http://eiop.or.at/eiop/texte/2004-008a.htm
: 6.7.2004| This paper's comments page | Send your comment! to this paper |
| Chasing Maastricht: The Impact of the EMU on the Fiscal Performance of Member States |
| Marius R. Busemeyer |
| European Integration online Papers (EIoP) Vol. 8
(2004) N° 8; http://eiop.or.at/eiop/texte/2004-008a.htm |
| Date of Publication in
: 6.7.2004 |
| | Abstract |
Back to homepage |
PDF | | This paper's comments page | Send your comment! to this paper | |
1
This paper examines the impact of the Economic and Monetary Union (EMU) on the fiscal performance of EU member states. Fiscal performance is hereby defined as the degree to which a country achieves certain normative goals in fiscal policy: Lower deficits, higher surplusses, lower debt levels, or a change rate towards lower debt levels improve fiscal performance; higher deficits, higher debt levels, or change rates towards higher debt levels are signs of deterioration in fiscal performance. In this paper, I will show that the process towards EMU and the accompanying Stability and Growth Pact (SGP) have increased fiscal performance in the EU member states. Small EU states have generally performed better than big states, because they were more constrained by the institutions of the SGP: During the period of accession, the threat of being excluded from the final stage of EMU was more severe than for the large states in the core of the EU (Germany and France). Second, as members of the EMU smaller states will encounter higher difficulties in finding allies to avoid punishment through sanctions in the Council.
The effectiveness of EMU in raising the sustainability of fiscal policies in member states is highly debated. Many critics consider the EMU and the SGP, which is supposed to guarantee a better fiscal performance of the member states, to be unsustainable in itself (McKay 1999; Ferguson / Kotlikoff 2000). Bernaldo de Quiros (1999) speaks of a “political time bomb” that will explode when strategic debt accumulation of member states (Beetsma / Bovenberg 2002) forces the European central bank to loosen its strict monetary regime. This would then impose negative externalities in the form of higher interest rates and lower private investment on the other member states. Although some argue (Buti / Giudice 2002; Rotte 1998) that the race for membership during the 1990s has indeed had a substantial effect on the member states’ fiscal performance, this effect is not expected to last as the question of membership has been settled and the remaining sanctioning mechanisms rely on the mutual condemnation of the states (Freitag / Sciarini 2001). But when all the states face the same fiscal pressures from demographic changes and slowing economic growth, this mutual condemnation is less probable.(1)
2
This paper challenges the far-reaching skepticism regarding the problem-solving capacity of the EMU. The next section shows with the help of a pooled time series analysis, that there has been a positive effect on fiscal performance, even if one controls for many other variables that have been found to have an influence on public deficits. In addition, I look at the changes in debt levels. The third section of this paper compares differences in means between EU and non-EU countries as well as between EMU and non-EMU countries for the time period between 1980 and 2002. This analyis confirms the findings of the previous section. In the fourth section, data on differences in means between small and big EU countries is presented. This sections also deals with possible explanations for these findings. The final section concludes and comments on the latest developments in the struggle around the SGP.
This chapter shows in a cross-sectional time series analysis the impact of the process of monetary and economic integration. The analysis looks at 22 OECD countries(2) in the period between 1980 and 2000. The hypothesis to be tested is that membership or prospective membership in the EMU had a significant and positive impact on fiscal performance, i.e. led to lower deficits or higher surplusses. I also control for other variables that have been deemed important for the analysis of public debts.
The dependent variable is the current budget balance (current receipts minus current disbursements) in % of GDP, taken from the OECD Economic Outlook Database. Disbursements include interest payments, so my dependent variable is not the primary balance. Discussion in the literature is split about which dependent variable to use: Some argue in favor of using the primary balance, because it is deemed to be a better indicator of political activity (Freitag / Sciarini 2001: 166). However, an incumbent government has no choice but to accept the inherited debt of its predecessor. Interest payments on inherited debt can destroy the outside appearance of fiscal prudence, although the government may be running high primary surplusses. Still, accession to the EMU was not made dependent on primary balances, but on total balances. Governments that had inherited a high public debt were in a worse starting position, because interest payments tilted the budgets downwards. The decisive variable however remained the overall budget deficit. Therefore, I chose it as dependent variable.
Another important strand of the literature argues in favor of using year-to-year changes in the public debt to GDP ratio as dependent variable (De Haan / Sturm 1994: 162; Roubini / Sachs 1989b: 904-905). Looking at changes in the debt level means adopting a strong long-term point of view. It is imaginable, that a government runs high surplusses without reducing the level of public debt, because public money is re-invested. At least theoretically, under special circumstances, like the process towards EMU, high deficits can also go along with reductions in debt levels. Therefore, I will also use changes in gross debt levels as dependent variables to test the robustness of the findings.
The choice of independent variables reflects the different explanatory approaches of comparative political economy:
3
First, there are of course economic control variables like GDP growth, inflation (CPI), and unemployment rates, taken from the OECD Economic Outlook Database as well as from OECD Historical Statistics. In the context of public debts, there are good theoretical reasons for economic factors to have a significant impact on budget deficits. High economic growth, c.p., creates more tax revenue and thus reduces the deficit. On the other hand, high unemployment leads to higher spending (in the case of public unemployment insurance, in some countries also for active labour market policies) and to lower revenue (less contributions from employees and employers) at the same time. Thus, economic growth is hypothesized to have a positive impact on budget balances, while high unemployment is expected to have a negative impact. The overall impact of economic factors is expected to be very significant. Budgets contain a number of entitlement programs with fixed benefits, contributions and eligibility criteria (fixed at least in the short term), so that changes in the economic environment directly lead to changes in revenue and spending. But the story does not end here. Political, institutional and partisan factors can be expected to increase the overall explanatory power of the argument considerably. What is more, when economic factors are given the same status as independent variables as political variables, there is a certain risk to underestimate the impact of the political factors, because economical factors are assumed to be completely exogenous, while in reality, they also depend to a certain extent on political and institutional factors (Franzese 2002: 167).
Second, the influence of the governing parties will be assessed both directly and in conjunction with other factors (unemployment rates). Early studies (Hibbs 1977) found that leftist parties in government tend to reduce unemployment even if this leads to higher inflation, while rightist party governments pay more attention to low inflation rates with less regard to unemployment rates. It has been conventional wisdom that left governments tend to spend more than right governments (Cameron 1985). But from this, it does not necessarily follow that left governments run higher deficits. Instead, a number of studies (Cameron 1985; Cusack 1999; Wagschal 1996; Franzese 2002) found that leftist governments run lower deficits. Right parties in government often lower taxes without reducing spending, which creates higher deficits, whereas left governments seem to finance their expansionary spending programs by creating higher revenues (Wagschal 1996: 183). Carlsen (1997) adds one caveat: Leftist governments, at least for the period of the 80s, tend to run higher deficits when high unemployment prevails. They seem to be more sensitive towards deteriorating economic conditions, mainly because high unemployment strongly affects their main constituencies (employees). In this analysis, my independent variables are the share of seats in the cabinet taken by left, right, and christian democratic parties. I also add a variable that measures the combined impact of the power of socialdemocratic parties in government and unemployment rates(3).
