How does globalization affect the implicit tax rates on labor income, capital income, and consumption in the European Union?

Onaran, Ozlem ; Boesch, Valerie ; Leibrecht, Markus 等

I. INTRODUCTION

This article analyzes the effects of globalization on the taxation of labor income, capital income, and consumption in the EU Member States. The theory of capital tax competition argues that, as capital becomes increasingly more mobile, firms are able to avoid high taxes by choosing countries with a low capital tax burden, which may result in inefficiently low taxes on capital income and inefficiently low public good provision (Brueckner 2000; Bucovetsky 1991; Krogstrup 2004; Oates 1972; Tanzi 1995; Wilson 1999; Zodrow and Mieszkowski 1986). However, fewer possibilities to tax mobile capital also implies that more immobile tax bases, notably labor income and consumption, should bear the tax burden necessary to finance a given level of public expenditures.

Indeed, an already extensive empirical literature is available which explores the impact of globalization on the level of the capital income tax burden as well as the tax burden on more immobile tax bases. In this literature, five different measures of tax burden are used: (1) statutory tax rates (STRs) on corporate income, (2) corporate income tax revenues as a percentage of GDP (gross domestic product) or total tax revenues, (3) implicit tax rates (ITR) (1) on capital income, corporate income, labor income, and on consumption expenditures of the Mendoza, Razin, and Tesar (1994) type, and (4 and 5) effective marginal and average tax rates on corporate income (EMTR and EATR) of the Devereux and Griffith (1998) type.

Studies that test the effect of globalization on the level of STR or EATR (see Clausing 2007; Slemrod 2004; Swank and Steinmo 2002 on STR, and Dreher 2006a; Garretsen and Peeters 2007; Krogstrup 2005; Loretz 2008 on EATR) generally find a negative effect. Yet, the empirical results on the effects on the ITR on capital income are inconclusive: Dreher, Gaston, and Martens (2008), Swank and Steinmo (2002), and Swank (2006) find no effect; Dreher (2006a) estimates a positive effect, whereas Winner (2005) finds a negative effect. Using the ITR on corporate income, Adam and Kammas (2007), Bretschger (2008), and Bretschger and Hettich (2002) also find a negative effect of globalization. Moreover, Quinn (1997) finds a positive effect on corporate tax revenues as a percentage of GDP and total revenues.

Concerning the ITR on labor income, Adam and Kammas (2007), Winner (2005), and Dreher, Gaston, and Martens (2008) find a positive effect of globalization. However, Dreher (2006a) finds no effect, whereas Swank and Steinmo (2002) find a negative effect. Furthermore, Bretschger and Hettich (2002) explore the globalization impact on the ratio of the ITR on labor income to that on corporate income. They find a positive relationship which signals that the tax burden is shifted to labor income. Similarly Winner (2005) and Adam and Kammas (2007) find a shift from capital to labor income taxation. Regarding the globalization effects on the ITR on consumption, the picture is more conclusive as most available studies find no effect (e.g., Dreher 2006a; Dreher, Gaston, and Martens 2008; Swank and Steinmo 2002).

Concerning tax law-based tax measures (EA TR, EMTR, STR), the empirical evidence available so far points toward a negative relationship with globalization. Yet, notable features of the existing literature are that these studies (1) are predominantly based on a sample of advanced Organization for Economic Cooperation and Development (OECD) countries and (2) do not separate the globalization effects by welfare regimes or country groups.

This article adds to the literature by exploring the differences in globalization effects across welfare regimes within the EU15 based on an "augmented" Esping-Andersen typology. We also distinguish between countries that have tax rates higher than average compared to those with a lower tax rate. (2) A second novelty of this article lies in including the Central and Eastern European New Eastern European Members States (CEE NMS) in the country sample and, thus, by focusing on differences in globalization effects between western countries (EU15) and CEE NMS.

With respect to the separation by welfare regimes, Campbell (2005) suggests that national political and economic institutions mediate how states react to the pressures of globalization. Thus, the institutional environment limits the degree to which national tax regimes converge in response to globalization. Specifically, the institutional configuration of national politics shapes actors' political tax policy strategies. This is in line with the argument that welfare regimes display path-dependency, and only change gradually within specified paths (Esping-Andersen 1996; Scharpf and Schmidt 2000; Swank 2001). Different welfare states create different expectations and dependency relations among the citizens, which cannot be changed quickly given electoral considerations (Kautto and Kvist 2002). This also affects tax policy. For instance, Campbell (2005) argues that the type of labor and business organizations affects the sorts of tax policies which these organizations are willing to support. If labor is politically influential, unions support relatively high taxes because they expect this to finance expenditures like social protection. Indeed, Campbell (2005) finds that social-democratic countries set the highest tax rates on both labor and capital income, as they utilize the tax revenue to finance welfare spending and the tax burden is lowest in the liberal welfare state.

Although most of the CEE countries reformed their tax systems along western lines during the transition period (Campbell 2005), analyzing the CEE NMS in isolation is meaningful for several reasons: first, CEE NMS governments have been especially active in using cuts in effective corporate income tax rates (e.g., Bellak and Leibrecht 2009), introducing flat rate personal income taxes (e.g., Keen, Kim, and Varsano 2008), and creating special economic zones (e.g., World Bank 2008) to attract foreign capital (especially Foreign Direct Investment, FDI). Second, the tax reforms have probably led to a profit-shifting response of multinational enterprises, which have tried to make use of the lower statutory corporate tax rates in the transition countries (Devereux 2007). Third, the transition crisis has posed quite extensive fiscal needs in terms of increased unemployment and early pension schemes, which may have shaped taxation decisions (Havlik and Landesmann 2005). Fourth, the tax structure is quite different in the CEE NMS, for example the share of indirect taxes in tax revenues is on average much higher than in the EU15 (e.g., EU Commission 2009: 58). Fifth, the extent of the informal economy, which is much higher than in the EU15, may have increased the preference of the governments for lower tax rates on income to create incentives for formalization as well as a higher share of taxes on consumption (Duman 2009).

On the basis of a panel data set, we find supportive evidence for an increase in the ITR on labor income in the EU15. We do not find a significant effect of globalization on the ITR on capital income. There is some evidence of convergence in terms of the ITR on consumption, as countries with higher than average ITR on consumption respond to globalization by decreasing their tax rates. However, there are important differences among the welfare regimes within the EU15. Social-democratic countries have decreased the tax burden on capital, but increased that on labor due to globalization. Globalization exerts a pressure to increase taxes on labor income in the conservative and liberal regimes as well. Taxes on consumption decrease in response to globalization in the conservative and social-democratic regimes. In the CEE NMS, there is no effect of globalization on the ITR on labor and capital income but we find a negative impact on the ITR on consumption. Furthermore, this again points at convergence as the CEE NMS with higher than average ITR on consumption decrease their tax rates in response to economic globalization.

This article is structured as follows: Section II reviews the literature on welfare regime typologies. Section III describes how we measure tax burdens and the globalization process. It also includes some empirical, stylized facts. Section IV introduces the control variables used in the empirical analysis. Section V presents the estimation methodology and Section VI discusses the results. Section VII concludes.

II. GROUPING OF COUNTRIES INTO WELFARE REGIMES

The welfare state literature indicates considerable heterogeneity among the Western European countries related to the institutional setting of a country. In the literature about the effects of globalization on the government budget, Leibrecht, Klien, and Onaran (forthcoming) is the only article to the best of our knowledge that directly addresses the relevance of welfare regimes; however, this article only discusses the effects on social spending.

