136 3.5.4 Summary of results We found that governments are trying to increase their re-election chances by reducing the tax burden in the election year and even more in the year before the election. Governments are scheduling the implementation (and not so much the adoption) of tax reforms in an opportunistic way. These findings support our first two hypotheses on opportunistic government behavior. Furthermore, opportunistic behavior is strongest in the wage and income taxwhere tax reform effects are most directly visible to voters. Strong reductions in the wage and income tax are implemented especially in the year before the elections, while increases are concentrated in the two years after the elections. On the other hand we find only very limited evidence for opportunism in the less visible taxes. Only the mineral oil tax increases were scheduled after elections, while opportunistic behavior played no role with respect to the tobacco tax and the VAT. These findings support our third opportunism hypothesis. Altogether we therefore find strong support for all three opportunism hypotheses. 3.6 The partisan politics approach 3.6.1 General approach and related literature While all approaches discussed so far focus only on the re-election motives of governments, partisan approaches (inspired especially by the early work of Hibbs (1977)) argue that party governments and political parties can be characterized by distinctive ideological goals resulting from differences in their clienteles. Based on the partisan politics approach, left-wing governments are expected to pursue different economic policies than right-wing governments. The influence of partisan orientation on economic policy and economic outcomes is studied in the literature mainly with respect to monetary policy, government expenditures and deficit spending.239 The political business cycle theory240 focuses on partisan influence on macroeconomic outcomes via monetary policy and government expenditure.241 Another branch of literature restricts itself to analyzing how partisan preferences of governments are affecting public expenditures without discussing macroeconomic consequences.242 Finally following Cowart ?s 1978 study, 239 Applications of partisan theory to Germany can be found e.g. in Cusack (1995), Lehmbruch (2000), Renzsch (1998) or Schmidt (1996). 240 Inspired e.g. by Nordhaus (1975), Hibbs (1977), Tufte (1978), Paldam (1979) or Alesina, Roubini and Cohen (1997). 241 See for an overview Belke (1996). 242 See e.g. Swank (1988). 137 there exists a lot of work testing the hypothesis that leftist governments tolerate higher rates of government deficits than conservative governments.243 Although tax policy plays an important role in the political debate between different parties,244 the influence of partisan orientation on tax policy is seldomly incorporated in the analysis of tax policy.245 The relatively limited empirical work that analyzes the effects of partisan orientation on taxation consists mostly of cross-national studies. One of the first studies was pursued by Cameron (1978) who – although he was mainly interested in the influence of trade openness on government revenues – analyzed the influence of partisan orientation on tax revenues in a cross-national empirical sample. Garrett (1998) analyzed the effects of partisan orientation on total tax revenues and separately the effects on income taxes, consumption taxes and corporate taxes. Boix (1998) looked at the effect of leftist governments on average tax rates and the progressiveness of income tax systems. Most of the empirical findings support a partisan hypothesis. Cameron (1978) discovered a significant influence of partisan orientation on taxation: leftist governments increase tax revenues while right-wing governments have a tendency to cut back taxes. Garrett (1998) found a significant positive impact of left-wing governments on overall and income tax revenues but not on consumption or corporate income tax revenues. Boix (1998) found evidence for the hypothesis that left-wing-governments implement higher tax rates and a steeper progression in the personal income tax while conservative governments push down the tax rates especially for high incomes. For Germany most of the existing analyses of partisan politics in tax policy are restricted to case studies.246 Federal governments in Germany have always been dominated either by the Christian-democratic CDU/CSU or by the Social-democratic SPD. Most of the time, the CDU/CSU or the SPD governed together with a smaller coalition partner, the FDP or the GREENS. In the grand coalition (1966 to 1969) CDU/CSU and SPD governed together. All ministers of finance since 1964 (with the exception of Rolf Dahlgruen who was minister of finance from 1962 to 1966 and was a member of the FDP) have been associated to the CDU/CSU or the SPD. Therefore, we focus on the CDU/CSU and SPD for our partisan analyses and abstract from the smaller coalition partners. What are the interests of these two parties in tax policy and tax reforms? We discuss partisan preferences and derive hypotheses in the next part. 243 See Cowart (1978) or e.g. Roubini and Sachs (1989). 