Third, and most importantly, I control for institutional variables. I will focus on government institutions and their impact on fiscal performance, because these variables have been most prominent in comparative political economy. It can be argued that these institutions have a more direct impact on performance than macro-level political institutions like corporatism or the constitutional veto-structure of a given country, although the latter are not without explanatory power (see Wagschal 1996, who finds a positive impact of corporatism on fiscal performance). Looking at government institutions, the weakness of an incumbent government has been found to lead to higher deficits / lower surplusses (Roubini / Sachs 1989 a,b; Borrelli / Royed 1995; Edin / Ohlsson 1991; Grilli / Masciandaro / Tabellini 1991; Hallerberg / von Hagen 1999). Others (Sakamoto 2001; Wagschal 1996) have challenged this hypothesis. The theoretical reasoning behind it points to three factors: the number of coalition parties, the type of government (majority of minority government) and the average tenure of governments.
4
First, the higher the number of coalition parties, the weaker the government and the worse the fiscal performance. Each governing party has its own constituencies that have to be served (Sakamoto 2001: 533). In a climate of austerity, when the distribution of losses happens more often than the distribution of gains, each party is a veto player in the decision-making arena of the government. Above all in mininum winning coalitions, each player has the institutional power to block or water down spending cuts by threatening to leave the coalition to protect his constituencies in order to increase his chances of re-election. A similar logic applies to the variable “number of ministers in a cabinet”. The more spending ministers there are, the higher the deficits (Kontopoulos / Perotti 1999, who find that the number of spending ministers is more important than the number of parties in a coalition). This is somewhat surprising, because the linkage between a single minister (as member of a party) and a certain constituency is not as strong and direct as between a party and its electoral base. But each minister heads a department of the government, which by itself can be a powerful bureaucracy with strong interests in power expansion. Pressurized by his departmental administration and led by his own interests in gaining more power vis-á-vis his ministerial colleagues, each minister has an incentive to increase the size of his budget and get involved in log-rolling to help others expanding their budget. Only the prime minister and the finance minister do not have this incentive (for further elaboration on this point see below).
The second factor is the type of government. Roubini and Sachs construct an index to measure to degree of political coherence of the government (Roubini / Sachs 1989 a,b). This index unfortunately mixes the distinctive dimensions of number of parties in the coalition and type of government. This has been rightly critized (Sakamoto 2001: 532; Edin / Ohlsson 1991; Borelli / Royed 1995),(4) because it unnecessarily gives away opportunities to increase information and explanatory power. The critique focusses on the treatment of minority governments by Roubini and Sachs, which are ascribed to the highest degree of weakness, because they are expected to be bound by fluctuating and unstable coalitions in parliament. But there is no theoretical reason to assume that minority governments are weaker than multi-party majority governments per se. Not only do majority governments face the risk of being held accountable for budget decisions to a greater extent (Sakamoto 2001: 532), minority governments can also sometimes take advantage of their minority position by building different coalitions for different policy projects. They can also threaten potential coalition partners to form a policy coalition with other parties. Thus, the dimensions of size and type of government have to be kept seperately. Therefore, I will not use the Roubini and Sachs index, but instead control for the individual variables.
The third factor is government tenure or government turnover. If a government anticipates remaining in office only for a short period of time, then the temptation to increase deficits becomes stronger as the government can expect not to be held accountable for the resulting deficit problems. Thus, the hypothesis is: The shorter the average government tenure or the higher the government turnover, the higher the deficits (for others Grilli / Masciandaro / Tabellini 1991: 358).
Hallerberg and von Hagen (1999) try to connect the weak / deficit-prone government hypothesis to the structure of the party and electoral system. They argue that in proportional voting systems the number of parties in parliament is higher, and therefore, the necessity to build multiparty government coalitions instead of single-party governments is, too. In majority systems, the chances of obtaining single-party governments are considered to be much higher. The consequences for fiscal policy, besides the argument that many parties in the coalition tend to produce higher deficits, are in their view that single party governments tend to overcome the collective action problem of how to keep the spending ministers at bay by delegating a lot of power to a strong finance minister.
5
The incentives of a finance minister are expected to be different from ‘conventional’ spending ministers, because the finance minister gains prestige and power not by expanding his budget, but by controlling the other spending ministers and circumscribing their spending phantasies (see also Alesina / Perotti 1999). To extensively test the Hallerberg / von Hagen thesis, one would have to test their inferences from the party system and the electoral systems to the strength of government and the position of the minister of finance first before then testing the impact on fiscal performance.(5) In fact, the Hallerberg / von Hagen-thesis adds just another element to the causal chain: They do not question the weak / deficit-prone government hypothesis, instead they try to link the existence of weak governments to institutional features of the electoral and the party system. Because the expected direct effect of these factors is probably weak, I will not include them in the empirical analyses later on.
To operationalize the institutional hypotheses, I use variables from different sources. The number of cabinet members (AKAB) was taken from Schmidt (2003). This variable substitutes the ‘number of spending ministers’ variable, which has been tested by Kontopoulos / Perotti 1999. This is possible because it can be assumed that the number of ‘non-spending’ ministers (normally only the finance minister) is more or less constant between countries. What is more, the distinction between spending and non-spending ministers introduces an element of arbitrariness: Questions about the status of the head of government arise: Does he really belong to the ‘non-spending’ fraction or does he have a strong incentive to avoid serious spending cuts, because he will be the one to be punished most severely in the next election? Thus, it seems reasonable to use the variable ‘number of cabinet members’.
Another variable reflects the majority that a government coalition has in parliament (MAJ) on a yearly basis. This variable is taken from the Database of Political Institutions (DPI) (Beck et al. 2000) to operationalize the type of government. It is a continuous variable and therefore preferable to a simple dummy variable on the minority / majority status of a government. Two other variables from the DPI are used to operationalize government turnover: One measures the number of years that the chief executive has remained in office (YRSOFFC) and the other the number of years that the party of the chief executive has remained in charge of governing the country (PRTYIN).