Esping-Andersen (1990) developed a widely used typology of three welfare regimes, grouping countries based on their stratification, decommodification, and the mix between private and public social security institutions. The first group consists of social-democratic regimes which are universalistic and egalitarian with high degrees of decommodification, little stratification, and social security payments provided universally by the state (Sweden, Finland, Denmark, Norway). The second comprises conservative regimes strongly associated with employment protection, with the family at the heart of the protection, a medium decommodification, and social security provided partly by the state and partly by the market (Germany, France, Austria, Belgium, Italy, Japan, Switzerland, and the Netherlands). Finally, the third group encompasses liberal regimes with low decommodification, high stratification, restricted role of the state, a low level of social security, and a significant private insurance contribution (United Kingdom, United States, Ireland, Canada, and Australia). This classification is later extended by adding a separate welfare regime group for the southern European countries (Italy, Spain, Greece, and Portugal) by Ferrara (1996) and Bonoli (1997). As opposed to previous research which treats countries such as Spain, Portugal, and Greece as latecomers on the same path of continental conservative welfare states, Ferrara (1996) argues that southern countries are inter alia characterized by a highly fragmented and polarized welfare regime with generous pensions paired with substantial gaps in the social safety net, a departure from the corporatist tradition in the field of health care, a highly collusive mix between public and private institutions in the welfare sphere and the persistence of clientelism in the distribution of cash subsidies.

Owing to its wide use in the literature, we use Esping-Andersen's classification. Although these welfare regimes are developed based on social expenditure structures, they reflect institutional and political structures that might determine the influence of social actors on tax policy. Furthermore, expenditure structures create a path-dependency that also locks in tax regimes (Esping-Andersen 1996; Scharpf and Schmidt 2000; Swank 2001). As discussed in Section III on the stylized facts of the tax rates on capital and labor income and consumption, the tax structures of welfare regimes can be quite different.

Although some studies see welfare states in the CEE NMS within the liberal regime based on a mix of social insurance and social assistance, and a partial privatization of social policy with just a few corporatist attributes (e.g., Ferge 2001, Standing 1996), others argue that the CEE NMS constitute a separate post-socialist regime type (Aidukaite 2004; Lelkes 2000). Noelke and Vliegenthart (2009) argue that the CEE NMS form a dependent market economy model that differs from coordinated or liberal market economies due to its dependency on foreign capital. Fenger (2007) distinguishes a "post-communist European type" and a "former USSR type," where the former mixes characteristics of both the conservative and the social-democratic types. Bohle and Greskovits (2007) distinguish between a neoliberal type in the Baltic States, an embedded neoliberal type in the Visegrad states, and a neo-corporatist type in Slovenia. Alternatively, Orenstein and Haas (2005) distinguish between European and Eurasian post-communist welfare states where the European category includes all CEE NMS as well as other former Yugoslav republics. According to Orenstein and Haas (2005), good prospects of joining the EU pushed for the development of welfare states in the CEE NMS, and they therefore find less of a difference between the Baltic countries and the other countries within the CEE NMS.

These studies suggest that the countries in the CEE NMS constitute a welfare regime in transition different from those found in the EU15. Therefore, we estimate the effects of globalization on taxes separately for the EUI5 and the CEE NMS. However, further tests of diversity among the CEE NMS are not possible due to limited degrees of freedom.

III. MEASURING TAX BURDEN AND GLOBALIZATION

A. Measuring the Tax Burden on Capital Income, Labor Income and Consumption

As mentioned, different types of tax rates are used as dependent variables in empirical studies. STRs on corporate income are taken directly from the tax code; however, they do not account for the changes in the tax base. EMTR and EATR on corporate income are calculated based on the neoclassical investment theory and on actual and future tax law data. They measure the tax burden on a hypothetical investment project (Devereux and Griffith 1998). (3) These rates are thus forward looking rates.

ITRs are calculated by dividing the total tax revenue from capital income, labor income, or consumption by the pre-tax income of the respective production factor or consumption (e.g., corporate income tax revenues divided by gross operating surplus). (4) These rates are backward looking rates and are available not only for corporate income, but also for capital, labor income, and consumption. Therefore ITRs are especially suitable for exploring whether globalization has led to a decline in the tax burden on capital and an increase in the tax burden on labor and/or consumption within a unified framework.

For this reason, we base our analysis mainly on ITRs. For the ITRs, we use Eurostat data (see European Commission 2009). The advantage of the Eurostat data set is that it is the first to cover all 27 EU member states, including the CEE NMS. Two data sources are used for the ITRs: the data for the period starting in 1995 is taken from the Eurostat database. For the period prior to 1995 the European Commission provides the data in its publication (European Commission 2000). The growth rate of this data is used to extend the Eurostat data backwards from 1995 to 1970 or 1980. (5) The time period for most CEE NMS ranges from 1995 to 2007. The data on the ITR on capital in Romania, Bulgaria, and Slovenia, which only begins in the late 1990s, is further extended backwards to 1995 with own calculations, based on the method used by the European Commission (2000). (6) In summary, the data reaches back to 1970 for nine EU15 countries, to 1980 for six EU15 countries, and to 1995 for most CEE NMS (Table A1).

Figure 1A-C shows the development of the ITR on capital income, consumption, and labor income in the EU15 grouped by welfare state regimes and the CEE NMS (GDP-weighted average).

The ITR on capital income stayed rather stable at 30.4% (weighted sample mean) due to a broadening of the tax base, particularly in the case of corporate income (rising profits, legal changes regarding deductions, allowances, etc.; see Devereux, Griffith, and Klemm 2002; European Commission 2008). Regarding the different welfare regimes, since the 1970s the ITR on capital income has fallen in the liberal welfare regime from 50.6% to 40.6% (weighted regime mean) and risen in the social-democratic regime from 22.6% to 36.5% with stronger upturns and downturns. In the southern regime, the ITR on capital income has developed similar to that of the social-democratic regime rising from 24.4% to 35.0%. In the conservative regime, it decreased from 33.3% to 29.7%, apart from a slight increase since the mid-1990s. Within the conservative regime, France is a particular exception with a continuously increasing ITR on capital. In the CEE NMS, the mean of the ITR on capital income is much lower than in the EU15 (19.7% compared to 30.6%). The highest rate is for the Czech Republic with 25.9% and it is as low as 10% in the Baltic countries. The average in the CEE NMS decreased until 2000 and then slightly increased.

However, there are also important differences in the trends among the CEE NMS. Although in Slovenia, the ITR on capital has risen from a very low level, it has decreased in Latvia and Estonia, and it plummeted from a higher level in Slovakia while remaining quite stable in most other CEE NMS.

The ITRs on labor are on average higher than those on capital, with a weighted mean of 33.4%. The development has been more homogenous among the different welfare regimes in the EU15 with an overall increasing trend in most countries compared to the early 1970s. Countries in the social-democratic welfare regime have the highest ITR on labor income which has risen until the turn of the millennium and has decreased since then. Overall the rate increased from 37.9% to 40.8%. The lowest ITR on labor income is levied by the liberal regime with a mean of only 25.1%. It rose until the late 1980s, since which it has decreased slightly. The ITR on labor income in countries of the conservative regime lies between those of the social-democratic and the liberal regime, and has constantly risen from 33.7% to 39.5% almost converging with that of the social-democratic regime eventually. The ITR on labor income in the southern regime has also constantly risen from 24.9% to 38.2%, although it is still slightly lower than that of the conservative and social-democratic welfare regimes. The ITR on labor income in the CEE NMS is higher than in the EU15, although the difference is relatively small (36.5% to 33.4%), with Hungary, the Czech Republic, and Slovenia showing the highest rates. On average, the ITR on labor in the CEE NMS remained quite stable or declined slightly. Combining this with the lower ITR on capital, it can be argued that the CEE NMS rely more heavily on labor taxes.

