Fiscal Reaction Functions of Different Euro Area Countries

Fiscal Reaction Functions of Different Euro Area Countries Klaus Weyerstrass Institute for Advanced Studies Department of Economics and Finance Josefstädter Strasse 39, A-1080 Vienna, Austria E-Mail: klaus.weyerstrass@ihs.ac.at; Tel. +43-1-59991-233 Preliminary version, 9 September 2015 Abstract This paper addresses the sustainability of public finances of different Euro area countries. Not least the financial and economic crisis and the subsequent European debt crisis has put the fiscal sustainability of several Euro area countries into question. However, the development of public finances varies considerably between Euro area member states. Estimations of fiscal reaction functions in this paper show that fiscal policies of Austria, Germany, Italy, Spain, Portugal, Ireland and Greece cannot be regarded as totally sustainable. Even when fiscal rules and their interaction with the past debt level are taken into account, in many cases the primary budget balance does either react not at all or only insignificantly to past increases in the debt level. It is too early, i.e. there are too few data points available at present to test whether fiscal policies have strengthened recently in the aftermath of the European debt crisis. JEL classification: C22, E62, F34, H63 Keywords: fiscal reaction functions, fiscal adjustment composition Introduction Fiscal sustainability requires that a government is able to repay its debt at some point in the future. The primary budget balance, i.e. the budget balance net of interest payments, is a key determinant of public debt dynamics. Bohn (1998) argues that the reaction of the primary balance to the government debt stock can be taken as an indicator of the sustainability of public finances. If an increase in the debt stock is followed by a strengthening of the primary budget balance, fiscal policy 1

is sustainable, since such a policy prevents the debt level from ever increasing. In this paper, various fiscal reaction functions are estimated for seven Euro area countries: Germany, Austria, Ireland, Greece, Spain, Portugal and Italy. These countries represent different fiscal policy stances and developments over the past years. In a first set of equations, it is tested whether the primary budget balance or the overall balance of the different countries reacts at all to an increase of public debt in the previous year. Then, it is investigated whether public spending or revenues, and if so, which parts of spending and/or revenues have been used to improve public finances. Past experience shows that budget consolidations at the expenditure side have been more successful and long-lasting than consolidations at the revenue side. As a reaction to the public debt crisis, the European countries have implemented fiscal rules aiming at containing public debt and correcting fiscal imbalances. In the empirical estimations, it is investigated whether the existence of such rules has indeed improved the response of fiscal policies to the past development of the debt level. The paper is structured as follows. In the next section, the concept of sustainability of public finances is addressed, followed by a section on fiscal reaction functions. Then, the empirical estimations and the results are described. The final section summarises the main findings and draws conclusions. Sustainability of public finances Public debt is sustainable if the intertemporal budget constraint is fulfilled. This requires that current spending on goods and services, i.e. public consumption, plus the cost of servicing previously issued debt equals current tax revenues plus the issuance of new debt. In this definition, taxes are net taxes, i.e. government revenues of taxes and social security contributions net of transfer payments. Following the presentation in Balassone and Franco (2000) as well as in Neck and Sturm (2008), this requirement of public finances can be formalised as follows. It it assumed that government borrowing takes the form of one-period bonds that pay an interest rate i t > 0 in period t. Government consumption in period t is denoted as G t. T t are (net) tax revenues in period t, and B t denotes government debt issued in period t. With these assumptions and difinitions, the government s budget constraint in period t is given by G t + (1 + i t )B t 1 = T t + B t This condition may be expressed in terms of ratios, i.e. in relation to nominal GDP. Let g t, t t and b t be the ratios of government spending on goods and services, tax revenues, and debt issuance to nominal GDP in period t, respectively. Since it is assumed that public debt takes the form of one- 2