It is interesting to look at macro-level institutional variables. The independence of the national central bank is an important example. Governments that are confronted with a highly independent central bank will not run high deficits, because they cannot artificially reduce debt levels with the help of inflation (Franzese 2002: 146). They can neither rely on the central bank to bail them out if debt levels have risen too much (the EC treaty f.e. explicitly rules out bail-outs for national governments by the ECB). And third, they cannot force the central bank to lower interest rates if these have risen because of massive public borrowing and resulting crowding-out effects. Therefore, independent central banks improve the fiscal performance of states.
I include a second macro-level institutional variable: the number of vetoplayers. Some studies (Obinger / Wagschal 2000) found that many vetoplayers have been a reason for retarded welfare state development in some states (Switzerland, USA), because powerful “countermajoritarian institutions” (Schmidt 2000: 351-354) have only reluctantly transferred spending power to the central government. Extending this argument, one could expect better fiscal performance for countries with a strong constitutional veto structure, because spending pressures are reduced. I expect a different outcome, because the fiscal climate has changed “from expansion to austerity” (Pierson 2001). In times of austerity, when the reduction of deficits becomes an important policy goal in order to create more fiscal leeway for other policy decisions, a strong constitutional veto structure can hinder consolidation initiatives from the central government.
6
For example, a strong second parliamentary chamber can increase the chances of the opposition to exert influence, especially when the opposition controls a majority of seats in the second chamber (f.e. Germany for the most part of the 1990s until today). Party competition promotes the adoption of a contra-consolidation policy strategy by the opposition in order to increase their chances of government take-over in the next elections. Another example are public social insurance funds: when these are not part of the general government budget but instead self-administrated and financed by contributions instead of general tax revenue, this creates another strong institutional barrier against quick consolidation. These examples show intuitively that a strong constitutional veto structure can be expected to increase deficits. To measure the constitutional veto structure, I use Manfred Schmidt’s (2000: 352-353) index of veto points (‘VETO’).(6)
Finally, two subsidiary variables have to be added. One is the lagged current account deficit (the lagged dependent variable). Although this variable has to be added because of methodical reasons, there is also a theoretical consideration behind it, namely policy inheritance: High budget deficits are hard to dismantle. Consolidation policies go together with high political costs, because they are often linked to spending cuts or tax hikes. Especially, when deficits are high, the policy inheritance effect will be very strong. When countries run budget surplusses (which is rather seldom), the policy inheritance effects will be not as strong: Money is easier spent than saved.
The second subsidiary variable is trade openness (measured by the sum of imports and exports as percent of GDP, ‘OPENNESS’). The ongoing economic internationalization since the 1980s has increased the pressure on governments to avoid unsustainable spending practices. Yet, it is still highly debated if increasing openness leads to higher compensatory spending and larger deficits (Sakamoto 2001: 539) or to lower deficits because of considerations of competitiveness (Wagschal 1996: 146).
It is interesting to assess the impact of EMU vis-á-vis the impact of the general trend towards economic internationalization. The question is if EMU had an independent impact or if the variations in the fiscal performance can be explained by the variations in the openness of a country towards forces of economic internationalization.
For EMU membership, I constructed two dummy variables. One (“EU member after 1992”) gives each EU member from 1992 on the value of 1 (Austria, Finland and Sweden from 1995). This dummy is supposed to capture the effects of the Maastricht process for all EU member countries, regardless if they become a member of the Eurozone later on or not. The other dummy variable (“later member of Eurozone”) adopts the value of 1 for countries of the later Eurozone after 1992, countries (Denmark, Sweden, UK) that have opted out, continue to receive 0. Greece became an EMU member in 2001 only, but gets 1 after 1999 to control for consolidation efforts prior to accession to the EMU.
Table 1(7) presents the results of the analysis. We can draw four conclusion from the findings: First, the coefficient of the lagged dependent variable is reasonably small, ranging from 0.869 to 0.907.
7
Second, most of the political and institutional variables do not have very much explanatory power in the sense that they are all statistically significant, although t-values for some are reasonably high. Only the number of ministers in a government and the number of years of the chief executive party in office seems to have a significant impact in all the models. Partisan variables perform better: The negative sign on the coefficient of the interaction variable between socialdemocratic share of seats in government and unemployment rates indicates that socialdemocratic governments seem to be willing to run higher deficits in times of high unemployment (confirming Carlsen 1997). In other times, left governments seem to run higher surplusses / lower deficits than christian democratic or conservative governments. This can be seen as an indication of the hypothesis that right wing governments tend to increase deficits, because they focus on lowering taxes without implementing the necessary spending cuts (Cameron 1985; Cusack 1999). The magnitude of the effect is not negligible: A complete turnover from a purely christian democratic government to a purely socialdemocratic one is associated with an improvement in current deficits of 1.7%, controlling for all the other independent variables.
Third, let us look at macro-institutional variables: Central bank independence seems not to be associated with changes in current deficits, although it may be important for gross debt levels. The results show that a high number of vetoplayers in a country are associated with higher deficits. Vetoplayers often play the role of countermajoritarian institutions, thereby making it hard for governments to implement spending cuts or non-expansive welfare state reforms, because vetopoints allow opposition forces to enter the political struggle and to blame the governments for unpopular measures. The government is thus very reluctant to implement these reforms.
Fourth, let us look at the international variables. The sign of the trade openness variable gives limited support to the ‘compensation thesis’, i.e. higher exposure to international competition leads to more fiscal prudence. EMU-membership or membership in the EU during the 1990s (being part of the wider process of fiscal consolidation that followed the enactment of the Maastricht Treaty) does not seem to have too much of an impact on current fiscal deficits. Furthermore, there is no difference between countries in the Eurozone and the ones that stay outside. But the sign of the coefficient indicates that EMU membership has had a positive impact on the fiscal performance of member states. What is more, this effect is discernible even when we control for central bank indepence (CBI). Membership in EMU is therefore not only about making your country’s central bank more independent, there is also a positive effect on fiscal policy independent of CBI.
Table 2(8) presents the results of a re-analysis with the same independent, but a different dependent variable, namely changes in gross debt levels as percentage points changes between gross debt levels in % of GDP.(9) Because the lagged dependent variable is not included as independent variable,(10) the R² is siginificantly lower (it varies between 0.3 and 0.4). On the other hand, it has more explanatory power in the sense that a higher R² now provides more substantive information about the goodness-of-fit of the model.
In general, the findings are confirmed. The sign of the coefficient on the ‘number of ministers in cabinet’ indicates that a higher number of ministers in a cabinet results in higher levels of gross debt. The larger the majority of the party in government in parliament, the more probable it is that the government will reduce or avoid high increases in gross debt levels. A government with a large majority in the legislature is less willing to increase gross debt levels, because the visibility of its actions is very high and it can be easily blamed for incautious fiscal policies.