With respect to the ITR on consumption, the social-democratic countries once again show the highest ITR by far. It rose in the late 1980s, decreased in the early 1990s and since then has stayed more or less constant. Over the whole period, it increased from 27.6% to 29.3%. In the liberal regime, the ITR on consumption increased during the late 1970s and early 1980s, and then stayed rather constant in a range of 19.0%-20.0%, as in the conservative regime. Again in southern countries, the ITR on consumption is lower than in the EU15 with a mean of 15.4%. However, it has been constantly rising, with a particularly strong increase in the late 1980s and early 1990s. The level of the ITR on consumption in the CEE NMS corresponds to the level in the conservative and liberal regime.

Overall, the descriptive analysis of the ITRs underlines the relevance of estimating different effects of globalization among welfare regimes and country groups (West vs. East).

B. Measuring the Globalization Process

Globalization is a multifaceted phenomenon comprising economic, social, institutional, and political aspects (Dreher 2006b; Dreher, Gaston, and Martens 2008). In the empirical literature, globalization is frequently measured by a country's openness to trade or FDI (see Gemmel, Kneller, and Sanz 2008 for a review). However, using either trade or FDI to measure globalization, thus excluding other flows of income and capital or changes in de jure measures or social and political dimensions of globalization might result in biased estimates (Dreher, Gaston, and Martens 2008). The estimations by Leibrecht, Klien, and Onaran (forthcoming) about the effects of globalization on social spending might be significantly misinterpreted if narrow measures are used. Following Leibrecht, Klien, and Onaran (forthcoming), we therefore base our analysis on the Konjunkturforschungsstelle (KOF; Swiss Economic Institute) globalization indices, developed by Dreher (2006b) and Dreher, Gaston, and Martens (2008), which incorporate these different dimensions. These are weighted indices of various globalization variables, where the weights are determined via principal component analysis. Each variable entering the KOF measure is transformed to an index on a scale from 1 to 100, where 100 is the maximum value of the variable in the period 1970 to 2006. The data is transformed according to the percentiles of the original distribution. For the current analysis, the KOF indices have the additional advantage of also being available for the CEE NMS.

[FIGURE 1 OMITTED]

We use two different KOF indices, one capturing economic globalization and one also considering social, political, and institutional aspects of the globalization phenomenon: (a) KOFecon which incorporates actual FDI, income and trade flows, as well as legal restrictions on FDI and trade. The "flow-part" of the index brings together FDI stock and flows, exports and imports, portfolio investments (stock of assets and liabilities), and income payments to foreign nationals, all normalized by GDP. The "restrictions-part" of the index includes de jure measures of formal openness such as hidden import barriers, mean tariff rates, taxes on international trade as a percentage of current revenue and capital account restriction; (b) KOFglobal which combines economic globalization with social and political globalization, incorporating the number of embassies and high commissions in the country, the number of international organizations of which the country is a member, the number of international treaties signed, personal contacts, information flows, and cultural proximity (Dreher 2006b). The EU countries have gone through two types of global integration: at the global level and the European level. Although KOF globalization indices capture the effects of overall global integration, they do not distinguish the effects of the intra-EU integration. It is not possible to construct a new index for the intra-EU integration within the scope of this article; however, in order to reflect the relevance of this EU dimension, and test for convergence effects, we also introduce a dummy variable in the empirical estimation for countries with tax rates higher than EU average. Furthermore, we explore the robustness of our results by testing the effects of intra-EU trade and FDI and extra-EU trade and FDI separately. Note, however, that it is not possible to get bilateral data for the other components of the KOF economic globalization indices; therefore the results for trade and FDI effects are not strictly comparable to the effects of the broader KOF indices.

IV. CONTROL VARIABLES

In addition to the globalization variables, further explanatory variables capturing mainly domestic determinants of tax rates are included in the estimations as control variables. The choice of control variables is based on the related previous empirical literature (e.g., Adam and Kammas 2007; Bretschger and Hettich 2002; Dreher 2006a; Swank and Steinmo 2002; Winner 2005).

Total expenditures of general government as a ratio to GDP (variable expenditures) should ceteris paribus be positively related to ITRs as they induce higher financial needs.

The general government consolidated gross debt as percent of GDP (debt) can have either a positive or negative effect on ITRs. On the one hand, public debt can serve as substitute for taxes: when taxes are lowered, expenditures have to be financed by debt. On the other hand, there exists a borrowing effect: the higher the public debt, the more taxes have to be levied to pay for the debt. Thus, the expected sign of this variable is ambiguous a priori.

The population older than 65 years as a share of total population (oldage) is expected to be positively correlated with ITRs: a higher share of dependent population results in higher fiscal needs.

The growth rate of real GDP (growth) aims to capture cyclical effects and is expected to have a negative effect on labor and capital tax rates. Based on a tax competition model with a balanced budget, rational governments will lower tax rates when growth is high as low economic growth leads to a lower interest rate and capital exports (see Adam and Kammas 2007; Bretschger and Hettich 2002). However, alternatively, governments may engage in countercyclical tax policy, thus lower the tax rates when growth is low. Thus a priori the sign of growth is ambiguous.

Inflation, measured as the change in the GDP deflator (inflation), is expected to affect taxes through different channels. If tax law contains an amount expressed in nominal values (e.g., levels of tax brackets with progressive taxation or an amount of personal deduction for the income tax), inflation might affect tax revenues positively and negatively (Thuronyi 1996): First, taxpayers are pushed into higher labor income tax brackets and tax revenues rise at a higher rate than the tax base. The same may happen in the case of proportional business taxes if depreciation allowances are based on historical values. Second, due to collection lags in tax administration, inflation may lead to a decrease in ITRs as the tax base increases at a higher rate than tax revenues. Third, if the tax base is measured in non-indexed nominal values, as is sometimes the case with property taxes or when (consumption) tax revenues are based on "specific taxes," inflation may cause an erosion of tax revenues and a decrease in the ITR. Thus the sign of the inflation coefficient is ambiguous a priori.

Government Party (gov_party) is an ordinal variable ranging from 1 to 5 that controls for the partisan effect based on leftist and non-leftist parties' shares of cabinet seats (1 = hegemony of fight-wing [and center] parties, i.e., 100% cabinet seats share of non-leftist party or parties; 2 = more than two-thirds and less than 100% of cabinet seats held by non-leftist party; 3 = stalemate, i.e., left and non-left: more than one-third and less than two-thirds of cabinet seats; 4 = social-democratic dominance, i.e., leftist parties hold more than two-thirds and less than 100% of cabinet seats; 5 = social-democratic hegemony, i.e., 100% cabinet seats held by leftist parties). Its calculation is following Schmidt (1982, 1992) and is provided by Potolidis, Gerber, and Leimgruber (2010) and Armingeon et al. (2010), who reproduce the Schmidt index. (7) The tax rates are expected to be lower, the more right-wing the governing political parties tend to be, assuming that they would advocate a more neoliberal economic policy stance with tight fiscal policy and lower public expenditures, as well as lower taxes to stimulate business and increase labor supply. Hence, we expect a positively signed coefficient.

As smaller countries are typically more open than larger countries, a country's relative size (size) is included in the set of regressors following Winner (2005) in order to cope with a possible small country bias. This variable is measured as the proportion of a country's GDP to the average sample GDP. We expect a positive sign on the coefficient of this variable. For the ITR on capital income, this directly follows from the literature on asymmetric tax competition (Bucovetsky 1991). However, larger countries provide mobile firms, workers and consumers with taxable agglomeration rents inter alia due to lower trade costs, higher real wages, and more product varieties (e.g., Brakman, Garretsen, and yon Marreijk 2009, chapter 11 for mobile capital). Thus, we expect that a larger country size is also positively related with the ITR on labor and consumption.

Tables A1 to A6 in the appendix contain information on the measurement of the variables, the databases used, and descriptive statistics.

