the fact that the Commission tries to avoid
open conflict with member states, preferring softer forms of steering, or
that member states still retain enough power throughout the process to
block - or discourage - Commission’s action. Given Dawson (2015b) praise
for RQMV, whose application is only activated after the triggering of an EIP,
the Commission’s inability - or unwillingness - to trigger such a procedure
hints that careful attention has to paid to the steps leading to it, where
Council (Commission) could battle over blocking (pushing for) it.
Moreover, apart from few notable contributions (Bekker, 2016; Savage &
Verdun, 2016), little attention has been given on how the Commission
fulfils its new mandate in monitoring post-crisis macroeconomic
develop-ment.
Research question
The goal of this research is thus to shed light on the relations between
member states and the Commission in the new MIP framework. The main
focus will be the its final phase, when the Council agrees on Commission’s
assessments and draft policy-recommendations, maintaining the right of
amending them.
11On a similar way, the Excessive Deficit Procedure (EDP) in the SGP framework has been opened several times - due to the stringent numeric criteria to respect - yet in the only case where sanctions were about to become a reality, the Commission proposed a symbolic 0% fine to both Spain and Portugal (European Commission, 2016b)
12Bulgaria, France, Sweden, and Italy presented imbalances since the first round of monitoring, with the latter being categorized as under excessive imbalances for the last four years (see Table 1 on page 38).
The crisis of trust toward EU institutions, the rising influence of national
movements campaigning for opting-out of the euro or the whole EU, and
the still ongoing consequences of the great recession, make a valid point
in favour of a better understanding of the current shortfall of a mechanism
aimed at averting future crisis. Perfecting it would enhance Commission’s
(and in general EU’s) output legitimacy, that in turn could facilitate
fur-ther integration where needed. Moreover, given the willingness of EU’s
institutions to make more incisive use of MIP’s corrective arm and
capabil-ity to steer structural reforms (J.-C. Juncker, Tusk, Dijsselbloem, Draghi,
& Schulz, 2015), a clear understanding of the mechanisms at play is of
paramount importance.
Our research question is thus as follow:
Why are some member states capable of modifying the assessments of
Theory
Regulation No 1176/2011, as previously detailed, involves an high
de-gree of vagueness, and empowers the Commission with discretion hardly
seen in other surveillance mechanisms13 (Chalmers, 2012). Such an
out-come can be explained by the uncertainty that faced the EU during the
adoption of the Six-Pack. At the time, even the European Council
Presi-dent Herman Van Rompuy considered the developments within the EMU
as an existential threat to single currency and possibly to the existence of
the EU as a whole (Rumpoy, 2012).
This chapter will provide the main arguments discussed in the
litera-ture, from which hypotheses and the research design framework will be
built.
International contracts and MIP
The decision of member states to agree on international forms of
co-operation can be analysed as a mechanism to guarantee commitments
against short-term opposing interests - time inconsistency - or political
in-stability (Kydland & Prescott, 1977). Delegation is thus a reinforcing tool
apt to depoliticize certain fields, guaranteeing compliance through an
over-sight system that is (at least partially) away from their control, enhancing
its credibility and time consistency (Gilardi, 2002, 2005; Majone, 2001).
Such form of restraint, based on delegation to an actor insulated from
both time inconsistency and political instability, is what Lijphart (1999,
p. 5) defines as divided power. Moreover, future politicians will hardly be
13For example, the SGP, while offering leeway to Commission’s assessments (Euro-pean Commission, 2015e), has a limited amount of numerical indicators, to which non-compliance will be evident.
able to reverse decisions on delegation, guaranteeing delegation’s
credibil-ity (Wallace, Pollack, & Young, 2015, p. 58).
Most evident in the case of EU, where deepening of integration is based
on cross-border-transactions that require supranational coordination
fol-lowed by further institutionalization (Sweet & Sandholtz, 1997),
coordi-nation often requires the establishment of hard mechanisms, based on
independent assessments of compliance and possible sanctioning
pow-ers against breaching parties (Thatcher & Sweet, 2004; Pollack, 1997;
Tallberg, 2000). Majone (1994, p. 94) has argued that the Commission,
because it is pan-European and not democratically elected, is more
insu-lated from political pressures and is therefore more likely to take difficult
decisions and less likely to be captured by vested interests than national
regulators.