8
Also, the years that the party of the head of government stays in power (government turnover rate) has a positive impact on fiscal performance. Parties that stay in power longer seem to adopt a more long-term view on fiscal policy issues, which is rational, because the probability that they will still be in power then is quite high. These findings are in line with the mainstream in political economy literature (going back to Roubini / Sachs 1989b).
Partisan variables perform well, too, and give some support on the above mentioned thesis on the unability of right wing parties in governments to reduce debt levels. This is especially true for christian democratic parties, who are more reluctant towards spending cuts in the welfare state than conservative parties (Van Kersbergen 1995). The signs and hypotheses about CBI and vetoplayers are also confirmed. It is striking that the CBI variable performs much better than in the preceding analysis in terms of statistical significance. A possible explanation for this might be that for governments, the central bank’s grip on interest rates has a deterring effect only for gross debt levels, whereas the deterrence effect for running current account deficits is not as strong.
Most important for our analysis is however the strong performance of the EMU variables. EU-membership during the 1990s and participation in the process of fiscal consolidation triggered by the Maastricht treaty has had a positive impact of fiscal performace. This is not only true for the later members of the Eurozone, but also for the countries that did not join the third stage of EMU (as can be seen from the higher magnitude of the “EU member after 1992” dummy). Trade openness also had a positive and statistically significant impact on fiscal performance.
The overall conclusion of this analysis is that (prospective) membership in the EMU has had a significant and positive impact on the fiscal performance of member states. It is remarkable, that this EMU effect can be traced even when controlling for a variety of other variables, that have been regarded as most influential for fiscal performance of nation states.
The question that arises at the end of this chapter is if this effect can be ascribed to the supranational institutional arrangements of EMU (i.e. the Stability and Growth Pact) or if membership in the EMU was just another argument in the domestic political battle between governments, social actors, and voters. A positive answer to the first hypothesis would support liberal and functional interpretations of the process of European integration, while a positive answer to the latter would imply a stronger role of the nation state. A crucial test for both perspectives is to look at the fiscal performance after 1999, when the impact of supranational arrangements is expected to be even stronger.
This chapter tries to trace the Maastricht effect beyond 1999, contrasting the development in EU and EMU member countries with non-members.(11) The impact of EMU could be rated positive if we can trace a process of continuous improvement in deficits and debt levels in EU and / or EMU countries compared to non-EU members over the past two decades.
Table 3 displays some key figures. Generally, EU member states have performed worse, no matter if fiscal performance is measured by looking at the current fiscal balance, the change in debt levels, or the gross debt level itself. Deficits are higher, debt levels have increased faster, and the level of gross debt is generally higher in EU member states than in non-EU countries. Above all, the difference in means regarding fiscal balances and gross debt levels is highly significant, indicating that for these variables the differences are especially pronounced.
9
Looking at the dynamics though gives a more detailed picture. While in the period between 1980 and 1991, when the process towards EMU was more a distant future vision than a real policy project, the differences between EU and non-EU countries was most pronounced. EU countries had an average debt level of 63.47% of GDP compared to 45.84% for non-EU countries, a striking difference of 17.62 percentage points! EU member states also ran significantly higher deficits (5.64% of GDP versus 1.85%). One reason for the generally lower debt levels in non-EU countries is the different nature of welfare state regimes. These countries are among others Japan, USA, New Zealand, Australia, and Canada, all belonging to the liberal family of welfare state regimes (Esping-Andersen 1990; Obinger / Wagschal 2000). This group of countries is denoted by common features like a lean welfare state, low public spending on social policy, and a tendency towards the use of regulation instead of spending to steer the economy.
During the period between 1992 and 1997, the process of economic integration was set into motion and the prospect of the decision on the number of countries to join EMU in 1997 fuelled serious efforts to improve fiscal performance. The data shows that this period is a turning point. Above, I already mentioned that under special circumstances, high deficits can go along with slower increases in debt levels. And exactly this seems to be the case. While current fiscal balances in EU member states remain higher than in non-EU countries, the debt levels have not increased as fast as in the other countries. In the EU, the average increase of debt levels in this period was 1.27% of GDP per year, compared to 2.65% for non-EU countries. The difference in debt levels between the two groups has fallen from 17.62% to 12.54%, although debt levels in the EU countries are still significantly higher.
The trend seems to continue in the period between 1998 and 2002. Arguments that question the effectiveness of the EMU in constraining member states in spending in the long run expect the resurgence of high deficits after the threat of being excluded from the highly prestigious project of EMU is over. Because the decision on the number of member countries is made and has been made in a very compromising way, interpreting the convergence criteria taking political considerations into account, the external constraint of compliance is weakened.
Liberal or institutionalist theories would argue, that just after 1999, when the position of ECB was strengthened and the final stage of the economic and monetary union had been entered, fiscal performance should have increased further or at least not deteriorated. Functionalist theories have expected the spill-over from the single market to monetary and economic integration and predict a further step towards a common economic policy that encompasses monetary and fiscal policy. The final stage of EMU has institutionalized a Central Bank that is even more independent than its German model, because its legal framework is laid down in an international treaty that is harder to modify than the simple law upon which the German Bundesbank was based. The Stability and Growth Pact (SGP) is the second important component of the EMU. There has however not been a step towards a higher degree of institutionalization of the SGP after entering the third stage.
The data in table 3 indicate that the trend of the 1990s has not been reversed after 1998, but reinforced. The difference in debt levels has decreased further 6.64% of GDP, making the difference in means statistically insignificant. The same logic applies to current fiscal balances. The trend in the changes in debt levels variable is also more pronounced and, for the first time since the 1980s, has led to a real reduction in debt levels, not only to a lower increase. EU members have decreased their debts on average by 2.26% of GDP per year, compared to only 0.54% for non-EU countries.
10
A caveat has to be added to this analysis: The data also shows a general trend towards an enhanced fiscal performance in the 1990s. The EU member states have followed this trend, but at the same time, they have outperformed non-EU countries in relative terms, although they only managed to reduce the difference and not to outperform non-EU countries in absolute terms. The picture gets sharper when we look at EMU countries instead of EU member states.
In the period between 1992 and 1997, the difference in debt levels shrinks to 6.85%. Differences in current balances are not significant any more (-3.84% of GDP for non-EMU countries, -3.54% for EMU members). With regard to changes in gross debt level, EMU countries clearly outperform non-EMU states, indicating that the efforts of EMU countries to join were mainly aimed at reducing the debt levels, less at reducing the deficit.