V. ESTIMATION METHODOLOGY

We explore the effect of globalization for the EU15 countries on the various ITRs by using the baseline model shown in Equation (1):

(1) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where j is the capital, labor, or consumption. The country index i ranges from 1 to 15 for the EU15; t is the time index ranging from 1970 to 2007. [ITR.sub.ji(t-1)] represents the lagged-dependent variable. [G.sub.it-1] stands for the 1-year lagged globalization indicator. [C.sub.it-1] and [M.sub.it] are the matrices capturing the 1-year lagged and the contemporaneous control variables as further detailed below. [[alpha].sub.ji] captures country-fixed effects, [[omega].sub.jt] captures time fixed effects, and [[epsilon].sub.jit] is the remainder error term.

The second step is to investigate if there are differences of the effect of globalization between countries with a high and low ITR, respectively.

(2) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].

In Equation (2) a dummy variable [H.sub.jit] is interacted with the lagged globalization indicator, which is 1 if the ITR of the respective country is higher than the mean ITR of the EU15 countries in time t. For most countries [H.sub.jit] is either 1 or 0 for all years; but there are also intermediate country cases where the dummy changes value for some years. In order to avoid an ad hoc classification of these intermediate countries as either high- or low-tax countries for all years, we opted for using a time variant dummy. Thus for most cases, [H.sub.jit] captures a structural effect for high- versus low-tax countries. In the intermediate cases where a country changes position, for example, a country that had traditionally lower than average ITR, might find itself with a higher than average ITR in 1 year when several other countries decrease taxes, and they may try to respond to this next year. Since [H.sub.jit] is time variant, we also added the intercept dummy itself despite the presence of country-fixed effects (see, e.g., Brambor, Clark, and Golder 2006). Furthermore a trend [H.sub.ji] x T is introduced, which is specific to the countries with higher than average ITR.

As described in Section II, we further test for the heterogeneity of the globalization effects in different Western European welfare state regimes, for example the social-democratic, conservative, liberal, and southern regime. Therefore, the following equation is estimated:

(3) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where [D.sub.k] is a dummy variable representing the different welfare regimes. [D.sub.1] stands for the social-democratic regime. [D.sub.1] is 1 if a country belongs to the social-democratic welfare regime and 0 otherwise. [D.sub.2] represents the southern regime and [D.sub.3] the liberal regime. The conservative regime is the base regime. The trend T is also interacted with the regime dummies to account for path-dependency of the welfare regimes. As we include a full set of time dummies, one welfare-specific trend cannot be identified. Therefore only three welfare-specific trends are included into Equation (2). Including the fourth welfare-specific trend would not change the results as its impact is captured by [[omega].sub.jt]. Owing to data limitations, other control variables are not interacted. We also do not add the regime intercept dummies, as we already have country-specific time effects.

We further test whether globalization affects the CEE NMS and the EU15 countries differently:

(4) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].

Thus, in Equation (4) a dummy [N.sub.i] is introduced into the baseline equation that is 1 if the country is a CEE NMS. This dummy is interacted with the globalization indicators and the trend. The country index i ranges from 1 to 25, (8) the time index t from 1995 to 2007.

In the final equation once more a dummy [H.sub.jit] is introduced, capturing differences between high- and low-tax countries:

(5) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].

Thereby the effects of globalization in the EU15 countries with lower than average ITRs can be identified by the coefficient [[beta].sub.jg], in the EU15 countries with higher than average ITRs by the coefficient [[beta].sub.jg] + [[delta].sub.jg], in the CEE NMS with lower than average ITRs by the coefficient [[beta].sub.jg] + [[sigma].sub.jg], and in the CEE NMS with higher than average ITRs by ([[beta].sub.jg] + ([[sigma].sub.jg] + [[delta].sub.jg]).

The globalization indices as well as all the control variables except the fraction of elderly people and the government cabinet gravity, enter into Equations (1)-(5) with a 1-year lag. This is done for two reasons: first, to cope with time lags in the political and fiscal decision process and, second, to mitigate potential problems due to endogeneity. A better way to cope with endogeneity issues would be to apply a generalized method of moments (GMM)-approach. However, due to the low number of cross-sections (countries), meaningful GMM-based estimation of the Arellano and Bond (1991) type is precluded (also see Potrafke 2009). (9) A second best approach is to use lagged values of the right-hand side variables (see Wooldridge 2002: 301). Moreover, as we include fixed country-effects in addition to the lagged-dependent variable results are biased for short T applications (see Nickell 1981). This poses only minor problems in case of the EU15 country sample. In this case, the time span ranges from 1970 to 2007. As shown in Judson and Owen (1999), the least square dummy variable (LSDV) estimator with lagged-dependent variable performs comparably well in applications based on unbalanced panels. (10) In case of the aggregate pool including the CEE NMS, the time span covered is lower (1995-2007). The results for this pool of countries must therefore be seen as indicative rather than conclusive.

In each estimation the variance-covariance matrix of the remainder error term, [[epsilon].sub.jit], is calculated using the approach developed by Newey and West (1987). Therefore, standard errors are fully robust with respect to serial correlation as well as general heteroscedasticity (see Baum, Schaffer, and Stillman 2007). (11)

VI. ESTIMATION RESULTS

A. Globalization and ITRs in the EU15

The first six columns of Table 1 show the results for the basic specification in Equation (1) for the EU15 using two alternative KOF globalization indices among the explanatory variables. The lagged dependent variables are significant in all cases. However, globalization seems to matter only for the ITR on labor income. Specifically, KOFecon carries a significant positive coefficient which implies that ceteris paribus economic globalization exerts an upward pressure on the ITR on labor income. Specifically, an increase in economic globalization by 1 percentage point increases the ITR on labor income by 0.75 percentage points. (12)

Concerning control variables, Table 1 shows that most variables fall short of statistical significance. This is due to the inclusion of the lagged-dependent variable which absorbs most of the explanatory power of regressor variables. In line with the theoretical predictions (see Brakman, Garretsen, and von Marreijk 2009; Bucovetsky 1991) and empirical findings (also see Adam and Kammas 2007; Bretschger and Hettich 2002; Garretsen and Peeters 2007; Winner 2005), we can establish that a country's relative size has a positive effect on the ITR on capital income: larger countries have a higher tax burden on capital.

The fraction of elderly people, total government expenditures as well as the debt level have a positive effect on the ITR on labor income. Thus, higher fiscal demands due to increasing total expenditures and the shift of expenditures toward oldage contributions as well as debt payments are financed via taxes on labor income.

In columns (7) to (12) a dummy, which is 1 if the ITR of a country is higher than the mean ITR of the EU15, is interacted with the KOF indices. This intends to show whether countries with above average ITRs react differently to globalization pressures than low ITR countries do. F tests for the joint significance of the KOF index and the interaction dummy are reported at the end of the table. Note that we do not include the dummy variable as a separate regressor as we have country-fixed effects included. Compared to columns (1) to (6), results for the ITR on labor and capital income remain stable. Column (9), however, indicates that countries with an ITR on consumption higher than average decrease their tax rate in response to globalization (although the statistically significant effect is due to total globalization and not economic globalization), while there is no significant effect on ITR on consumption in the countries with a below average ITR. Thus there is a convergence in terms of the ITR on consumption rather than an overall decline.