Generally speaking, bounded rational governments are unable to
fore-see possible future developments during international negotiations (Cooley
& Spruyt, 2009, p. 29). As a consequence, recurring to incomplete
con-tracts, where the principal (in our framework, the Council) sets the main
goal and priorities, while the agent (the Commission) is tasked with sorting
out how to achieve them, can lower transaction costs during the agreeing
phase.14 Moreover, the dynamic development of the contract guarantees
its adaptability to changing environments, facilitating the agent’s task to
fulfil its mandate. Granting the agent the right to rewrite the contract when
new informations are available avoids the need for agreement between
nu-merous principals to renegotiate the contract, provided informations are
verifiable with the counterpart - the principal in this case - (Hart & Moore,
1988)
At the same time incomplete contracts, not having detailed rules grant the
agent a variable degree of flexibility. Flexibility in international agreements
allows member states to accept deeper commitments otherwise costly, due
to the constrains that they will be putting on an uncertain future (Johns,
2014; Mansfield, Milner, & Rosendorff, 2002; Kucik & Reinhardt, 2008).
If flexibility is exploited by one of the actors under surveillance, the
commitments of all signatory parties would lose enforceability,
undermin-ing the rationale and credibility of the surveillance mechanism as a whole
(Chapman, Fang, Li, & Stone, 2015; Copelovitch, 2010). Differently from
domestic law, where violations of the contract entails zero-sum payments
from breaching to breached-against party (as it is formal non-compliance
within the MIP), breaches in the process are not zero-sum (Guzman, 2005),
meaning that in case of a member’s violation, whose certification and
sanc-tioning is blocked at a certain step of the surveillance process, the
reputa-tional loss of the breaching party is not offset by benefit to the Commission.
In the case of MIP, member states agreed upon a new mechanism that
was layered on top of earlier legislation and Treaty provisions (Verdun,
2015). According to historical institutionalism’s theory (Pierson, 2000;
Capoccia & Kelemen, 2007; Mahoney, James y Schensul, 2006; DiMaggio
& Powell, 1983, for an introduction), setting up new institutions can
eas-ily lead to the recreation of existing frameworks, given the knowledge of
their functioning, vagueness on their goal, and technical uncertainty,
of-ten leading to isomorphic institutional change. Path dependency provides
incentives to recreate institutional choices inherited from the past, even if
why the MIP was created with characteristics already criticised in the SGP.
Yet, although MIP expanded existing surveillance over further policy-fields,
its architecture incorporated the contemporary novelties reinforcing the
SGP, thus augmenting uncertainties on member states’ commitment to
the new procedure. Reinforced mechanisms made sense in the case of SGP
because they were the result of a process of learning (Dunlop & Radaelli,
2015), so to resolve the criticality exposed during previous years. Their
reproduction in a new framework of surveillance, with different goals and
a broader focus, was not assured to be useful, and uncertainty was thus
even higher than on the SGP reform.
To overcome these issues, the agent has been tasked to come up with time
line, procedural rules, indicators to be analysed, and methodologies for the
assessment (Discussed in Section ‘Reforms and Critiques’). Such reliance
on the Commission suggests that member states recognized its authority
and importance in coordinating economic policies, other than committing
themselves credibly to international agreements and to benefit from the
policy-relevant expertise provided by supranational actors (Wallace et al.,
2015, p. 32).
In order to better understand Commission’s powers, preferences, and
subsequently its relation with member states (the principal), it is useful to
dive into the role of such an institution.
European Commission: policy entrepreneur and watchdog
The building block of the Commission’s role is inscribed in the Treaty on
European Union (TEU), whose article 17 describes the two main functions
The Commission ‘promote the general interest of the Union and take
appropriate initiative to that end’. EU integration scholarship describes
such role as policy entrepreneurship (Copeland & James, 2014). It is an
agent whose interest lays in expanding its role though more delegation of
competencies. Important in such context is also the ability to take
advan-tage of windows of opportunity to propose and push in favour of its own
solution to policy-makers (Kingdon, 1984).
This behaviour by the Commission can be specifically seen in its signalling
of the shortcoming of surveillance regarding macroeconomic imbalance
when member states were in many cases still profiting from emerging
im-balances (European Commission, 2008). When - soon after - the economic
and financial crisis hit the EU, in May 2010 the Commission produced a
Communication proposing - among others - the surveillance of
macroeco-nomic developments through a new mechanism (European Commission,
2010c), it was eventually agreed by European Parliament and Council in
November 2011. This decisive result scored by the Commission was also
politically useful to demonstrate its ability to remain relevant and act
de-cisively throughout the economic crisis (Armstrong, 2012, p. 214).