period bonds, debt issuance in period t equals total debt at the end of period t. Under these conditions, the above equation can be rewritten as d t + 1 + i t 1 + gy t b t 1 = b t In this equation, d t = g t - t t is the primary budget deficit ratio (i.e. the difference between government expenditures other than interest payments, and tax revenues, both expressed in relation to GDP), and gy t is the growth rate of nominal GDP. According to this equation, the debt ratio increases if the government runs a deficit and, at the same time, the nominal interest rate exceeds nominal GDP growth. In the long run, it is not possible for governments to use the issuance of ever increasing new debt to repay old debt and to finance interest payments. This is known as no-ponzi scheme condition. The above equation in connection with the no-ponzi scheme condition leads to the following intertemporal government budget constraint t (d t ( 1 + gy s )) + b 1 + i 0 = 0 s t=1 s=1 b 0 is the current debt ratio. This requirement of public debt sustainability, i.e. for the government to remain solvent, implies that the present discounted value of primary deficits plus the value of current debt must be zero. This also implies that running substantial deficits over a long period is consistent with sustainability as long as these deficits can be repaid in the future by sufficiently high budget surpluses. A second interpretation of sustainable fiscal policy considers the evolution of debt in the medium term. Here, sustainability is interpreted as a given reduction of the debt-to-gdp ratio over a given time horizon toward a target value like the 60% threshold of the Maastricht treaty and the Stability and Growth Pact. This interpretation of debt sustainability is mainly justified by the view that governments with high debt levels are less flexible to respond to adverse shocks, as high debt servicing costs leave little room for fiscal policy intervention (Neck and Sturm, 2008). Consider the budget constraint of the government, formalized in the following way: b t+1 b t+1 b t = (r gr)b t + d t+1 where r denotes the real interest rate, and gr is the real GDP growth rate. Thus, the public debt ratio is reduced if the primary budget surplus exceeds debt servicing. This can be written as d t+1 (r gr)b t. 3

According to this equation, the debt-to-gdp ratio will increase indefinitely if the real interest rate exceeds real GDP growth, unless the government runs a sufficiently large primary budget surplus. In this approach, the interest rate and the GDP growth rate are assumed to be exogenous. As the above conditions have shown, public finances are not sustainable if the government runs primary budget deficits over a long period of time without securing that the accumulated debt is repaid at some point. Hence, public finances are sustainable if the government reacts to an increasing debt level by improving its primary budget balance. Whether this is the case can be assessed by estimating fiscal reaction functions. These will be addressed in the following section. Fiscal reaction functions A fiscal reaction function relates the budget balance, public revenues or expenditures to an increase in the debt level. Bohn (1998) argues that the reaction of the primary budget balance to changes in public debt can be viewed as an indicator of the prudence or sustainability of fiscal policies. If an increase in the debt level is followed by an improvement of the primary balance, fiscal policy can be regarded as sustainable, since more resources are made available to service debt. Such estimations are evidently backward-looking and only uncover the feedback from public debt within the estimation sample, but they cannot predict the fiscal reaction of a government in the future and hence whether the government will pay its debt back (Baldi and Staehr, 2014). Bohn (1998) finds a significantly positive coefficient for the USA in the 20 th century and concludes that policy-makers have eventually reacted to the accumulation of public debt during this period of time. Wyplosz (2006) and Staehr (2008), among others, apply the same methodology to European data and also find some evidence of positive feedback from the debt stock to the primary balance, but they also conclude that the feedback is difficult to estimate precisely because the data series are short. Piergallini and Postigliola (2012) find that the primary balance in Italy reacted positively to the debt level and conclude that politicians have taken corrective measures to ensure sustainability of public finances in Italy. Estimating a fiscal reaction function for Brazil with monthly data, de Mello (2008) finds that the primary balance reacted positively and strongly to lagged debt. 4