244 See e.g. Kaltenborn (1999). 245 See for further reference especially the discussion in the monograph of Hettich/Winer (1999) or OECD (2001). 246 See for example Ganghof (2004), Renzsch (2000) or Zohlnhoefer (1999). For some quantitative analyses of German tax reforms, which do not find an effect of partisan politics on tax reforms, see Wagschal (2005), pp. 411 ff. 138 3.6.2 Hypotheses on partisan interests and tax reforms in Germany Following partisan theory, the ideological goals of different political parties should reflect the interests of the party clienteles. How do the party clienteles of the CDU/CSU and the SPD differ?247 Data source: Membership statistics of CDU/CSU and SPD. Federal Statistical Office. 0.1 0.2 0.3 0.4 0.5 0.6 1950ies 1970ies 1980ies Pensioners and workers as share of total membership CDU/CSU SPD 0.0 0.1 0.2 0.3 0.4 1950ies 1970ies 1980ies Entrepreneurs as share of total membership CDU/CSU SPD Average net household income (2000): Pensioners € 20,700 Workers € 28,900 Entrepreneurs € 88,500 PARTY MEMBERSHIP, SOCIAL GROUPS AND INCOMES Figure 62: Clienteles of the two main German political parties Unfortunately, we do not dispose of detailed data on the income distribution of the party memberships. But if we compare membership data with respect to social groups (workers, pensioners, entrepreneurs) for the two parties (see Figure 62), we find that pensioners and workers (the two groups with low average incomes) are more important for the SPD than for the CDU/CSU. Additionally, the share of entrepreneurs, which have an income above average, is far stronger in the CDU/CSU.248 The share of white-collar workers has been similar in both parties in the period analyzed. 247 See for a general discussion of left/right preferences Garrett/Lange (1991) or Laver/Hunt (1992). 248 This has to be interpreted carefully as membership statistics do not allow to discriminate in between “normal entrepreneurs” and independent farmers (two groups with very different average incomes). 139 Based on these different clienteles we can derive two hypotheses on partisan interests with respect to tax reforms. Our first hypothesis is concerned with the general direction of tax reforms. We can argue that the poorer SPD clientele profits stronger from increases in government spending that are financed by taxes. For example the current federal budget, which is largely financed by taxes, assigns nearly 50% of all expenditure for social and work-related issues. Additionally, the lower-income clientele of the SPD contributes less to the financing of this expenditure especially via the progressive wage and income tax as the most important tax. Based on this general argument we should therefore expect higher tax burden increases in case of a SPD-led government than in case of a CDU/CSU-led government (partisan hypotheses I). 0 1 2 3 4 5 6 7 8 9 10 1. De cile 2. De cile 3. De cile 4. De cile 5. De cile 6. De cile 7. De cile 8. De cile 9. De cile 10 . D ec ile VAT payments as % of disposable household income 0% 10% 20% 30% 40% 50% 60% 70% 1. Quintile 2. Quintile 3. Quintile 4. Quintile 5. Quintile INCOME TAXES PROGRESSIVENESS OF MAIN TAXES Value-added tax Wage and income tax Sources: VAT data according to Bach (2005); Income tax progression: based on data from the Federal Statistical Office. Share of taxable income Share of tax payments in wage and income taxes Income deciles Income quintiles % Figure 63: Progressiveness of VAT and wage and income taxes However, we do not have to restrict ourselves to the overall development of the tax burden, but we can analyze partisan motivation with respect to the tax mix as well. In revenue terms the two most important taxes are the wage- and income tax and the VAT. Both taxes have contrary distributional consequences for different income groups (see Figure 63). The VAT is a regressive tax resulting from the higher share of consumption spending of lower-income households. This is only slightly moderated by the reduced tax rates of 7% e.g. for sales of food which are more important in low-income households. Studies reveal that households in the first income decile 140 face a total tax burden from the VAT of 9% of