The lack of a significant effect on the ITR on capital income raises the question whether globalization in Europe has two different dimensions with different effects on taxes: intra-EU integration and integration with the rest of the world. To reflect this aspect, we differentiate between FDI and trade that is going to or originating from the EU countries and FDI and trade that is going to/originating from the rest of the world. Thus instead of the KOF indices in Table 1, we used intra- and extra-EU trade or FDI variables to measure the two dimensions of globalization. The trade variable is defined as bilateral exports plus imports as a ratio to GDP. Similarly, the FDI variable is defined as bilateral FDI inward plus FDI outward stock as a ratio to GDP. (13) Interestingly, trade with the rest of the world has a negative impact on the ITR on capital, but there is no significant effect of intra-EU trade. It seems plausible that external trade integration leads to a decline in tax rates on mobile factors, while integration between a small group of relatively homogenous countries may not create the same downward pressure. (14) There is no statistically significant impact of trade with other EU countries or the rest of the world on the ITR on labor income and the ITR on consumption in the EU15 countries. Globalization measured by FDI inward and outward stock also does not have an effect on the ITR on capital income and labor income. Yet, it has a positive effect on the ITR on consumption in case of FDI that is originating in or going to the EU countries. Although these results are not at odds with those derived using the broader KOF index, the latter variable better serves the purpose of capturing the multidimensional globalization process compared to single components of globalization. However a drawback of the KOF index is the impossibility of distinguishing the intra- and extra-EU dimensions of globalization.

Taken together the results contained in Table 1 imply that globalization does not exert any downward (or upward) pressure on the ITR of capital which is in line with many related studies (see below) but somewhat against the public perception. Yet, we find robust evidence that globalization increases the taxation of labor income, especially in countries with an above average ITR. Thus, a relative shift toward taxation of labor income is likely due to globalization pressures.

How do our results compare to related literature? Our results pertaining to the absence of a significant effect of globalization on the ITR on capital income are in line with Dreher, Gaston, and Martens (2008), Swank and Steinmo (2002), and Swank (2006). However, Dreher (2006a) finds a positive effect, whereas Winner (2005) finds a negative effect of globalization on the ITR on capital income. Similarly, Adam and Kammas (2007), Bretschger (2008), and Bretschger and Hettich (2002) find a negative effect of globalization, but they use the ITR on corporate income as the dependent variable. Their measures of globalization are also limited to the trade volume, the Quinn indices on capital or goods market integration, or the IMF index of restrictions on capital mobility. The difference in the sample (a subsample of the OECD countries vs. EU15) can also explain part of the difference. As we show below, there are differences in how different welfare regimes filter the effects of globalization on the ITRs. Thus, the results are sensitive to the country sample. Swank (1998, 2006) provides a further reason why globalization might have no effect on ITR on capital income. He argues that there has been a general policy shift from "market conforming" to "market regulating" policy caused by a spread of neoliberal policy and theory. Government's aim changed from redistribution to promoting growth by encouraging investment. So the change in tax policy may not only be caused by globalization, but also by a general change in policy. This policy shift is also in line with the cutting of the tax rates and broadening of the tax base, which means that the policy goals shifted from redistribution towards efficiency as the deadweight-loss of taxation depends in a nonlinear way on the statutory tax rate and on the possibilities to substitute taxable income with non-taxed items.

Regarding the ITR on labor, the positive effect of globalization in the EU15 is in line with Dreher, Gaston, and Martens (2008), Adam and Kammas (2007), and Winner (2005). Bretschger and Hettich (2002) also find a shift of the tax burden to labor income, although they do not estimate the effect on the ITR on labor income separately, but rather in the form of the ratio of taxes on capital to labor. However, Dreher (2006a) finds no effect and Swank and Steinmo (2002) find a negative effect on the ITR on labor income.

Our finding of no globalization effect on the ITR on consumption is well established in the literature (see, e.g., Dreher 2006a; Dreher, Gaston, and Martens 2008; Swank and Steinmo 2002). Our findings indicate that there is only a downward convergence effect in the countries with higher than average ITR on consumption; however, other studies do not test such a specification.

Given the upward pressure of globalization on the ITR on labor income, an interesting follow-up question is how the composition of tax revenues changes in response to globalization. Although the share of a tax base in total tax revenues does not measure the average effective tax burden born by a particular category, the trends of its share shows whether revenues from a particular tax base are especially influenced by the globalization process. To address this question, we estimate the effect of globalization on the share of taxes on labor and capital income and consumption in total tax revenues using the same control variables. The results are reported in Table 2. Interestingly, there is no significant effect of globalization on the share of a particular tax base in total tax revenues. In particular, no positive effect on the share of taxes on labor can be established despite the robust positive effect of globalization on the ITR on labor income. This signals that labor income, the denominator of the ITR on labor, increases less than the total tax revenues over time. The growth in total tax revenues usually maps the growth in GDP. Thus, our result of no impact of globalization on the share of labor income taxes in total income taxes despite a positive globalization impact on the ITR on labor income is in line with the declining share of wages in GDP in all the EU15 countries since the 1980s.

B. Globalization, Taxes, and Welfare Regimes in the EU15

Table 3 reports the results in case the EU15 are split into four welfare regimes based on the (extended) Esping-Andersen classification. Again, the lagged-dependent variable is highly significant in each case. Yet, there is an interesting difference between the welfare regimes. According to our estimations, there is a significant negative effect of globalization on the ITR on capital income in the social-democratic welfare regime. The pressure in the social-democratic countries on the ITR on capital indicates a convergence toward the lower European average. Thus, our results indicate convergence rather than a race to the bottom in effective capital income taxation.

Both globalization indices have a positive effect on the ITR on labor income in the conservative regime (the base group). Results also imply that economic globalization exerts an upward pressure on labor income taxes in the social-democratic and the liberal welfare regimes: An increase of economic globalization by 1 percentage point leads to an increase of the ITR on labor income by 0.95 percentage points in the conservative, 2.06 in the social-democratic, and 0.70 in the liberal regime. (15) Although the individual coefficients, which represent contrast to the base group are not statistically significant, the F tests on joint significance of the coefficients (base group plus contrasted regime) imply statistical significance. Thus in all regimes except the southern regime globalization exerts a pressure to increase taxes on labor income. This is interesting, as southern countries showed the strongest increase in labor taxes. Apparently, this rise is unrelated to globalization pressures. The economic effect of globalization on the ITR on labor income is strongest in the social-democratic countries. This might be explained by the prevailing social consensus in these countries about having high taxes on labor income.

Regarding ITR on consumption, results displayed in columns (5) and (6) are in favor of a negative economic globalization impact in the conservative welfare regime. Moreover, a negative pressure is also possible in the social-democratic regime. Yet, in this case the statistical evidence is rather weak as the p value of the F test on joint significance of the coefficients is only borderline significant (p value of .091). These results are consistent with our findings above about convergence in the ITR on consumption via a decline in the countries with higher than average ITR.

Taken together, these results shed more light on the relationship between globalization and ITRs. In particular, globalization has an impact on all three types of effective tax rates, capital, labor, and consumption, which contrasts to the results displayed in Table 1 (where only ITR on labor income is related to globalization). Globalization determines tax policy especially in the social-democratic and the conservative welfare regime. Social-democratic countries have decreased the tax burden on capital, but increased that on labor due to (economic) globalization. This might be considered in line with the broad agreement of the labor unions in the social-democratic countries to preserve the competitiveness of their firms while preserving a certain level of social welfare regime. In the conservative welfare regime especially labor and consumption taxation are hit by globalization pressures. However, even in the liberal regime globalization exerts a positive impact on the ITR on labor income. Thus, the positive relationship between globalization and the ITR on labor income as established in Table 1 represents a common development across three of the four West European welfare regimes.

Finally, we performed several robustness checks for the EU15. First, we excluded inflation. Second, we used unemployment instead of growth as proxy for cyclical effects, with and without inflation. The results regarding globalization are robust and can be received upon request.