Such an example, however, represents more an exception than the rule,
especially on major reforms. In fact, as pointed out by Marks and Hooghe
(2009), with the end of the ‘permissive consensus’ that characterised
Eu-ropean’s ambitions toward integration, and member states’ ideational
con-vergence toward the so called ‘Maastricht orthodoxy’ (Andor, 2013), which
implies a more pronounced turn toward intergovernmental decision-making,
Commission’s entrepreneurship is constrained by member states’
2013), having renounced aiming at an ‘ever closer Union’, as the White
pa-per on the future of Europe suggests, proposing for the first time a scenario
with EU’s retrenchment (European Commission, 2017c).
Even though Commission’s role in setting the economic agenda is
re-duced, article 17 TUE also gives it ‘coordinating, executive and
manage-ment functions’. Such tasks are of paramount importance in the
frame-work of MIP, given the nature of incomplete contract on which the
mecha-nism is established.
Due to the broad definition of imbalances, the unsettled question on when
an imbalance should be considered excessive, and the sensitivity of policy
recommendations to curb them - reforms can be required in a range of
fields encompassing labour market protection, pension reforms, taxation,
banking and housing sectors’ developments (to name a few) - the
Commis-sion has an high level of responsibility and discretion at its hands. This
is even more evident in relation to members of the euro zone, for which
non-compliance can lead to financial sanctions.
In order to ensure this task the Commission has to be perceived as
le-gitimate. Lacking input legitimacy, and being throughput legitimacy
ques-tioned due to the opaqueness of its assessments, attention to be as
trans-parent as possible, inclusive in producing assessments with all engaged
actors, and to deliver tangible and indisputable results (output legitimacy)
is among the main priorities of the Commission (Schmidt, 2013), otherwise
member states could have incentives to disregard its prescriptions, and in
the worse case to roll back the delegated powers.
Regarding throughput legitimacy, the Commission pays lot of attention to
see for example). Other than involving national actors more deeply in the
process, through periodically revamped frameworks such as the 2015 one
by the Juncker Commission, from which member states now have more
time for national discussion (Figure 1 on page 11 presents the revamped
Semester) and reinforced dialogue with the institutions (European
Com-mission, 2015b; Stuchlik, 2016), the Commission needs to establish itself
as a credible, authoritative and prepared actor. Even though an internal
reorganization took place to provide for better analytical skills (Savage &
Verdun, 2016), it still faces an asymmetry of informations against each
member state,15 hence it focuses on collaborating with them to achieve
its goals (European Commission, 2015d). An interesting insight in this
matter, exposed by Schout (2017), is that in many cases national voters
trust EU institutions more than their own government, thus enhancing
Commission’s legitimacy vis-à-vis its counterparts in the Council.
While not being the main focus of this research, it is interesting to note
that even if the Commission had the best knowledge of each member state
situation and was able to craft policies capable of solving every issue under
the MIP, as pointed out in Jones (2016): ‘then reform [would be] more a
process of politicians imposing the best solution policy advisers can offer
on powerful groups while at the same time selling the necessity for
sac-rifice to the electorate’. On a similar vein, Scharpf (2011) highlights that
under the scope of MIP recommendations governments’ capacity to
exer-15Moschella (2014) compares MIP surveillance with well established IMF’s one. She concludes that although asymmetry of information is a characteristic of each and every international surveillance body, the reiterated relations between officials from the Com-mission and member states provide a better solution to this problem than the IMF instru-ments. The MIP offers the tools to produce tailored recommendations, instead of generic ones normally produced by the IMF’s staff.
cise indirect influences over variables such as nominal wages or private
savings vary greatly between countries (in some cases being deemed ‘non
existing’), making compliance potentially impossible, or politically
suici-dal.
As a consequence, while the Commission has to aim at the highest
an-alytical level possible, contestation by member states is inherent in the
process, due to both the residual uncertainty of the proposed measures,
and the political implication such recommendations have over national
decision-making.
Even before the reform of economic governance in the EU, the power
of the Commission appeared to be ample and uncomfortable for member
states. Already within the Lisbon Strategy, where the Open Method of
Coordination (OMC) guaranteed only soft surveillance powers to the
Com-mission, according to Zeitlin and Heidenreich (2009, p. 236) it was
com-mon that bilateral meeting between Commission and member state took
the form of a negotiation over the modification or withdrawal of an
‘embar-rassing’ recommendation.