Estimating fiscal reaction functions for seven Euro area countries Countries and fiscal variables In this paper, various fiscal reaction functions are estimated for seven Euro area countries: Germany, Austria, Ireland, Greece, Spain, Portugal and Italy. These countries represent different fiscal policy stances and developments over the past years. Ireland, Greece, Portugal and Spain came into liquidity problems and got financial support from the other EU countries and from the IMF, although in the case of Spain the support was confined to the banking sector. Austria s and Germany s fiscal stances were in general more prudent. However, while Germany has achieve small budget surpluses since 2012, the budget deficit in Austria is still above 2 percent in relation to GDP. Since 2012 Italy s budget deficit ratio has been close to the 3 percent threshold of the Maastricht treaty. This deficit was almost exclusively driven by interest payments due to the high debt level, while the primary budget has been almost balanced. The debt ratios of Italy, Portugal and in particular Greece already surpassed 130 percent; Spain s debt ratio is close to and Ireland s slightly above 100 percent. Austria s public debt level recently rose to almost 85 percent, while that of Germany decreased to slightly less than 75 percent of GDP. Fiscal reaction functions are estimated for (1) the primary budget balance, (2) the verall budget balance, (3) total government revenues, (4) total expenditures, (5) direct taxes, (6) indirect taxes, (7) government consumption, (8) subsidies, (9) public investment, and (10) transfers. This enables to analyse (i) if fiscal policy has reacted at all to an increase in public debt, and (ii) if so, whether this consolidation was mainly undertaken by rising revenues or by cutting consumptive expenditures or public investment. The data, covering at most the period 1991 to 2014, were taken from the database of Eurostat. Fiscal rules In addition to estimating the reaction of the budget balance, revenues and expenditures, respectively, to the debt level, it was investigated whether the existence and the rigour of fiscal rules increases the reaction of the budget balance to a rising debt level. Information on fiscal rules for the Euro area countries have been taken from a comprehensive fiscal rules dataset compiled by the European Commission. The EU Commission, Directorate General for Economic and Financial Affairs (DG ECFIN), has compiled a dataset on fiscal rules in the 28 EU countries in force in the time period since 1990. This dataset is based on information collected directly from the EU Member States. The dataset covers all types of numerical fiscal rules (budget balance, debt, expenditure, and revenue rules) at all levels of government (central, regional, and local, general government, and social 5

security). The information was collected in a survey, starting in 2006 with annual updates since 2008. The latest available update is currently providing information for 2013. The questionnaires of these surveys requested information on the description and definition of the fiscal rule and its coverage, its statutory base, monitoring and enforcement mechanisms, as well as experience with the respect of the rule. To be effective in containing budgetary imbalances, fiscal rules need to be equipped with appropriate characteristics within the institutional framework of budgetary policy. Whether a fiscal rule will be respected or not depends to a large extent on its institutional features. To capture the influence of these features, DG ECFIN has constructed an index of strength of fiscal rules, using information on (i) the statutory base of the rule, (ii) room for setting or revising its objectives, (iii) the body in charge of monitoring respect and enforcement of the rule, (iv) the enforcement mechanisms relating to the rule, and (v) the media visibility of the rule. Based on the fiscal rule strength index for each rule, a comprehensive time-varying fiscal rule index for each Member State was constructed by summing up all fiscal rule strength indices in force in the respective Member State weighted by the coverage of general government finances of the respective rule (i.e. public expenditure of the government sub sector(s) concerned by the rule over total general government expenditure). In the presence of more than one rule covering the same government sub-sector, the second, third and fourth rules obtain weights ½, ⅓, and ¼, to reflect decreasing marginal benefit of multiple rules applying to the same sub-sector. The assigned weights are mainly determined by the fiscal strength of the rule and its coverage. The fiscal rules database contains time series for the fiscal rule index over the period 1990 to 2013. 1 Estimation results Two sets of estimations have been performed. Firstly, panel estimations for all seven Euro area countries covered in this paper (Austria, Germany, Greece, Ireland, Italy, Portugal, Spain). In these estimations, the following dependent variables have been included: (a) primary budget balance, (b) headline balance, (c) total public revenues, (d) total expenditures, (e) direct taxes, (f) indirect taxes, (g) public consumption, (h) public investment, (i) subsidies, and (j) transfers. Secondly, separate estimations have been performed for each of the seven countries. In these estimations, exclusively the reaction of the primary budget balance to the lagged debt level has been investigated, 1 http://ec.europa.eu/economy_finance/db_indicators/fiscal_governance/fiscal_rules/index_en.htm. 6