C. Globalization and Taxes in the EU15 and CEE NMS

Next, we compare the effects of globalization on taxation in the EU15 and the CEE NMS. Since the data for the CEE NMS start in 1995, we estimate the effects in the EU15 also from 1995 onwards. Therefore the results for the EU15 in this section are not strictly comparable to the ones in Section VI.A. (16) At the first step, we test the differences in the reaction of the old and new member states to globalization. Again we do not estimate a full interaction model with dummies for the control variables to gain degrees of freedom. Table 4 shows the results for the pool of the EU15 and the CEE NMS for the period 1995-2007 with the interaction dummy for the CEE NMS as shown in Equation (4). The EU15 is the base group, and the effect of KOF indices in the CEE NMS is given by the summation of the base coefficient and the interaction dummy. In columns (7)-(12), a second interaction dummy is introduced that is 0 if the ITR of a country is lower than the EU25 mean ITR, and 1 otherwise (Equation (5)). Thus, four different country groups can be distinguished: first, the EU15 countries with low ITRs as the base group; second, the EU15 countries with high ITRs; third, the CEE NMS with low ITRs; and fourth the CEE NMS with high ITRs. The summation of the coefficients of KOF index and the interaction dummy for the CEE NMS shows the effect of globalization in the CEE NMS with ITR lower than average, whereas the summation of the coefficients of KOF index, the interaction dummy for the CEE NMS, and the interaction dummy for the countries with ITR higher than EU25 average shows the effect of globalization in the CEE NMS with ITR higher than average. The F tests for joint significance are reported at the end of the table.

Again, the lagged dependent variable is statistically significant in each case. Economic globalization has now a negative effect on the ITR on capital income in the EU15 in the period after 1995, in contrast to the estimations in Table 1 for the whole sample. The effect is valid for the EU15 countries with both a lower and higher than average ITR on capital as can be seen in columns (7) and (8).

The ITR on labor income in both groups in the EU15 (high and low ITR countries) is positively affected by KOFecon, but there is no significant effect in the CEE NMS.

Globalization (both indicators) exerts a negative impact on the ITR on consumption in the CEE NMS; but when the countries with higher than average ITR are distinguished, interesting results emerge, particularly in the case of economic globalization: The CEE NMS with a higher than average ITR on consumption decrease their tax rates, whereas there is no significant response in the CEE NMS with lower than average ITR. Again there is some evidence of convergence. The effect of overall globalization is negative and statistically similar in both groups of CEE NMS, though. This may be related to the reforms during the process of EU accession in all the CEE NMS. Specifically, CEE NMS had to realize the acquis communautaire of the EU which is rather far-reaching in the field of indirect taxation. In the EU15, although there is no significant effect on the ITR on consumption in the aggregate, the results in column (12) show that the EU15 countries with lower than average ITR on consumption increase their tax rate in response to globalization, while the other EU15 countries keep their tax rates stable. These results are consistent with the results for the EU15 with the full sample; now convergence is taking place via an increase in the tax rates in countries with lower than average ITR on consumption rather than a decline in those with taxes higher than average.

The reason for the lack of a significant effect in the CEE NMS concerning the ITR on labor or capital income could be that the increase in the formalization of these economies along with international integration might offset any negative effects of globalization on taxes (Duman 2009). However, we are not able to test this argument due to lack of time series data on the informal economy. Of course, another problem can be the limited data availability leading to high standard errors.

Regarding the control variables, GDP growth and inflation have a positive effect on the ITR on capital income. The inflation rate has a negative impact on labor taxation, whereas size, government expenditure, and public debt have positive effects. The ITR on consumption is negatively affected by the relative size of a country, as well as oldage and government party. Government expenditures have a positive impact on consumption taxes.

Again we tested the robustness of our results using the intra- and extra-EU trade and FDI variables instead of the aggregate KOF indices. However, especially with regards to the rather short time series for the CEE NMS (only 4 or less years for most countries in the case of FDI and 9 years in the case of trade), the results have to be handled with care. In this aggregate pool for the post-1995 period, there is no effect of trade on the ITR on capital income. We find a positive effect of trade with the rest of the world on the ITR on labor income for the EU15 countries and a negative effect of trade with the rest of the world on consumption taxes for the CEE NMS. Concerning FDI, there is no effect on the ITR on labor income, but a negative effect of FDI originating from or going to the rest of the world on the ITR on capital income in the EU15 and a positive effect of FDI within the EU on the ITR on consumption in the EU15. Thus the positive effect of globalization on the ITR on labor, and the negative effect on ITR on capital in the EU15, and the negative effect of globalization on consumption taxes in the CEE NMS which are also found in Table 4, seem to originate from extra-EU trade or FDI.

Taken together, in the CEE NMS there is no effect of globalization on the ITR on labor and capital income but we find a negative impact on the ITR on consumption, particularly in the CEE NMS with higher than average ITR in the case of economic globalization. In the EU15 the upward pressure of globalization on the ITR on labor are once again robust findings. There is some evidence of a negative effect on the ITR on capital in the EU15 in the post-1995 period. However, we would be cautious about this latter result as it stands at odds with the more reliable findings based on the full EU15 sample. In the EU15, in the post-1995 period the EU15 countries with lower than average ITR on consumption increase their tax rate in response to globalization, whereas the other EU15 countries keep their tax rates stable.

Finally, we again estimate the effects of globalization on the composition of the tax revenues. The results are in Table 5. In the EU15 consistent with the results for the full sample, there is no significant effect on the share of taxes on labor income; but according to the results for the post-1995 period, the share of taxes on capital income is decreasing in response to economic globalization, and the share of taxes on consumption are increasing. So there is a shift towards indirect taxes with a more regressive nature, which fall more heavily on the immobile factor of production and a decline in the tax share of the mobile factor. Interestingly, the opposite is true for the CEE NMS: the share of taxes on capital increases and that on consumption decreases. These findings are consistent with the estimation results for the ITRs. The significant positive effect on the share of taxes on capital income in the CEE NMS may be related to the increase in FDI and profit income in these countries after the transition.

SUMMARY AND CONCLUSIONS

This article analyzes the effects of globalization on the ITR on labor income, capital income, and consumption with an emphasis on the differences among welfare regimes in the EU15 and also between the EU15 and CEE NMS.

Overall, our results confirm that globalization leads to a higher tax burden on labor income in the EU15: there is a positive effect of globalization on the ITR on labor income in the EU15, but no effect on the ITR on capital income. There is evidence of convergence in terms of the ITR on consumption as the countries with higher than average ITR decrease their ITR on consumption.

However, there are differences in the response to globalization across the welfare regimes within the EU15. Globalization has a significant negative effect on the ITR on capital income in the social-democratic regime, but no significant effect in the other welfare regimes. Regarding the ITR on consumption, there is a significant negative effect of globalization in the social-democratic and conservative regimes. In the case of the ITR on labor income, we find evidence that globalization causes an increase in all the welfare regimes except the southern regime.