their total disposable income while households in the highest income decile face a burden of less than 6% of their total disposable income.249 In wage and income taxes the picture looks quite different. The tax revenue statistics show that many of the low-income households do not pay any wage or income taxes. Of all households with positive tax liabilities the 20% with the lowest taxable incomes contribute only 0.05% to the total tax revenues from the wage and income tax but dispose of 2.2% of total taxable income (2001 data).250 The 20% households with the highest taxable income dispose of 50.4% of taxable income but contribute 69.5% to total wage and income tax revenues. This reflects the progressiveness of the wage and income tax payments. If we combine the fact that the lower-income groups are more important for the SPD than for the CDU/CSU with the different progressiveness of VAT and the wage and income tax, we can derive our second partisan hypotheses: left-wing governments should increase taxation via tax reforms especially in the wage and income tax, while rightwing (CDU/CSU led) governments should shift the tax burden to the VAT (partisan hypothesis II). With respect to the other two important indirect taxes, the mineral oil and the tobacco tax, we could make a similar argument although the distributional consequences are not so clear-cut.251 3.6.3 Empirical analysis Based on our data-set we can directly test our two partisan hypotheses. We have compiled the data on the overall fiscal effects of tax reforms in Figure 64. The figure covers a total of 39 years (1965 to 2003). During three years a grand coalition of SPD and CDU/CSU governed. We can split the remaining 36 years evenly into 18 years of right-wing (CDU/CSU) government and 18 years of left-wing (SPD) governments.252 What do we see with respect to our two partisan hypotheses? First, we find that the annual average tax burden increases were slightly higher under the left-wing governments (compare the first box in Figure 64), but reductions were higher as well. Therefore, the overall net fiscal effect of tax reforms was an even higher reduction of the tax burden under the left-wing governments than under the right-wing governments, contradicting our first hypothesis on partisan politics in tax reforms (see the balance of reforms reflected by the line in Figure 64). During the grand coalition increases and reductions were both lower than under other governments 249 See Bach (2005) or Jacobebbinghaus (2003). 250 An even better indicator to determine the progressiveness of the wage and income tax would be market income (before reductions based on tax laws). However, this is not included in German tax revenue statistics. 251 For a discussion with respect to the progressiveness of the mineral oil tax see Poterba (1991). 252 The effect on the majority constellation in the second chamber, the Bundesrat, on partisan politics will be discussed in the following part on divided government. 141 and here the average net fiscal effects of tax reforms were positive (leading to an increase of the tax burden). TAX REFORMS BY PARTISAN CONSTELLATION DATA OF ADOPTED REFORMS - 1965-2004 -0,02% 0,00% 0,02% 0,04% 0,06% 0,08% AV ER AG E LE FT RI G HT G RA ND CO AL IT IO N INCR/GDP RED/GDP BALANCE/GDP -0,80% -0,60% -0,40% -0,20% 0,00% 0,20% 0,40% 0,60% AV ER AG E LE FT RIG HT GR AN D C OA LIT ION INCR/GDP RED/GDP BALANCE/GDP REFORMS (TOTAL) -0,60% -0,50% -0,40% -0,30% -0,20% -0,10% 0,00% 0,10% 0,20% 0,30% AV ER AG E LE FT RIG HT GR AN D C OA LIT ION INCR/GDP RED/GDP BALANCE/GDP REFORMS (WAGE AND INCOME) -0,15% -0,10% -0,05% 0,00% 0,05% 0,10% 0,15% AV ER AG E LE FT RI G HT G RA ND CO AL IT IO N INCR/GDP RED/GDP BALANCE/GDP -0,01% 0,00% 0,01% 0,02% 0,03% 0,04% AV ER AG E LE FT RI G HT G RA ND CO AL IT IO N INCR/GDP RED/GDP BALANCE/GDP REFORMS (VAT) REFORMS (TOBACCO) REFORMS (MINERAL OIL) Fiscal effect of reforms/GDP Fiscal effect of reforms/GDP Fiscal effect of reforms/GDP Fiscal effect of reforms/GDP Fiscal effect of reforms/GDP Own calculations based on: Federal Ministry of Finance (2004), Federal Statistical Office(2007). INCR=increases; RED=reductions; BALANCE=Increases-reductions Figure 64: Fiscal effects of tax reforms by partisan constellation How does the reform pattern look like if we analyze tax reforms by kind of tax? With respect to the wage and income tax we see only very limited activity during the time of the grand coalition. Left-wing governments increased the tax burden in the wage and income tax less strongly than right-wing governments, but the average annual tax burden reductions were stronger than under the right-wing governments. This led to a stronger net-relief