controlling for the influence of economic growth. In the following, first the panel estimation results will be described, followed by the results of the individual country estimations. Panel estimation results In all panel estimations, fixed country effects have been included. Taking the data availability and the presence of lagged variables into account, the estimation period is generally 1992 to 2014 for the estimations without fiscal rules, and 1992 to 2013 with fiscal rules included. The following tables show the estimation results. For each dependent variable, two estimations have been performed: one without and one with fiscal rules. In the tables, these estimations are donated with (a) and (b), respectively. The existence of fiscal rules per se may not influence public finances, since fiscal rules are meant to improve the reaction of public finances to an increase of the debt level. This is captured by including the fiscal rules index not separately, but in interaction with the past debt level. Since revenues are positively related to the state of the economy, in the equations explaining the budget balance or revenues, GDP has been included. Whether nominal or real GDP in levels or in growth rates was taken was decided on the basis of the significance of the coefficients. In the tables, standard errors are shown below the coefficients. In addition, the significance of the coefficients is visualised by asterisks: *, **, *** means significance at the 10, 5, 1 percent level. As Table 1 shows, over the period 1991 to 2014 the primary budget balance in the panel of the seven Euro area countries has improved if in the previous year public debt has risen, but the reaction was not significant (panel a). The presence of fiscal rules increased the reaction of the primary budget balance to the past debt development, but it remained insignificant (panel b). In the case of the overall budget balance, i.e. including interest payments, the improvement as a reaction to past debt becomes significant once the interaction between lagged debt and fiscal rules is included. Even if the primary budget balance does not improve significantly as public debt rises, this might be the case for either revenues or expenditures. According to the results shown in Table 2, both total revenues and expenditures increase significantly if public debt has risen in the previous year. While the reaction of revenues goes into the right direction in terms of improving the budget balance, the contemporaneous increase of expenditures thwarts this improvement. The rise of public expenditures alongside with revenues is also the reason why the primary budget balance does not improve significantly. While the inclusion of the interaction of lagged debt with the fiscal rule index confirms the results obtained without fiscal rules, the reaction of revenues and expenditures does only change insignificantly. 7

Table 1: Panel estimation results for 7 Euro area countries (1) REAL GDP GROWTH NOMINAL GDP LEVEL PRIMARY BALANCE (a) -2.816 * (1.186) 0.016 (0.013) 0.805 *** (0.095) PRIMARY BALANCE (b) -2.871 * (1.263) 0.017 (0.014) 0.003 (0.004) 0.819 *** (0.101) HEADLINE BALANCE (a) -6.188 *** (1.322) 0.003 (0.015) 0.480 *** (7.334) HEADLINE BALANCE (b) -3.757 ** (1.544) -0.031 (0.019) 0.013 *** (0.005) 0.509 *** (0.066) ADJUSTED R² 0.323 0.315 0.360 0.405 NO. OF OBSERVATIONS 149 142 149 142 Fixed country effects are taken into account; standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. Table 2: Panel estimation results for 7 Euro area countries (2) REAL GDP GROWTH NOMINAL GDP LEVEL TOTAL REVENUES (a) 38.789 *** (0.724) 0.041 *** (0.008) 0.058 (0.033) TOTAL REVENUES (b) 38.996 *** (0.824) 0.038 *** (0.001) 0.001 (0.002) 0.062 (0.035) TOTAL EX- PENDITUES (a) 41.010 *** (1.261) 0.070 *** (0.016) TOTAL EX- PENDITURES (b) 40.231 *** (1.381) 0.079 *** (0.017) -0.007 (0.005) ADJUSTED R² 0.879 0.878 0.581 0.583 NO. OF OBSERVATIONS 146 139 149 149 Fixed country effects are taken into account; standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. 8