There are also important differences between the EU15 and the CEE NMS; in the latter, we find no effect of globalization on the ITR on labor and capital income but a negative impact on the ITR on consumption. Interestingly, there is again evidence of convergence, particularly in response to economic globalization as the CEE NMS with higher than average ITR on consumption decrease their tax rates in response to globalization.

doi: 10.1111/j.1465-7295.2011.00420.x

ABBREVIATIONS

AETR: Average Effective Tax Rate

CEE NMS: Central and Eastern European New Member States

EATR: Effective Average Tax Rate

EMTR: Effective Marginal Tax Rate

FDI: Foreign Direct Investment

GDP: Gross Domestic Product

GMM: Generalized Method of Moments

ITR: Implicit Tax Rate

KOF: Konj unkturforschungsstelle

LSDV: Least Square Dummy Variable

OECD: Organization for Economic Cooperation and Development

STR: Statutory Tax Rate

APPENDIX

TABLE A1
Data Availability

Implicit Tax Rates

1970-2007: BE, GE, DE, FR, IE, IT, LU (no ITR on capital
available), NE, UK 1980-2007: AT, ES, FI, GR, PT, SE 1995-2007: CZ,
EE, HU (ITR on capital since 2000), LT, LV, PL, SK, SI Later than
1995: BG 1999-2007, RO: ITR on capital 1998-2004, ITR on labor and
consumption 1999-2007 All data from Eurostat; ITR on capital for
BG, HU, RO, SI partly own calculations

TABLE A2
Data Sources and Description

ITR cap Implicit tax rate on capital income

ITR con Implicit tax rate on consumption

ITR lab Implicit tax rate on labor income

Captaxes Tax revenues on capital income as a share of total tax
 revenues
Labtaxes Tax revenues on labor income as a share of total tax
 revenues
Contaxes Tax revenues on consumption as a share of total tax
 revenues
KOFglobal KOF Index of economic, political, and social
 globalization, ranging from 1 to 100

KOFecon KOF Index of economic globalization, combining
 actual flows and restrictions

growth Growth rate of real GDP
debt General government consolidated gross debt
inflation GDP deflator
expenditures Total expenditures of general government as
 percentage of GDP
oldage Population > 64 as a percentage of total population
gov-party government cabinet composition (Schmidt index):
 (1) hegemony of right-wing (and center) parties;
 (2) dominance of right-wing (and center) parties;
 (3) balance of power between left and right parties;
 (4) dominance of social-democratic and other left
 parties; (5) hegemony of social-democratic and
 other left parties
size GDP of country i as share of total GDP

ITR cap Eurostat, European Commission 1970-2007
 (2000)
ITR con Eurostat, European Commission 1970-2007
 (2000)
ITR lab Eurostat, European Commission 1970-2007
 (2000)
Captaxes Eurostat, European Commission 1970-2007
 (2000)
Labtaxes Eurostat, European Commission 1970-2007
 (2000)
Contaxes Eurostat, European Commission 1970-2007
 (2000)
KOFglobal Dreher (2006a) updated in 1970-2007
 Dreher, Gaston, and Martens
 (2008)
KOFecon Dreher (2006a) updated in 1970-2007
 Dreher, Gaston, and Martens
 (2008)
growth AMECO database 1970-2007
debt AMECO database 1970-2007
inflation AMECO database 1970-2007
expenditures AMECO database 1970-2007

oldage AMECO database 1970-2007
gov-party Comparative political Dataset I 1970-2007
 and III (for the CEE NMS),
 University of Bern

size AMECO database 1970-2007

TABLE A3
Data Summary-EU15, 1970-2007

 Observ. M SD Min Max

ITR cap 471.00 26.92 9.02 7.04 56.01
ITR lab 510.00 33.26 7.58 8.17 49.40
ITR con 510.00 21.19 5.48 5.43 34.00
Captax 510.00 20.47 6.74 6.52 42.20
Labtax 510.00 49.08 8.25 25.21 65.90
Contax 510.00 30.76 5.94 17.70 50.76
KOFglobal 570.00 70.91 14.67 35.29 93.38
KOFecon 570.00 70.90 15.64 40.00 98.90
size 570.00 6.40 7.07 0.15 24.27
expenditures 432.00 47.93 6.70 29.27 71.68
debt 554.00 51.55 29.27 4.06 134.16
oldage 569.00 14.16 2.14 9.15 19.99
gov-party 552.00 2.65 1.47 1.00 5.00
inflation 570.00 6.31 5.55 -1.88 27.21
growth 570.00 2.97 2.43 -6.57 11.49

TABLE A4
Data Summary--CEE NMS and EU15, 1995-2007

 Obs M SD Min Max

ITR cap 295.00 23.76 9.67 4.90 49.80
ITR lab 317.00 36.39 5.85 24.10 49.40
ITR con 317.00 21.67 4.30 14.20 34.00
Captax 321.00 18.91 6.65 5.20 34.30
Labtax 321.00 48.30 7.08 31.60 62.80
Contax 321.00 33.00 5.96 23.60 56.50
KOFglobal 325.00 77.87 11.35 39.66 93.38
KOFecon 325.00 78.53 12.01 35.94 98.90
size 325.00 3.92 5.94 0.06 23.00
expenditures 320.00 45.04 6.77 31.46 65.10
debt 318.00 49.29 29.18 3.49 129.79
oldage 324.00 15.13 1.93 10.87 19.99
gov party 324.00 2.81 1.38 1.00 5.00
inflation 325.00 9.19 53.87 -1.88 948.28
growth 325.00 3.81 2.71 -9.40 12.23

TABLE A5
Correlation Matrix-EU15, 1970-2007

 ITR cap ITR lab ITR con Captax Labtax

ITR cap 1.00
ITR lab 0.04 1.00
ITR con 0.04 0.40 1.00
Captax 0.28 -0.50 -0.64 1.00
Labtax 0.07 0.76 0.39 -0.75 1.00
Contax -0.40 -0.54 0.20 -0.03 -0.63
KOFglobal -0.31 0.65 0.41 -0.39 0.44
KOFecon -0.10 0.29 0.33 0.03 0.02
size 0.43 -0.09 -0.47 0.31 0.05
expenditures 0.15 0.70 0.40 -0.52 0.67
debt -0.27 0.32 -0.25 0.22 -0.05
oldage 0.33 0.44 -0.17 0.13 0.16
gov-party 0.10 0.13 -0.01 -0.08 0.09
inflation -0.05 -0.57 -0.32 0.19 -0.36
growth -0.29 -0.16 0.07 0.12 -0.26

 KOFecon size expenditures debt oldage

KOFecon 1.00
size -0.49 1.00
expenditures 0.12 -0.14 1.00
debt 0.32 -0.21 0.31 1.00
oldage 0.16 0.19 0.20 0.31 1.00
gov-party -0.06 0.05 0.05 -0.05 0.23
inflation -0.53 0.00 -0.35 -0.18 -0.39
growth 0.19 -0.15 -0.40 -0.05 -0.21

 Contax KOFglobal

ITR cap
ITR lab
ITR con
Captax
Labtax
Contax 1.00
KOFglobal -0.17 1.00
KOFecon -0.04 0.66
size -0.47 -0.51
expenditures -0.39 0.36
debt -0.15 0.28
oldage -0.36 0.21
gov-party -0.06 0.04
inflation 0.32 -0.57
growth 0.25 0.05

 gov-party inflation

KOFecon
size
expenditures
debt
oldage
gov-party 1.00
inflation -0.03 1.00
growth 0.00 -0.11

TABLE A6
Correlation Matrix-CEE NMS and EU15, 1995-2007

 ITR cap ITR lab ITR con Captax Labtax

ITR cap 1.00
ITR lab 0.17 1.00
ITR con 0.27 0.29 1.00
Captax 0.50 -0.41 -0.28 1.00
Labtax 0.13 0.77 0.25 -0.62 1.00
Contax -0.67 -0.50 -0.02 -0.29 -0.57
KOFglobal 0.25 0.40 0.41 0.02 0.38
KOFecon 0.20 0.04 0.47 0.02 0.16
size 0.48 0.02 -0.26 0.37 0.13
expenditures 0.59 0.66 0.47 -0.06 0.54
debt 0.36 0.26 -0.11 0.45 0.03
oldage 0.28 0.29 -0.19 -0.01 0.30
gov-party 0.15 0.05 -0.06 0.03 0.06
inflation -0.38 -0.16 -0.21 -0.14 -0.18
growth -0.52 -0.25 0.00 -0.24 -0.25

 KOFecon size expenditures debt oldage

KOFecon 1.00
size -0.13 1.00
expenditures 0.08 0.18 1.00
debt 0.09 0.32 0.50 1.00
oldage 0.11 0.40 0.27 0.46 1.00
gov-party -0.03 0.13 0.17 0.13 0.22
inflation -0.43 -0.26 -0.32 -0.34 -0.27
growth 0.09 -0.40 -0.55 -0.46 -0.30

 Contax KOFglobal

ITR cap
ITR lab
ITR con
Captax
Labtax
Contax 1.00
KOFglobal -0.48 1.00
KOFecon -0.23 0.64
size -0.53 -0.12
expenditures -0.60 0.45
debt -0.50 0.37
oldage -0.34 0.14
gov-party -0.11 -0.09
inflation 0.35 -0.41
growth 0.54 -0.19

 gov-party inflation

KOFecon
size
expenditures
debt
oldage
gov-party 1.00
inflation 0.02 1.00
growth -0.06 0.14

REFERENCES

Adam, A., and P. Kammas. "Tax Policy in a Globalized World: Is It Politics After All?" Public Choice, 133, 2007, 321-41.