If such a situation occurs within the MIP framework a few considerations
are in order. First, Commission’s power can be counterbalanced by the
actions of some member states, especially if in the Council they are able to
create a coalition to suppress or water down certain inconvenient
recom-mendations. Second, we can deduce that not all member state have equal
capabilities and power to act that way, meaning that Commission’s
discre-tionary power would end up being stronger with weaker member states,
and weak with stronger ones. Finally, if some member states manage to
other member states would have an incentive to build coalitions to obtain
the same treatment, being a qualified majority needed to approve changes.
Member states’ preferences determinants
While assumed as strategic actors maximising their utility, member
states’ rationality is constrained by a number of factors over which they
have little control. this is most evident analysing their behaviour over time
(Moravcsik, 1993; Kydland & Prescott, 1977).
Member states do not have fixed preferences, meaning that even parties
favouring integration in specific domains might in the future oppose the
very contract they pushed for in the first place, due to changing (economic)
environment, unforeseen developments (both internally and
internation-ally), ideational change in their elite or electorate, or strategic opposition
for bargaining/electoral reasons (Börzel, 2002; Börzel & Risse, 2000).
An example of it can be the lobbying from Germany and France for the
cre-ation of the SGP, their breaching the pact soon after, subsequent reform on
a more flexible way, followed by Germany’s opposition against SGP’s
flexi-bility in the midst of the crisis (Heipertz & Verdun, 2010; Valero, 2015).
Respecting international agreements can be a sword of Damocles for
member states: the two level game existing in the EU (Putnam, 1988) can
be exploited to retain consensus while taking unpopular measures,16 or it
can encourage internal opposition to government incapable of standing up
against EU institutions leading to government change.17
16As in the recently analysed case of Latvia (Eihmanis, 2017), where national govern-ment implegovern-mented stricter austerity measures than the ones asked by the Commission and yet was able to be re-elected.
17Walter (2015) models why some countries implemented austerity measures quite smoothly while others faced problem. In the first two years of the economic and financial crisis in Europe Bosco and Verney (2012) defined electoral results as ‘incumbent epidemic’.
Drawing from multilateral lending and conditionality theories,
histori-cally developed for understanding behaviour inside International Monetary
Fund (IMF) and World Bank (WB), a powerful state is an influential one in
decision-making and rule bending (Thacker, 1999). Moreover, a
power-ful state is capable of shaping the size of conditionality and the strings
attached to financial lending (Copelovitch, 2010). We assume that
simi-lar mechanisms will be generally in place even inside EMU’s architecture.
Member states thus are not created equals, and their power determines
their ability to oppose Commission’s actions throughout the process of
oversight, instead of just fighting back once a problem is brought up.
In the EU, power can be categorized according to the voting rights of each
member states, or considering which tends to be the main actors around
whose positions coalitions are built (Kauppi & Widgrén, 2004; J. E. Lane
& Mæland, 2000; Aksoy, 2010), being the two often overlapping.
Different factors appears to impact on the influence of member states
in international negotiations.
Public support for instance can both favour or hinder member state power.
As researched by Chaudoin (2014b) regarding international trade
agree-ments, voters favour honouring the commitments only if they have no
preference over the issue. This can also be reversed as theorising thus
that where public opinion support European institutions, member state’s
power vis-à-vis the Commission will decrees. Moreover, in another paper
Chaudoin (2014a) found that voters are sensitive toward respect or
disre-spect for international agreements mainly during election periods.
Consequently, elections are sensitive periods for member states (Hibbs,
1991) to avoid antagonizing particular and powerful interests groups, other
than temporarily approve some policies to increment voters’ support
(Fernández-Villaverde, Garicano, & Santos, 2013; Efthyvoulou, 2012).
The outcome of elections can also influence member state power in
differ-ent ways: weak governmdiffer-ents have lower negotiating power internationally
due to the constraints under which each minister is kept accountable for
in national parliaments (Martin & Vanberg, 2004; Pahre, 1997; Goetz &
Meyer-Sahling, 2009). Moving toward the left side of the political
spec-trum for a government tends to mean that complying with financial
com-mitments will be harder (Grieco, Gelpi, & Warren, 2009). Moreover, in
the Council, the partisanship of a minister could have an impact on the
outcome notwithstanding the party affiliation of the prime minister: the
opaqueness of voting in the Council may provide incentives for a minister
to promote his agenda instead of the government agreed upon one.