Table 3 shows that governments use both direct and indirect taxes to raise revenues if public debt increases. However, the reaction of direct taxes is significant (at the 10 percent level) with and without taking fiscal rules into account, while the increase of indirect taxes is significant only when fiscal rules are included. Table 3: Panel estimation results for 7 Euro area countries (3) REAL GDP GROWTH NOMINAL GDP LEVEL DIRECT TAXES (a) 10.308 *** (0.286) 0.013 *** (0.003) 0.026 * (0.014) DIRECT TAXES (b) 10.595 *** (0.833) 0.008 ** (0.004) 0.002 * (0.001) 0.030 ** (0.015) INDIRECT TAXES (a) 12.239 *** (0.325) 0.003 (0.003) 0.046 *** (0.016) INDIRECT TAXES (b) 12.454 *** (0.338) 0.0005 (0.039) 0.002 ** (0.001) 0.049 *** (0.016) ADJUSTED R² 0.884 0.885 0.756 0.762 NO. OF OBSERVATIONS 143 136 143 143 Fixed country effects are taken into account; standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. Turning to the expenditure side, even though total outlays are increased rather than reduced as would be necessary to improve public finances (see Table 2), individual expenditure items may actually be reduced. In fact, both public consumption and investment decrease as a reaction to a past rise of public debt (see Table 4). But while the reduction of consumptive expenditures is not significant, the cutback of investment is significant. In terms of the sustainability of budget consolidations as well as regarding the strengthening of the economic growth potential and thus of the potential to generate future revenues, it is problematic to cut public investment. It would be preferable to retain public investment, and to cut back on consumptive expenditures. 9

Table 4: Panel estimation results for 7 Euro area countries (4) REAL GDP GROWTH -1 PUBLIC CONS. (a) 19.681 *** (0.388) -0.007 (0.004) -0.262 *** (0.030) PUBLIC CONS. (b) 19.624 *** (0.446) -0.006 (0.005) 0.001 (0.001) -0.251 *** (0.032) PUBLIC INVEST. (a) 5.361 *** (0.325) -0.027 *** (0.002) PUBLIC INVEST.(b) 5.381 *** (0.179) -0.027 *** (0.002) -0.0007 (0.0006) NOMINAL GDP LEVEL ADJUSTED R² 0.594 0.590 0.757 0.779 NO. OF OBSERVATIONS 149 142 149 142 Fixed country effects are taken into account; standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. Table 5: Panel estimation results for 7 Euro area countries (5) REAL GDP GROWTH NOMINAL GDP LEVEL SUBSIDIES (a) SUBSIDIES (b) TRANSFERS (a) TRANSFERS (b) 1.325 *** 1.216 *** 16.662 *** 16.656 *** (0.108) (0.446) (0.410) (0.497) -0.002-0.0003 0.057 *** 0.057 *** (0.002) (0.002) (0.005) (0.006) -0.001 *** -0.001 (0.0004) (0.002) -0.451 *** -0.457 *** (0.033) (0.035) ADJUSTED R² 0.695 0.728 0.940 0.941 NO. OF OBSERVATIONS 149 142 149 142 Fixed country effects are taken into account; standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. 10