in the progressive wage and income taxes under leftwing governments. With respect to the VAT, where reductions play normally no role, the left-wing governments increased the tax burden on average stronger than the right-wing governments. Here the grand coalition was more active than the leftand the right-wing governments. This resulted partly from the fundamental transformation of the sales taxes in a value-added tax and the following rate increase in 1968. Both reform patterns strongly contradict our second partisan hypothesis on tax reforms. Left-wing governments decreased the tax burden in the wage and income tax more strongly than right-wing governments and increased the tax burden more 142 strongly in the VAT.253 A look at the reform patterns in the mineral oil and tobacco taxes reveals a similar finding. Especially in the tobacco tax, left-wing governments acted contrary to our partisan hypothesis and increased the tax burden stronger than right-wing governments. However, one could argue that our data was strongly affected (and for this analysis distorted) by reunification in 1990 which created huge revenue needs. As the CDU/CSU and FDP coalition governed from 1982 to 1998, the data for right-wing governments were directly affected. To finance reunification, the introduction and increase of the solidarity surcharge on the wage and income tax were the most important tax reform measures.254 To control for these effects we have therefore recalculated the fiscal effects of tax reforms by partisan constellation excluding the introduction of the solidarity surcharge and its following increases. TAX REFORMS BY PARTISAN CONSTELLATION DATA OF ADOPTED REFORMS - 1965-2004 -0,02% 0,00% 0,02% 0,04% 0,06% 0,08% AV ER AG E LE FT RI G HT G RA ND CO AL IT IO N INCR/GDP RED/GDP BALANCE/GDP -0,80% -0,60% -0,40% -0,20% 0,00% 0,20% 0,40% 0,60% AV ER AG E LE FT RIG HT GR AN D C OA LIT ION INCR/GDP RED/GDP BALANCE/GDP REFORMS (TOTAL) -0,60% -0,50% -0,40% -0,30% -0,20% -0,10% 0,00% 0,10% 0,20% 0,30% AV ER AG E LE FT RIG HT GR AN D C OA LIT ION INCR/GDP RED/GDP BALANCE/GDP REFORMS (WAGE AND INCOME) -0,15% -0,10% -0,05% 0,00% 0,05% 0,10% 0,15% AV ER AG E LE FT RI G HT G RA ND CO AL IT IO N INCR/GDP RED/GDP BALANCE/GDP -0,01% 0,00% 0,01% 0,02% 0,03% 0,04% AV ER AG E LE FT RI G HT G RA ND CO AL IT IO N INCR/GDP RED/GDP BALANCE/GDP REFORMS (VAT) REFORMS (TOBACCO) REFORMS (MINERAL OIL) EXCLUDING SOLIDARITY SURCHARGEI I I Fiscal effect of reforms/GDP Fiscal effect of reforms/GDP Fiscal effect of reforms/GDP Fiscal effect of reforms/GDP Fiscal effect of reforms/GDP Own calculations based on: Federal Ministry of Finance (2004), Federal Statistical Office(2007). INCR=increases; RED=reductions; BALANCE=Increases-reductions Figure 65: Fiscal effects of reforms by partisan constellation (adjusted for reunification effects) 253 Without the important reductions of wage and income taxes, which were triggered by an increase of the tax exemption level demanded by the constitutional court under the right-wing government from 1982 to 1998, the effect would have been even more pronounced. 254 See for a discussion of the influence of the solidarity surcharge on wage and income taxes part IV.2.1. We abstract from the solidarity surcharge on the corporate profit here as it was much less important in revenue terms. 143 Figure 65 shows the adjusted results. We see that the exclusion of the solidarity surcharge lowers the average tax burden increases in the wage and income tax under right-wing governments. However, the exclusion of the solidarity surcharge does not change our general results. There is again no support for our partisan politics hypothesis. To make sure that this is not due to other distortive effects of reunification we have as well re-calculated reform effects excluding the years from 1991 to 1993. Even this did not change our results substantially. While the reform effects in the indirect taxes are comparatively clear, one could argue that the reforms in the wage and income tax are strongly influenced by a correction for cold progression resulting from inflation. Inflation was especially high under the left-wing governments from 1969 to 1982. However, there is little evidence for this argument. Generally, the data do not indicate that tax burden reductions in the wage and income tax were stronger in times of high inflation.255 Furthermore, we calculated the average annual reductions in the wage and income tax under three different left-wing governments from 1969 to 1982 and related them to the average inflation rates in these three legislative periods. Even under the same partisan government we did not find a correlation of tax burden reductions and inflation.256 Our data show that the tax burden reductions were strongest in the period from 1976 to 1980, although inflation was highest from 1969 to 1972. Therefore, we think that a critique of our results based on an argument of cold progression is not valid in this context. 