The period between 1998 and 2002 includes some surprises, because we can discern first signs of a deteriorating fiscal performance. Although EMU countries managed to balance their budgets, non-EMU member countries performed better by achieving surplusses of 1.14 of GDP on average. The difference between the two groups has increased again and is almost statistically significant. EMU countries continue to outperform non-EU states in efforts to reduce debt levels, although the difference is decreasing here, too. The difference in levels of gross debt could be reduced further from 6.85% of GDP for the period between 1992 and 1997 to 5.91% between 1998 and 2002. But EMU members still have a higher level of public debt than non-EMU countries (67.49% compared to 61.58%).
To sum up, the inspection of differences in means has produced some insights. First, there has been an overall trend towards improved fiscal perfomance in all 22 OECD states. Second, EU member states have decreased the performance gap towards non-EU countries, achieving almost similar performance levels at the end of the 1990s. Third, EU countries were more successful in reducing debt levels than they were in reducing deficits. Fourth, trends for EMU members were similar in direction, but more pronounced in size. Therefore, membership in the EMU (or prospective membership for the period between 1992 and 1997) has enhanced fiscal performance. This confirms the results of our analysis in chapter 2. Finally, there are some signs for the latest period that point towards a deterioration in fiscal performance for EMU members. At this point of the analysis, one cannot say if these signs are systematically linked to weaknesses in the institutional arrangement of the EMU or if they are temporal fluctuations. The next chapter will deal with this question.
This chapter tries to trace differences in performance between big and small member states of the EU.
‘Big’ countries in this analysis are France, Germany, Italy, Spain, and the United Kingdom, small countries are all the others. This is a very crude distinction with regard to actual population levels, but it is a distinction that has proven to be politically very significant in the EU context.
11
Table 5 shows that small countries have in general performed better than big countries. But interestingly, on average, small countries have a higher debt level (76.71% compared to 74.09% for the big countries). But this changes for the latest period in which small countries only had a debt level of 62.62% compared to 76.15% for big countries. Small countries have managed to generate budget surplusses of on average 0.91%, while big countries are still struggling with deficits of about 1.51%. Small countries also have a higher negative rate of debt level change (-2.64% versus -1.51%), indicating that small countries managed to reduce debt levels faster than big countries. In general, small EU member states perform better with regard to debt levels, changes in debt levels, and deficits than big countries do. What is more, the relationship between size of a country and fiscal performance seems to hold only for the EU context of the 1990s. Plugging the variable ‘population size’ in the analytical framework of chapter 2 (in bivariate or multiple regressions) does not yield statistically significant results.
The findings from the data are supported by the analysis of Von Hagen / Hallett / Strauch (2001), who distinguish between successful and unsuccessful consolidation strategies. They challenge the simple assumption that large consolidations are the most successful ones. Instead, they point to the importance of the nature of the consolidation. Successful consolidations are in their view those that are driven by expenditure cuts without reducing public capital investments, unsuccessful consolidations those that focus on creating more revenue without implementing the necessary spending cuts that imply high political costs (ibd.: 8-9). Their data show that the effect of consolidation strategies that are based on expenditure cuts is more sustainable in the long term than the effects of revenue-based strategies. They find successful consolidation predominantly in Austria, Finland, Sweden, and the UK, unsuccessful consolidation are to be found in Greece, Portugal, and Germany (ibd.: 39). However, the fiscal performance of the UK has been deteriorating since and in 2003, it breached the 3% ceiling. Greece and Portugal are small countries and are prone to clientelism, which makes fiscal consolidation hard. What is more, Portugal had shown some real effort to consolidate after having violated the 3% criterion in 2001.
There are two explanations for the better performance of small countries. First, in the period before the final stage of EMU was entered in 1999, the threat of being excluded from EMU because of the failure to meet the convergence criteria was more real for small countries than for the large countries in the core of the EU. Smaller countries, especially in the European periphery (Portugal, Greece, Ireland, Finland), had to cope with the serious threat of not being included into the EMU. This was also true for states in the European core like Belgium or the Netherlands, both of which had to undergo serious reforms to convince the bigger countries (mainly Germany) of their ability to join. Spain and Italy were borderline cases (Boix 2000). Speculation if these countries could have joined even if they had not been able to fulfill the convergence criteria at least to some extent is superfluous. But in the Italian case, domestic political actors had a strong incentive to use EMU as a “vincolo esterno” (Dyson / Featherstone 1996) to overcome domestic barriers against reform. In Germany and France, the credibility of the threat of being excluded from EMU was very limited.(12) It was rather obvious, that if Germany and France were not able to meet them, the criteria would be interpreted more flexibly or the whole project of monetary union would be postponed. Thus, the threat of being excluded could not be used by the French and German governments as effectively to overcome domestic barriers to reform as in other countries.
The second explanation for the better performance of small states lies in the institutional arrangement of the SGP. Because the SGP regime on fiscal policy still contains important elements of intergouvernmentalism, power relations between small and big states are mirrored in the extent to which compliance with supranational rules can be enforced.
12
Small states are constrained more, because coalition-building in the Council to avoid sanctioning is harder for them than for big states with more power resources. First, I briefly go through the legal provisions of the EC treaty to show that in the field of fiscal policy, the role of the nation states is more pronounced than in other fields, f.e. monetary policy. Second, a look at the first cases of violations of the SGP (Portugal, France, Germany) clarifies that the ability of the Commission to sanction and therefore the extent to which member states comply with the rules of the SGP depends on the size and power of the condemned country.
Art. 104 of the EC treaty regulates the Excessive Deficit Procedure (EDP) that is binding for every EU member. Council Regulation 1467/97 covers the specifics of the Stability and Growth Pact (SGP). In contrast to the field of monetary policy, fiscal policy within the SGP still has a strong intergovernmental component. Monetary policy is in the hands of the highly independent European Central Bank, whereas in the EDP after the Comission has already reached a decision that deficits in a country are ‘excessive’ the Council can decide if it wants to follow the Commission by qualitative majority. This intergovernmental element in decision-making weakens the power of the SGP, because it challenges the effectiveness of the sanctions mechanism. A monetary union like the EMU is always faced with massive free-riding problems: Some countries enjoy the benefits of memberships (high currency stability) without paying the price (sound fiscal policy). Because Germany feared that countries like Italy would become the free-riders of the new monetary union, the former finance minister Theo Waigel strongly promoted the establishment of the SGP. The problem of free-riding was believed to be minimized by establishing an institutional arrangement like the SGP that gives clear guidelines on fiscal policy and contains a strong sanctions mechanism. The Council can force the compliance of member states by requiring deposits that can eventually be transformed into fines with an upper limit of 0.5% of the violating state’s GDP if the member state in question does not follow the recommendations of the Council and does not reduce its deficit accordingly.