In addition to public consumption and investment, subsidies to companies and transfers to households may be reduced if public finances deteriorate. If this is case for the panel of seven Euro area countries investigated in this paper is shown in Table 5. Subsidies are indeed reduced as required to improve public finances, but this reaction becomes only significant in the presence of fiscal rules. On the other hand, transfers are not cut back, but even increase if public debt has risen in the previous year. The sign changes with fiscal rules, but the reduction of transfers in this case is insignificant. Summing the results of the panel estimations up, the budget balance improves only insignificantly as a reaction to a past increase of public debt. Both total revenues and expenditures rise. On the revenue side, primarily direct taxes are used to improve public finances. On the expenditure side, in particular public investment and subsidies are reduced to improve the budget balance. The presence of fiscal rules, which has been included in the estimations in interaction with the past debt level, slightly improves the reaction of public finances to a past debt increase. Estimations for individual countries In addition to the panel estimations described in the previous section, for each of the seven Euro countries separately it has been investigated whether the primary budget balance reacts significantly to the past development of public debt. Similar to the panel estimations, for each country two estimations have been performed, one without the fiscal rule, and one including the interaction between the past debt level and the fiscal rule index. Again, since the primary budget is positively connected to the state of the economy, nominal or real GDP, in levels or in growth rates has been included, depending on the significance of the coefficients. The results of these estimations can be found in Table 6 (Austria and Germany), Table 7 (Italy and Spain), Table 8 (Portugal and Ireland), and Table 9 (Greece). In Austria, the primary budget balance even deteriorates if in the previous year public debt has risen, but this reaction is not significant. When taking the interaction between the lagged debt level and the fiscal rule into account, the primary balance improves, but again the coefficient is not significant. In the case of Germany, both without and with the fiscal rule term the budget balance reacts into the correct direction, i.e. it improves as public debt has previously risen, but also in this case the coefficients are insignificant. The same pattern can be observed for Spain. 11

Italy is an interesting case. Without takin the interaction between lagged debt and the fiscal rule into account, the primary budget balance improves significantly as public debt rises, but once the fiscal rule is in addition taken into account the primary balance actually deteriorates significantly. Similar results have been found for Portugal, with the exception that the deterioration of the budget balance once the fiscal rule is taken into account is not significant. Only for Greece can a significant improvement of the budget balance after an increase of the debt level be found, both with and without taking the fiscal rule into account. Table 6: Estimation results for PRIMARY BALANCE as dependent variable (1) NOMINAL GDP LEVEL DUMMIES AUSTRIA (a) AUSTRIA (b) GERMANY (a) GERMANY (b) 4.739 * 8.151 * -2.673-2.314 (2.688) (4.031) (2.013) (2.462) -0.098-0.115 0.050 0.043 (0.002) (0.064) (0.030) (0.040) 0.015 0.002 (0.012) (0.008) 0.149 0.149 (0.121) (0.121) 0.00001 0.000002 (0.000009) 1995, 2004, 2009 (0.00001) 1995, 2004,02009 1995 1995 ADJUSTED R² 0.360 0.383 0.481 0.454 NO. OF OBSERVATIONS 23 22 23 22 Standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. 12

Table 7: Estimation results for PRIMARY BALANCE as dependent variable (2) REAL GDP GROWTH ITALY (a) ITALY (b) SPAIN (a) SPAIN (b) -5.262 * -5.176 * -7.270 *** -7.119 *** (4.004) (4.313) (1.666) (1.597) 0.068 * 0.062 0.033 0.015 (0.037) (0.041) (0.030) (0.027) -0.015 ** 0.008 (0.007) (0.005) 0.431 ** 0.453 ** (0.158) (0.147) 1.007 *** (0.092) 1.093 *** (0.104) ADJUSTED R² 0.293 0.453 0.864 0.876 NO. OF OBSERVATIONS 23 22 20 20 Standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. Table 8: Estimation results for PRIMARY BALANCE as dependent variable (3) PORTUGAL (a) PORTUGAL (b) IRELAND (a) IRELAND (b) -7.434 *** -9.487 *** -2.856 * -2.103 (1.606) (2.072) (1.447) (1.528) 0.059 *** 0.083 ** -0.016-0.031 (0.017) (0.023) (0.019) (0.022) -0.010 0.011 (0.007) (0.008) 0.050 *** 0.546 *** 0.576 *** 0.661 *** (0.119) (0.118) (0.084) (0.087) DUMMIES 2010 2010 2010 2010 ADJUSTED R² 0.685 0.708 0.897 0.905 NO. OF OBSERVATIONS 20 20 20 19 Standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. 13