Logic of action in the Council
Given the way surveillance is conceived in the MIP - if automatism was
in place - member states at risk would be pinpointed by the AMR and
sub-sequent IDRs would expose critical policy fields to be addressed through
CSRs. Targeted countries then would take actions to correct imbalances in
order to avoid be put under the corrective arm of MIP, where enforceability
will be stricter and non-compliance punished with financial sanctions.
However Commission’s discretion plays a role in assessing member states,
hinting that similar situations could be treated differently for reasons still
uncovered in the literature. Moreover, member states still retain power in
have in incentive to ‘correct’ Commission’s prescriptions before they are
endorsed. This move would simplify their task internally and avoid giving
ground to the Commission to step up its monitoring powers.
Member states’ behaviour is thus an important, yet not systematically
analysed, aspect of multilateral macroeconomic surveillance.
Bargaining strategies are particularly interesting in case of member
states facing imbalances that join forces to collectively suppress unfavourable
assessment by the Commission. Literature here affirms that lax
enforce-ment of international rules leads to bad behaviour by other member states
(Finance, 2004).
Hypotheses
The incomplete nature of the contract regulating the MIP is the
subopti-mal result of member state unwillingness to delegate further competencies
to an independent body that could force them to adopt policies that might
undermine national interests and traditional models of policy (Hall, 2014;
Hall & Soskice, 2001), forcing institutional convergence as a way to reach
the economic one.
On the other hand, coordination was deemed necessary due to the
inter-dependence of member states’ economies, as demonstrated by the harm
caused by rising imbalances and diverging trends prior to the economic
and financial crisis. The Commission thus was enabled with oversight and
steering competences over member states economic policies, empowered
by a mechanism of enforceability that in the case of euro zone members
can lead to financial sanctions, underpinning its hard power.
ca-pable of watering down Commission’s recommendations, and what
deter-mines these outcomes.
We can then hypothesise member states to exploit their own power to avoid
harsh assessment, which could lead to uncomfortable recommendations.
Recommendations that might on the one hand impact national
govern-ments’ re-election chances in case of implementation, and their loss of
credibility in the face of other member states and financial markets on the
other (notwithstanding be fact that they could face sanctions if members
of the euro zone).
H1.1 Member states’ power determines the final outcome of
Commis-sion’s assessments, followed by
H1.2 Bigger member states have more success in modifying
recommen-dations.
Given the nature of enhanced surveillance in the corrective arm, we can
theorise that
H1.3 Member of the euro zone - ceteris paribus - will be more invested
in watering down Commission’s assessment than non euro zone member
states.
Given the two-level political game in the EU we can expect that:
H2.1 The higher Euroscepticism, the higher the chances for member states
of modifying Commission’s assessments
That can also be reversed into:
H2.2 The higher support for the Commission, the lower the chances for
member states of modifying Commission’s assessments.
Considering the internal dimension of member states part of international
sit-ting governments regarding their strategic behaviour, thus we expect:
H3.1 In the run-up to elections, member states will increase changes to
Commission’s assessment due to more attention toward their achievements
and failures. Moreover:
H3.2 After elections resistance by member states against Commission
in-terference will increase , so to establish themselves as authoritative actors
in EU’s negotiation at the eyes of their voters.
H3.3 Fragmented governments, whose cabinet positions are split
be-tween competing parties, will have more chances of modifying Commission’s
assessments, due to the higher criticism and difficulty of implementing
sug-gested policies they will face internally
H4.1 The more worrisome the imbalance (and the higher the number and
precision of prescriptions to comply with), the more the chances for member
states of modifying Commission’s assessments.
H4.2 Non-compliance with the other pillar of surveillance - the SGP - will
increase the chances for member states of modifying Commission’s
assess-ments. So to avoid having a too large set of recommendations and a way
too harsh assessment to be presented to national parliaments.
H4.3 Member states experiencing a recession will have more chances of
modifying Commission’s assessments. This is due to the fact that they’ll
receive more recommendations and national voters will be already
com-plaining about incumbent government, hence the need to water down bad
Research Design
In order to test the aforementioned hypotheses, both qualitative and
quantitative designs could be employed.