The weak point in the EDP is the possible discretionary leeway that is given to the Council. After the first report of the Commission the Council can decide if the deficit is ‘excessive’ “having considered any observations which the Member State concerned may wish to make” (Art. 104 (6) EC Treaty). The Council makes recommendations to the member state how to solve the problems. Only when the deficit remains excessive over a period of time may the Council “decide to give notice to the Member State to take, within a specified time-limit, measures for the deficit reduction which is judged necessary by the Council in order to remedy the situation“ (Art. 104 (9) EC Treaty). After the Council has made a decision according to Paragraph 9, the Council may then decide to require deposits or fines (Art. 104 (11) EC Treaty). The EDP is thus a long process that is vulnerable to all sorts of political pressure on the various stages. The Council for example decided on January 21st and on June 3rd 2003 that the deficits of Germany and France, respectively, are to be called ‘excessive’, but just a few months later the Council refused to follow the recommendations of the Commission to press ahead with sanctions.
A telling example for the different impact the SGP and EMU had on fiscal performance is the comparison between the cases of Portugal, Germany, and France. Portugal was the first country to breach the 3% ceiling in 2001. The Council found in November 2002 that an excessive deficit exists in Portugal. The newly elected center-right government under José Manuel Durao Barroso promised to cut the deficit from 4.1% in 2001 to 2.8% in 2002, although it had come to power only in April 2002. Portugal’s finance minister, Ms Manuela Ferreira Leite, said: “The end of the world for us would be finding that we weren´t taking the necessary measures and that we could face sanctions.” (The Irish Times, Nov 6 2002, p. 16)
13
Barroso reacted to the Council’s condemnation by imposing spending cuts, increases in the value-added tax, delayed tax reductions, and a freeze in civil recruitment (Financial Times, Dec 12 2002, p. 4), which triggered a country-wide 24-hour general strike in December 2002. The Portuguese government reached its aim for 2002 and reduced the deficit to 2.7% of GDP. Although the prospects for future budget balances do not look as rosy anymore, the case of Portugal can still be regarded as an example of a country in which the rules and institutions of the SGP have led to an enhancement in fiscal performance, even though it meant large political costs for the incumbent Barroso. The disciplining impact of the SGP was strong, because Portugal could expect to be sanctioned by the Council for not complying with the SGP. Because it is a small country with only few allies other than Spain and maybe Luxembourg, it could not count on being left unpenalized.
The case is different for France and Germany. Both countries have breached the 3% ceiling for 2002, 2003, 2004 (Financial Times London, May 5, 2004, p. 6), and are expected to do so for 2005 (FT London, May 10 2004, p. 6) . Britain has joined its partners in violating the EU’s budgetary rules for the year 2003 (The Daily Telegraph, Feb 28 2004, p.30). In November 2003, the Council did not follow the Commission’s proposals to impose sanctions on Germany and France(13). Britain has supported both in their efforts to form a qualitative majority to avoid condemnation. The Netherlands, Austria, and Finland as well as Sweden protested violently against the Council’s decision to let Germany and France get away. Gerrit Zalm, the Dutch finance minister, even said that some countries had been “intimidated” by France and Germany (Sunday Times, Nov 30 2003, p. 10). The constellation big versus small states is clearly identifiable. Governments of small states were furious, because they had invested much political capital in complying with the Stability Pact. Now they had to witness the deconstruction of the Pact, because big states did not manage to keep their finances under control. This becomes even more painful, when one keeps in mind that the SGP had been introduced following an initiative by the very same states that now did not manage to comply. Many of the small states had to buy their membership of the EMU at a high price, because the fiscal conservatism of the SGP was against their political preferences. This is not only true for social democratic countries like Sweden, Finland, Austria, or Denmark, but also for Belgium that used public spending to cope with its internal divisions, or Greece and Portugal which had to come clear of clientelism.
The failure of imposing sanctions on the two most important countries in the EU shows the weaknesses of the institutional arrangement of the SGP. Letting the Council decide on sanctions introduces a strong element of intergovernmental decision-making. Instead of multilateralism, which implies a certain degree of equality between states, intergovernmentalism tends to reinforce existing inequalities in power. To allow the Council a discretionary decision on the final imposition of sanctions can be regarded as a ‘safety mechanism’ deliberately installed by the political actors when they were negotiating about EMU during the Maastricht conference. However, this safety mechanism could tear down the whole project of the SGP.
This paper has shown that the process towards EMU and the SGP has had a positive impact on fiscal performance in the last decade, although there are signs of deteriorating performance at the beginning of the new decade. The second finding of this paper is that small states have generally perfomed better. The reason for this is presumably the weak position of the small states in the intra-EU power struggle. Small states had to cope with the serious possibility of being excluded from EMU. They could also not rely on a strong coalition of allies in the Council to bail them out. The implication of these findings is that in the field of fiscal policy the role of the nation states remains more important than the seemingly ambitious EMU integration project would make one expect.
14
The deteriorating fiscal performance of big states in the last years indicates that the institutional arrangement of the SGP alone cannot prevent unsound fiscal policies. Thus, in the field of fiscal policy, intergovernmental theories seem to have more explanatory power than functionalist approaches, which expect a continuing high fiscal performance.
The idea of the project of the SGP was to impose an external constraint on governments to stymie them in unresponsible spending. This constraint was supposed to function similarly as the independent central bank for monetary policy did. The problem of the SGP is that it had never been an external constraint, because it was not independent of the governments’ decisions in the Council. The fundamental flaw in the construction of the SGP is not necessarily its rigidity, which has led the head of the Commission, Romano Prodi, to call the Stability Pact “stupid” in October 2002. The fundamental flaw lies more in the fact that the SGP was never an external constraint, but merely a temporary self-constraint. In that it is similar to a self-prescribed healthy diet, including the vicious ‘yoyo’ effect in the aftermath.
The position of the Commission to sue the Council for not upholding treaty law is understandable, but dangerous. The non-compliance of Germany and France seriously questions the supranational basis of the European integration project. Out of this perspective, the Commission has to adopt an uncompromising stance if it wants to uphold its position and the supranational nature of the European project. The Commission fully understood the scope of this challenge, as this statement of a spokesman shows: “It had to be done. It’s our role to see that treaty provisions are upheld. This kind of approach could be repeated if we don’t clarify the situation.” (The Daily Telegraph, Jan 14 2004, p. 29)
The upcoming enlargement increases the vulnerability of the supranational project. The failure of making old EU member states comply with treaty regulations could coincide with a tendency to re-interpret the European project that results from a possible overstretch of European institutions in an enlarged union with an ever wider variety of nations, cultures, and languages. A weakening of the position of the Commission could complicate efforts to uphold the supranational nature of the EU.