Table 9: Estimation results for PRIMARY BALANCE as dependent variable (4) GREECE (a) -1.043 (6.135) -0.009 (0.049) 0.368 (0.250) GREECE (b) 23.079 ** (11.142) -0.205 ** (0.088) 0.042 ** (0.020) 0.207 (0.234) ADJUSTED R² 0.685 0.447 NO. OF OBSERVATIONS 20 19 Standard errors in brackets; *, **, *** mean significance at the 10, 5, 1 percent level. Summary and conclusions Sustainability of public finances requires that a government is able to repay its debt at some point in the future. The primary budget balance is a key determinant of public debt dynamics. If an increase in the debt stock is followed by a strengthening of the primary budget balance, fiscal policy is sustainable, since such a policy prevents the debt level from ever increasing. In this paper, various fiscal reaction functions are estimated for seven Euro area countries: Germany, Austria, Ireland, Greece, Spain, Portugal and Italy. These countries represent different fiscal policy stances and developments over the past years. In a first set of equations, it is tested whether the primary budget balance or the overall balance of the different countries reacts at all to an increase of public debt in the previous year. Then, it is investigated whether public spending or revenues, and if so, which parts of spending and/or revenues have been used to improve public finances. The panel estimations show that the budget balance improves only insignificantly as a reaction to a past increase of public debt since total revenues and expenditures rise. On the revenue side, primarily direct taxes are used to improve public finances. On the expenditure side, in particular public investment and subsidies are reduced to improve the budget balance. The presence of the interaction between fiscal rules and past debt slightly improves the sustainability of public finances. Estimations of fiscal reaction functions separately for each of the seven investigated countries find that only for Greece a significant improvement of the budget balance after an increase of the debt level be found, both with and without taking fiscal rules into account. 14

The empirical estimations in this paper have shown that fiscal policies of Austria, Germany, Italy, Spain, Portugal, Ireland and Greece cannot be regarded as totally sustainable. Even when fiscal rules and their interaction with the past debt level are taken into account, in many cases the primary budget balance does either react not at all or only insignificantly to past increases in the debt level. It is too early, i.e. there are too few data points available at present to test whether fiscal policies have strengthened recently in the aftermath of the severe European debt crisis. References Balassone, F., Franco, D. (2000): Assessing fiscal sustainability: A review of methods with a view to EMU. In Fiscal Sustainability, ed. Banca d Italia, 21 60. Rome, Bank of Italy. Baldi, G., Staehr, K. (2014): The European Debt Crisis and Fiscal Reactions in Europe 2000-2014. Paper presented at the Workshop Asymmetries in Europe: causes, consequences, remedies. Pescara, 27th-28th April 2015. Bohn, H. (1998): The behavior of U.S. public debt and deficits, Quarterly Journal of Economics, vol. 113, no. 3, 949-963. de Mello, L. (2008): Estimating a fiscal reaction function: the case of debt sustainability in Brazil, Applied Economics, vol. 40, no. 3, 271-284. Neck, R., Sturm, J.-E. (eds.) (2008): Sustainability of Public Debt. The MIT Press. Cambridge, MA. Piergallini, A., Postigliola, M. (2012): Fiscal policy and public debt dynamics in Italy, 1861-2009, CEIS Research Paper, no. 248, Tor Vergata University, CEIS. Staehr, K. (2008): Fiscal policies and business cycles in an enlarged euro area, Economic Systems, vol. 32, no. 1, 46-69. Wyplosz, C. (2006): European Monetary Union: the dark sides of a major success, Economic Policy, vol. 21, no. 46, 207-261. 15