A small-n in-depth comparative case study, assessing a few member states
over specific cycles of the European Semester, could offer a detailed
ac-count of the logic underlying the outcome we observe in these cases,
pro-viding a thorough explanation for formal rules and informal behaviours
of the actors involved. Still, deeply interpreting only a few cases makes
generalization difficult, and more often that not, entails using too many
variables to explain an outcome (Schneider, 2008, p. 278). Moreover, the
connection between the specific cases and existing theories needs to be
really strong (Toshkov, 2016, p. 285), and external validity is not ensured
(Toshkov, 2016, p. 304). If we wish to analyse a weak or rare effect, we will
then most surely conduct research on deviant cases, where such an effect
would be most evident, for which very specific conditions could mislead
us to consider the causal mechanism that explains an outlier a common
feature of all cases in which such effect occurs. The abundance of
vari-ables makes even harder weighting their impact, discouraging
generaliza-tion even more.
Given the exploratory nature of this thesis, on a path - to our knowledge
- still untouched, uncovering general trends and possible causal
mecha-nisms would pave the way for more specific analyses, establishing a new
venue of research. Thus, hoping to contribute to member states’ and
Com-mission’s bargaining power literature, we rely on a large-n design, due to
relationship (Toshkov, 2016, p. 200).
One of the main shortcomings of applying quantitative approaches to
decision-making in the Council and its relations with the Commission is that only
revealed behaviours will emerge (Heisenberg, 2008, p. 270), whereas
strate-gic behaviours and informal norms are a strong component of Council
decision-making (Puetter, 2012).
Notwithstanding such pitfall, already providing a general set of variables
that influence the outcome of bargaining in the MIP framework would ease
the process of uncovering more subtle and informal independent variables
in single case or small-n comparative studies.
Finally, since the MIP is a new procedure, we are interested in testing
established theories of international political economy to it, so to uncover
similarities and differences due to specific governance arrangements. We
thus follow a deductive structure of theory testing: abstract propositions
are tested empirically and can either be confirmed or disproved (Toshkov,
2016, p. 39).
We will try to confirm our four hypotheses. However, due to the reduced
time-span of the analysis, the presence of an economic and political shock
at the beginning of our analysis that could bias behaviours, and the
pres-ence of unobserved variables that could shape outcome, the prospect of
theory confirmation is reduced. Still, following Popper’s falsifiability
crite-rion, our goal in pursuing scientific discovery should be falsification
Member states’ behaviour
Since the establishment of the Macroeconomic Imbalance Procedure
a total of 296 MIP-related country specific recommendations have been
adopted by the Council throughout six European Semester’s cycles.
First of all, using text-analysis techniques, we will conduct an analysis of
both editing and weakening of the texts assessing National Reform
Pro-grams, at whose end CSRs are detailed, considering whether the wording
or the meaning was changed from the Commission recommendation to the
final one adopted by the Council. Moreover, in case some prescriptions
were deleted through the process, it will be noted. The creation of such
dataset would also provide new information for Council’s behaviour,
be-ing useful for different research questions, thus possibly contributbe-ing at
expanding our knowledge on decision-making in the reformed economic
governance framework.
Such an assessment will provide a first count of whether member states
are actually capable of watering down Commission’s suggestions, with
what frequency they are able to do it, and whether the successful text
modification is equally distributed among member states.
On a second moment we will then test if these results correlate with
different set of variables, relating to Member state power, their economic
situation, the composition of national governments and the proximity of
elections, the attitudes of national voters toward the EU, or euro
mem-bership. In case of positive results we would then try to better uncover
plausible causal mechanisms. Obtaining negative results will still provide
(Bauer & Becker, 2014) and opening a venue of research on how the
Com-mission is capable to constrain member states’ prerogatives.
Another interesting venue of research, partially explored by Gros and
Alcidi (2015), relates to implementation of CSRs. It is considered low by
different accounts - similar at the rate of implementation of OECD’s ‘Going
for Growth’ recommendations. What seems overlooked is the relation
be-tween implementation, results of subsequently monitored indicators, and
following further recommendations. Process tracing analysis regarding the
Netherlands and MIP18 suggests that member states can forcefully reject
Commission’s intrusion in national politics, making the institution
with-draw non complied recommendation, although no change in
macroeco-nomic indicators would suggest an improvement of the national situation.
Thus, combining an analysis on non-implemented recommendations and
later developments in the scoreboards could uncover other cases in which
such phenomenon took place, offering a precise target for further in-depth
analyses. Moreover, comparing analyses of weakening with the follow-up
National Reform Programme could help understand how member states
exploits these changes in dealing with future cycles of macroeconomic
surveillance. For reasons of time constraints and the effort required such
cases - if uncovered - will not be analysed in this paper. This research
still contributes to the scarce strand of literature targeting the MIP, thus
proposing a frame in which the proposed and other researches could be
conducted.