In the end, the show-down in front of the Court that the Commission will provoke by enhancing legal action against the Council could be counterproductive. Instead, it would have been better if the Commission had come forward with substantial and well-coordinated proposals on how to reform the SGP and the EDP, when criticism was growing in 2002 and 2003. The Commission losing the battle is a real danger, although the European Court of Justice has often ruled in favor of integrationist positions, even when this implied a serious bending of treaty provisions. The Court is not only a judiciary, but also a political institution that is interested in the continuation of the European integration process.
Alesina, Alberto / Grilli, Vittorio, 1993: On the Feasibility of a One-Speed or Multispeed European Monetary Union; in: Economics and Politics 5/2: 145-165.
Alesina, Alberto / Perotti, Roberto, 1999: Budget Deficits and Budget Institutions; in: Poterba, James M. / Von Hagen, Jürgen (eds.): Fiscal Institutions and Fiscal Performance (Chicago, London: University of Chicago Press), pp. 13-36.
Artis, Michael J. / Buti, Marco, 2000: ‚Close-to-Balance or in Surplus’: A Policy-Maker’s Guide to the Implementation of the Stability and Growth Pact; in: Journal of Common Market Studies 38/4: 563-591.
Beck, Nathaniel / Katz, Jonathan N., 1995: What to do (and not to do) with Time-Series Cross-Section Data; in: American Political Science Review 89/3: 634-647.
Beck, Nathaniel / Katz, Jonathan N., 1996: Nuisance vs. Substance: Specifying and Estimating Time-Series-Cross-Section Models; in: Political Analysis 6: 1-36.
Beck, Thorsten / Clarke, George / Groff, Alberto / Keefer, Philip / Walsh, Patrick, 2000: New tools and new tests in comparative political economy: The Database of Political Institutions, World Bank: Development Research Group, Federal Department of Foreign Affairs (Switzerland) <http://econ.worldbank.org/view.php?type=18&id=25467>
Beetsma, Roel M.W.J. / Bovenberg, A. Lans, 2002: Strategic debt accumulation in a heterogenous monetary union; in: European Journal of Political Economy 19/1: 1-15.
Bernaldo de Quiros, Lorenzo, 1999: The European Monetary Union: A Political Time Bomb; in: Cato Journal 19/1: 143-160.
Boix, Carles, 2000: Managing the Spanish Economy within Europe; in: South European Society and Politics 5/2: 165-191.
Borrelli, Stephen A. / Royed, Terry J., 1995: Government ‘strength’ and budget deficits in advanced democracies; in: European Journal of Political Research 28: 225-260. Buti, Marco / Giudice, Gabriele, 2002: Maastricht’s Fiscal Rules at Ten: An Assessment; in: Journal of Common Market Studies 40/5: 823-848.
Buti, Marco / Giudice, Gabriele, 2002: Maastricht’s Fiscal Rules at Ten: An Assessment; in: Journal of Common Market Studies 40/5: 823-848.
Cameron, David R., 1985: Does Government Cause Inflation? Taxes, Spending, and Deficits; in: Lindberg, Leon N. / Maier, Charles S. (eds.): The Politics of Inflation and Economic Stagnation: Theoretical Approaches and International Case Studies (Washington, D.C.: The Brookings Institution), pp. 224-279.
Carlsen, Fredrik, 1997: Counterfiscal policies and partisan politics: evidence from industrialized countries; in: Applied Economics 29: 145-151.
Clift, Ben, 2002: The political economy of the Jospin government; in: Modern & Contemporary France 10/3: 325-337.
Cusack, Thomas R., 1999: Partisan Politics and Fiscal Policy; in: Comparative Political Studies 32/4: 464-486.
De Haan, Jakob / Sturm, Jan-Egbert, 1994: Political and institutional determinants of fiscal policy in the European Community; in: Public Choice 80: 157-172.
Dyson, Kenneth / Featherston, Kevin, 1996: Italy and EMU as a ‘Vincolo Esterno’: Empowering the Technocrats, Transforming the State; in: South European Society and Politics 1/2: 272-299.
Edin, Per-Anders / Ohlsson, Henry, 1991: Political determinants of budget deficits: Coalition effects versus minority effects; in: European Economic Review 35: 1597-1603.
Esping-Andersen, Gøsta, 1990: The Three Worlds of Welfare Capitalism (Cambridge: Polity Press).
Ferguson, Niall / Kotlikoff, Laurence J., 2000: The Degenration of EMU; in: Foreign Affairs 79/2: 110-121.
Franzese, Robert Jr., 2002: Macroeconomic Policies of Developed Democracies (Cambridge: Cambridge University Press).
Freitag, Markus, 2000: Die politischen Rahmenbedingungen des Euro: Glaubwürdige Weichenstellungen oder Gefahr möglicher Entgleisungen? In: Zeitschrift für Politikwissenschaft 10/3: 967-997.
Freitag, Markus / Sciarini, Pascal, 2001: The Political Economy of Budget Deficits in the European Union; in: European Union Politics 2/2: 163-189.
Genschel, Philipp, 2000: Die Grenzen der Problemlösungsfähigkeit der EU; in: Grande, Edgar / Jachtenfuchs, Markus (eds.): Wie problemlösungsfähig ist die EU? Regierung im europäischen Mehrebenensystem (Baden-Baden: Nomos), pp. 191-207.
Grilli, Vittorio / Masciandaro, Donato / Tabellini, Guido, 1991: Political and monetary institutions and public financial policies in the industrial countries; in: Economic Policy 13: 341-392.
Hahm, Sung Deuk / Kamlet, Mark S. / Mowery, David C., 1996: The Political Economy of Deficit Spending in Nine Industrialized Parliamentary Democracies; in: Comporative Political Studies 29/1: 52-77.
Hallerberg, Mark, 2002: Introduction: Fiscal Policy in the European Union; in: European Union Politics 3/2: 139-150.
Hallerberg, Mark / von Hagen, Jürgen, 1999: Electoral Institutions, Cabinet Negotiations, and Budget Deficits in the European Union; in: Poterba, James M. / Von Hagen, Jürgen (eds.): Fiscal Institutions and Fiscal Performance (Chicago, London: University of Chicago Press), pp. 209-232.