3.6.4 Summary of results Taken together we find no evidence for our two partisan hypotheses. Left-wing governments did not increase the tax burden stronger than right-wing governments. Contrary to our hypotheses of partisan interests, left-wing (SPD) governments reduced the tax burden stronger in the wage and income tax than right-wing (CDU/CSU) governments. Furthermore, the tax burden increases in the regressive indirect taxes were larger under left-wing governments. Our results are stable even if we exclude the direct influence of reunification in form of the solidarity surcharge and even if we exclude the years after reunification completely (from 1991 to 1993). Furthermore, we found that our results are not distorted by tax policy reacting to cold progression. Therefore, we conclude that partisan interests are not reflected in the tax reform patterns of our data-set. Our findings are in line with other quantitative studies of tax reforms in Germany,257 but they contradict the findings of several qualitative studies for Germany and several international comparative studies258 which mostly rely on revenue data. 255 See discussion of the role of cold progression in part V.2. 256 Measured by the GDP deflator. 257 See Wagschal (2005), pp. 411 ff. 258 See our discussion in part V.3.6.1. 144 However, we argue that the data employed here are far better suited for an analysis of partisan interests than just case studies or revenue data and that our results are therefore more reliable. 3.7 The role of divided government: partisan governments in a bicameral legislature 3.7.1 General approach and related literature The German federal government is not unrestricted in tax policy.259 Approval of tax reforms in the second chamber, the Bundesrat, is necessary in all major taxes with exception of mineral oil taxes and the tobacco taxes. Therefore, the Bundesrat has a far-reaching ability to veto tax bills. As the federal government and the state governments are dominated by the two major parties (the CDU/CSU and the SPD), the majority constellations in the two chambers can either lead to divided (different partisan majorities in both chambers) or undivided government (majority of the parties of the federal government in the Bundesrat).260 If party majorities are decisive for policy-making,261 the importance of the second chamber as a veto-player would therefore depend on whether we are in a situation of divided or undivided government. If the government parties do not have a majority in the second chamber, the Bundesrat, the opposition parties could use the veto-player position of the Bundesrat to block legislation. This has led to a “gridlock hypothesis” in case of divided government, which is very prominent in the literature on German federalism.262 In the empirical literature based on legislative statistics, which uses indicators as e.g. the number of effectively vetoed legislative proposals or the prolongation of legislation in case of divided government, there is so far only limited evidence for the gridlock thesis.263 259 See the discussion in part V.3.1. 260 See for a general discussion of bicameralism and divided government Persson/Roland/Tabellini (1997), Riker (1992), Rogers (1998), Tsebelis (1995), Tsebelis/Money (1997), Alesina/Rosenthal (1996), Brennan/Hamlin (1992) and Diermeir/Myerson (1999). For the differences in between effects of bicameralism and of divided government see Elgie (2001). 261 Please note that the importance of party majorities is not directly linked to partisan interests discussed under the partisan politics approach. Party majorities can matter even if parties do not adhere to distinctive ideological goals. For example, policy proposals might be blocked not based on their content, but because they are brought in by the opposition. 262 See e.g. Scharpf (1985) or Braeuninger/Koenig (1999). See for the gridlock thesis in tax policy in Germany especially Ganghof (1999) and Wagschal (2005), but also Renzsch (2000) and Zohlnhoefer (1999). With respect to international comparisons of the effects of divided government/bicameralism refer to Hallerberg/Basinger (1999). 263 See Burkhart/Manow (2006), Bauer (1998), Lehmbruch (2000) and Renzsch (2000). In a recent contribution Burkhart and Manow (2006) employ an “auto limitation process” (based on Vanberg (1998)): They argue that federal governments refrain from confrontation with the