18An unpublished manuscript by Bokhorst (2016) presents the case of the Netherlands and its mortgage interest tax deductibility (MID), addressed by the Commission and re-moved by the Dutch delegation after having demonstrated that no political party would have tackled it.
Variables and Operationalization
Population and case selections.
Table 1
Macroeconomic Imbalance Procedure findings for EU member states
Year no In-depth review no Imbalance Imbalance Excessive EIP Program Country 2012 AT, CZ, EE, DE, LT, LV, LU, MT, NL, PO, SK BE, BG, CY, DK, FI, FR, HU, IT, SI, ES, SE, UK EL, IE, PT, RO 2013 AT, CZ, EE, DE, LT, LV, LU, PO, SK BE, BG, DK, FI, FR, HU, IT, MT, NL, SE, UK SI, ES CY, EL, IE, PT, RO 2014 AT, CZ, EE, LT, LV, PO, SK DK, LU, MT BE, BG, FI, FR, DE, HU, IE, NL, ES, SE, UK HR, IT, SI CY, EL, PT, RO 2015 AT, CZ, DK, EE, LT, LV, LU, MT, PO, SK BE, FI, DE, HU, IE, NL, RO, SI, ES, SE, UK BG, HR, FR, IT, PT CY, EL 2016 CZ, DK, LT, LV, LU, MT, PO, SK AT, BE, EE, HU, RO, UK FI, DE, IE, NL, SI, ES, SE BG, HR, CY, FR, IT, PT EL 2017 AT, BE, CZ, DK, EE, HU, LT, LV, LU, MT, PO, RO, SK, UK FI DE, IE, NL, SI, ES, SE BG, HR, CY, FR, IT, PT EL
Sources: European Central Bank (2012), European Commission (2013b, 2014, 2015a, 2016a, 2017a)
Since the establishment of MIP, six rounds of surveillance were
con-cluded at the time of writing (winter 2017). Our unit of analysis is
country-by-year. Given a population of currently 28 member states, minus
coun-tries that joined the EU after the establishment of the procedure, and
member states under specific programs of surveillance for which the MIP
does not apply, we obtain a population of n=149 cases. For details see
Table 1. Data availability reduces the sample to n=148: recommendations
for Estonia in 2013 are unfortunately not available in English from any
official source.19
Dependent variables.
The main dependent variable is member states’ success in watering
down Commission’s assessment. To capture this effect we consider the
level of variation between Commission’s assessments of National Reform
Programs and Council’s approved country specific recommendations.
19All documents were automatically retrieved from: http://eur-lex.europa.eu/homepage.html
Documents regarding Denmark (2017), Czech Republic (2017), and Slovenia (2015) were retrieved from the website of European Commission or Council. Documents regarding Estonia (2013) are not available at all.
Such variation is represented by two different variables, defined edit and
weakening, following the example of Baerg and Hallerberg (2016). They are
both coded by the author using two different methods: edits is the result
of automated text analysis, while weakening takes advantage of the
swift-ness of semi-automated data gathering to perform human coded textual
comparison.
Figure 2. Longitudinal data on the dependent variable edit
(a) Average edit by year (b) Average edit by Country
Edit is a count variable that represents the number of changes that
oc-curred between the two versions of the document (i.e. the Commission’s
draft and Council’s approved one). It is produced through an automated
procedure based on a script compiled using the coding language Python.
After having compared the two versions of the same text, we calculated
the Levenshtein (1966) scale of distance,20 a measure of the required
dele-tions, insertions or substitutions needed to match the two texts (Sörensen,
2007). Once we have the raw data, we can check the distribution of
changes across countries and across year. As Figure 2 show, there is
in-20Distance can be calculated at different levels. Micro changes would consider a letter as the unit of analysis, thus accounting for all textual modifications. Meso changes would focus on words, whereas macro changes focus on sentences as unit of analysis. We opted for micro analysis, in order to have the highest level of information.
deed quite some level of variation, suggesting that certain circumstances
favour or hinder the chances of editing Commission’s text.
The main problem with the variable edit is that it does not measure
content similarity. Two versions of the texts could have a Levenshtein
dis-tance of zero, while having different meanings due to textual substitutions.