Heinemann, Friedrich, 2000: Globalisation, EMU and Fiscal Discipline: Evidence from OECD Countries; in: Aussenwirtschaft 55/2: 211-232.
Hibbs, Douglas A., Jr., 1977: Political Parties and Macroeconomic Policy; in: American Political Science Review 71: 1467-1487.
Kittel, Bernhard, 1999: Sense and sensitivity in pooled analysis of political data; in: European Journal of Political Research 35: 225-253.
Kontopoulos, Yianos / Perotti, Roberto, 1999: Government Fragmentation and Fiscal Policy Outcomes: Evidence from OECD Countries; in: Poterba, James M. / Von Hagen, Jürgen (eds.): Fiscal Institutions and Fiscal Performance (Chicago, London: University of Chicago Press), pp. 81-102.
McKay, David, 1999: The Political Sustainability of European Monetary Union; in: British Journal of Political Science 29/3: 463-485.
Obinger, Herbert / Wagschal, Uwe, 2000: Ökonomie, Institutionen und Politik: Determinanten der gebremsten Sozialstaatlichkeit im Überblick; in: Obinger, Herbert / Wagschal, Uwe (eds.): Der gezügelte Wohlfahrtsstaat: Sozialpolitik in reichen Industrienationen (Frankfurt: Campus), pp. 365-390.
Obstfeld, Maurice / Peri, Giovanni, 1998: Regional non-adjustment and fiscal policy; in: Economic Policy 13: 207-259.
Persson, Torsten / Tabellini, Guido, 1996: Federal Fiscal Constituions: Risk Sharing and Moral Hazard; in: Econometrica 64/3: 623-646.
Pierson, Paul, 2001: From Expansion to Austerity: The New Politics of Taxing and Spending; in: Landy, Mark K. / Levin, Martin A. / Shapiro, Martin (eds.): Seeking the Center (Washington, D.C.: Georgetown University Press), pp. 54-80.
Rotte, Ralph, 1998: International Commitment and Domestic Politics: A Note on the Maastricht Case; in: European Journal of International Relations 4/1: 131-142.
Roubini, Nouriel / Sachs, Jeffrey, 1989a: Government Spending and Budget Deficits in the Industrial Countries; in: Economic Policy 8: 100-132.
Roubini, Nouriel / Sachs, Jeffrey D., 1989b: Political and Economic Determinants of Budget Deficits in the Industrial Democracies ; in : European Economic Review 33 : 903-938.
Sakamoto, Takayuki, 2001: Effects of Government Characteristics on Fiscal Deficits in 18 OECD Countries 1961-1994; in: Comparative Political Studies 34/5: 527-554.
Schmidt, Manfred G., 2000: Demokratietheorien, 3rd edition (Opladen: Leske + Budrich).
Schmidt, Manfred G., 2003: Die parteipolitische Zusammensetzung der OECD-Demokratien (Heidelberg: Institut für Politische Wissenschaft).
Schürz, Martin, 2000: Wirtschaftspolitische Problemlösungsfähigkeit in der Wirtschafts- und Währungsunion; in: Grande, Edgar / Jachtenfuchs, Markus (eds.): Wie problemlösungsfähig ist die EU? Regierung im europäischen Mehrebenensystem (Baden-Baden: Nomos), pp. 209-227.
Tsoukalis, Loukas, 2000: Economic and Monetary Union: Political Conviction and Economic Uncertainty; in: Wallace, Helen / Wallace, William (eds.): Policy-Making in the European Union, 4th edition (Oxford: Oxford University Press), pp. 149-178.
Van Kersbergen, Kees, 1995: Social Capitalism: A Study of Christian Democracy and the Welfare State (London, New York: Routledge).
Von Hagen, Jürgen / Hallett, Andrew Hughes / Strauch, Rolf, 2001: Budgetary Consolidation in EMU; Economic Papers No 148, ZEI, Bonn.
Wagschal, Uwe, 1996: Staatsverschuldung: Ursachen im internationalen Vergleich (Opladen: Leske und Budrich).
(1) Being one of the few exceptions, Freitag (2000) expects the Euro to be strong in the long term, because of the general political framework and the credibility of the fiscal policy regime of the EU (the SGP). Heinmann (2000) with his optimistic view is also an exception.
(2) Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Japan, Luxembourg, Netherlands, Norway, New Zealand, Portugal, Spain, Sweden, United Kingdom, United States.
(3) The constructed variable (SOZUNEM) is obtained by multiplication of the variable ‘percentage share of social democratic parties’ members in government’ (PROZ_SD) and unemployment rates (UNEMPL).
(4) See also De Haan / Sturm 1994, who find no effect of Roubini and Sachs variables on fiscal deficits even if they are disintegrated into categorical variables for each dimension.
(5) In fact, Sakamoto (2001: 535) finds that the weakness of government variables are only weakly connected to institutional features of the electoral or party system.
(6) This is an additive index as well. Its scale is from 0 to 10. Included dimensions: concordance democracy, federalism, central bank independence, Lijphart-Index of judiciary control of the legislative, EU-Membership, pronounced minority protection, bicameralism, coalition government, self-administrative structures in the public social insurances, direct democracy. I corrected Schmidt’s values for the countries that have not been members of the EU in the 1980s. The fact that EU membership is part of Schmidt’s index is somewhat problematic, because I also control for EU and EMU membership in dummy variables. But my dummy variables aim to capture the effect of the 1990s (and are 0 for the 1980s) and EU membership is only one out of ten criteria in Schmidt’s index.
(7) To test robustness I also ran the same regressions in fixed-effects models. The signs and the magnitude of the coefficients were similar, only the t-statistics are somewhat lower.
(8) Again, to test robustness, the regressions were also run in a fixed-effects model.
(9) Please be aware that the interpretation of the signs of the coefficients has turned around. While in the former analysis, a negative sign meant deteriorating fiscal perfomance, a negative sign in this analysis stands for a reduction in debt levels and is therefore performance-enhancing. For positive signs, vice versa.
(10) Lagged gross debt is a lagged level variable and not a lagged dependent variable.
(11) Membership in the EU is captured on a yearly basis: Greece joins 1981, Spain and Portugal in 1986, Finland, Austria, and Sweden in 1995. The Non-EU member states referred to are the remaining states of the OECD country sample defined above.
(12) Clift (2002) tells the French story of alternating periods of expansion and fiscal consolidation. It became clear in the last years, that the fiscal consolidation strategy that was followed by the French government in the late 1990s was not sustainable in the long term (cf. Von Hagen et al. 2001).
(13) This action by the Council has triggerd legal action by the Commission, on which I comment briefly in the concluding remarks section.