To overcome this effect of automated text analysis, we produced a second
dependent variable, called weakening. Using human coding, we assessed
the changes and their meaning between the two versions of the texts,
cod-ing a binary outcome for weakencod-ing (1, effect is present), and no change
(0, effect not present). In order to be considered an instance of weakening,
the Council text needs to be less critical toward a member state, or have
prescribed less stringent - or precise - outcomes. Few examples of
weaken-ing are presented for clarification in Table 2 in 42. There is no distinction
between the level magnitude of watering-down: as soon as weakening is
found the case is coded as such, without accounting for multiple
weaken-ing in the same text or different degrees of changes. Results are presented
in Figure 3.
Figure 3. Longitudinal data on the dependent variable weakening
Analysing both variables offers the opportunity to check whether
mem-ber states change texts with the only result of weakening the
recommenda-tions received, or whether also minor changes are important to them, due
for instance to the follow-up discussions inside national parliaments, or
the possibility for journalists to extrapolate part of the recommendations
to comment on the government’s results or unsuccessfulness. Data
pre-sented in Figures 5 on page 62 show that the instances of weakening are
quite scarce between countries and across years, while editing is a more
widespread practice.
As observed by Grimmer and Stewart (2013, p. 267) automated
meth-ods are no substitute for thinking about the problem and validating
gath-ered informations, thus the weakening variable provides a qualitative
as-sessment based on human judgement and benchmarking. That way,
al-though the two variables assess different outcomes, we are capable of
test related effect using different methodologies, reducing possible biases
G PREFEREN CES BET WEEN INS TITUTIONS 42 Table 2
Example of coding weakening variable
Country and year Commission assessment Council assessment
Netherlands 2014 Even though the 2013 recommendations called for
a focus on support for households most in need, this redirection has not taken place. Therefore, while the proposed measures are steps in the right direction, the overall pace of reforms is slow in ad-dressing underlying problems and thus needs to be stepped up, while continuing to ensure that so-cial housing is available to disadvantaged citizens unable to obtain housing at market conditions, in-cluding in high demand locations. The 2013 rec-ommendations suggested that the stepping-up of the housing market measures should be subject to overall economic developments. As the economy and the housing market are expected to continue recovering, consideration should be given to speed-ing up the pace of reforms.
The 2013 recommendations called for a focus on support for households most in need, this redirec-tion is work in progress. Therefore, while the pro-posed measures are steps in the right direction, the overall pace of reforms is slow in addressing un-derlying problems and thus needs to be stepped up when the economy allows for it, while continuing to ensure that social housing is available to disadvan-taged citizens unable to obtain housing at market conditions, including in high demand locations.
Germany 2016 Achieve a sustained upward trend in public
invest-ment, especially in infrastructure, education, re-search and innovation, by using the available fis-cal space and prioritising expenditure. Improve the design of federal fiscal relations with a view to ad-dressing the persistent public under-investment, especially at municipal level.
Achieve a sustained upward trend in public in-vestment, especially in infrastructure, education, research and innovation, while respecting the medium term objective. Improve the design of fed-eral fiscal relations, with a view to increasing public investment.
Cyprus 2017 Cyprus also continues to face critical challenges for
the functioning of the judicial system.
Cyprus has been taking measures to reinforce its judicial system but continues to face serious chal-lenges as regards the efficiency thereof.
Explanatory variables.
In order to test the hypotheses outlined in the previous chapter, a set of
independent variables deduced from the theory will be employed to
evalu-ate possible con-causal relations, other than providing answers expanding
our understanding of what factors impact MIP’s decision making.
Explanatory variables used in this research can be divided into four
cat-egories: institutional, European, political, and macroeconomic predictors.
They are as follow:
Institutional variables are included in matrix I={euro, large state, shared
behaviour}
Euro: Takes the form of a dummy variable describing whether the country
is part of the euro zone (1) or outside (0), so to assess if ceteris paribus
EZ member states are more invested in avoiding unfavourable
recom-mendations that, if disregarded, might lead to the opening of an EIP,
where RQVM would make it harder to avoid financial sanctions in
case of repeated non-compliance.
Large State: Takes the form of a dummy variable with the value (1) if
the country is among the four biggest EU member states (France,
Germany, Italy, or the United Kingdom).
Shared Behaviour: Takes the form of a spline smoother of the
year-by-year trend of both measures of the dependent variables, as discussed
in Baerg and Hallerberg (2016), to hint the likelihood of changing
textual outputs when other member states are doing it. A positive
coefficient indicate a positive correlation between changing the text
when others are doing it, a negative coefficient indicates no common