Aidukaite, J. "The Emergence of the Post-Socialist Welfare State: The Case of the Baltic States: Estonia, Latvia and Lithuania." Sodertorn Doctoral Dissertations, 2004.

Arellano, M., and S. Bond. "Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations." Review of Economic Studies, 58, 1991. 277-97.

Armingeon, K., R. Careja, S. Engler, M. Gerber, P. Leimgruber, and P. Potolidis. Comparative Political Data Set III 1990-2008. Institute of Political Science, University of Berne, 2010.

Baum, C. F., M. Schaffer, and S. Stillman. "Enhanced Routines for Instrumental Variables/GMM Estimation and Testing." Boston College Working Papers in Economics 667, Boston College Department of Economics, 2007.

Bellak, C., and M. Leibrecht. "Do Low Corporate Income Tax Rates Attract FDI?--Evidence from Central--and East European Countries." Applied Economics, 41(21), 2009, 2691-703.

Bjornskov, C., and N. Potrafke. "Politics and Privatization in Central and Eastern Europe: A Panel Data Analysis." Social Science Research Network, 2010. Accessed October 13, 2010. http://ssrn.com/abstract=1319730.

Bohle, D., and B. Greskovits. "Neoliberalism Embedded Neoliberalism and Neocorporatism: Towards Transnational Capitalism in Central-Eastern Europe." West European Politics, 30, 2007, 443-66.

Bonoli, G. "Classifying Welfare States: A Two-Dimension Approach." Journal of Social Policy, 26, 1997, 351-72.

Brakman, S., H. Garretsen, and C. yon Marreijk. The New Introduction to Geographical Economics. Cambridge: Cambridge University Press, 2009.

Brambor, T., W. Clark, and M. Golder. "Understanding Interaction Models: Improving Empirical Analyses." Political Analysis, 14, 2006, 63-82.

Bretschger, L. "Taxes, Mobile Capital, and Economic Dynamics in a Globalising World." CER-ETH Working Paper 05/43, 2008.

Bretschger, L., and F. Hettich. "Globalisation, Capital Mobility and Tax Competition: Theory and Evidence for OECD Countries." European Journal of Political Economy, 18, 2002, 695-716.

Brueckner, J. K. "A Tiebout/Tax-Competition Model." Journal of Public Economics, 77, 2000, 285-30.

Bruno, G. "Approximating the Bias of the LSDV Estimator for Dynamic Unbalanced Panel Data Models." KITES Working Papers 159, KITES, Centre for Knowledge, Internationalization and Technology Studies, Universita' Bocconi, Milano, Italy, revised July 2004.

Bucovetsky, S. "Asymmetric Tax Competition." Journal of Urban Economics, 30, 1991, 167-81.

Campbell, J. L. "Fiscal Sociology in an Age of Globalization: Comparing Tax Regimes in Advanced Capitalist Countries," in The Economic Sociology of Capitalism, edited by V. Nee and R. Swedberg. Princeton, NJ: Princeton University Press, 2005, 391-418.

Clausing, K. "Corporate Tax Revenues in OECD Countries." International Tax and Public Finance, 14, 2007, 115-33.

Devereux, M. P. "Developments in the Taxation of Corporate Profit in the OECD Since 1965: Rates, Bases and Revenues." Working Papers 0704, Oxford University Centre for Business Taxation, 2007.

Devereux, M. P., and R. Griffith. "The Taxation of Discrete Investment Choices." IFS Working Paper series no. W98/16, 1998.

Devereux, M. P., R. Griffith, and A. Klemm. "Corporate Income Tax Reforms and International Tax Competition." Economic Policy, 17, 2002, 451-95.

Dreher, A. "The Influence of Globalization on Taxes and Social Policy: An Empirical Analysis for OECD Countries." European Journal of Political Economy, 22, 2006a, 179-201.

--. "Does Globalization Affect Growth? Evidence from a New Index of Globalization." Applied Economics, 38, 2006b, 1091-110.

Dreher, A., N. Gaston, and P. Martens. Measuring Globalization--Gauging its Consequences. New York: Springer, 2008.

Duman, A. Flux or Fixed: Tax Reforms, Informal Economy and Foreign Investment in New EU Member States, Vienna University of Economics and Business, Department of Economics and Research Institute International Taxation Discussion Paper No 36, 2009.

Esping-Andersen, G. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press, 1990.

--. "After the Golden Age? Welfare State Dilemmas in a Global Economy," in Welfare States in Transition, edited by G. Esping-Andersen. London: Sage, 1996, 1-31.

European Commission. Structures of the Taxation Systems in the European Union. Luxembourg: Eurostat Statistical Books, 2000.

--. Taxation Trends in the European Union, Luxembourg: Eurostat Statistical Books, 2008.

--. Taxation Trends in the European Union, Luxembourg: Eurostat Statistical Books, 2009.

Fenger, H. J. M. "Welfare Regimes in Central and Eastern Europe: Incorporating Post-Communist Countries in a Welfare Regime Typology." Contemporary Issues and Ideas in Social Sciences, 3, 2007. Accessed September 5, 2010, http://journal.ciiss.net/index. php/ciiss/article/view/45/37

Ferge, Z. "Welfare and 'Ill-Fare' Systems in Central-Eastern Europe," in Globalization and European Welfare States--Challenges and Change, edited by R. Sykes, B. Palier, and P. M. Prior. Houndmills-Basingstoke, UK: Palgrave, 2001, 127-53.

Ferrara, M. "The Southern Model of Welfare in Social Europe." Journal of European Social Policy, 6, 1996, 179-89.

Garretsen, H., and J. Peeters. "Capital Mobility, Agglomeration and Corporate Tax Rates: Is the Race to the Bottom for Real?" CESifo Economic Studies, June 4, 2007.

--. "FDI and the Relevance of Spatial Linkages: Do Third-Country Effects Matter for Dutch FDI?" Review of World Economics, 145, 2009, 3219-338.

Gemmel, N., R. Kneller, and I. Sanz. "Foreign Investment, International Trade and the Size and Structure of Public Expenditures." European Journal of Political Economy, 24, 2008, 151-71.

Havlik, P., and M. Landesmann. "Structural Change, Productivity and Employment in the New EU Member States," in Economic Restructuring and Labour Markets in the Accession Countries. The Vienna Institute for International Economic Studies Research Project commissioned by EU DG Employment. Social Affairs and Equal Opportunities, 2005.

Judson, R., and A. Owen. "Estimating Dynamic Panel Data Models: A Guide for Macroeconomists." Economics Letters, 65, 1999, 9-15.

Kautto, M., and J. Kvist. "Parallel Trends, Persistent Diversity: Nordic Welfare States in the European and Global Context." Global Social Policy, 2, 2002, 189-208.

Keen, M., Y. Kim, and R. Varsano. "The 'Flat Tax(es)': Principles and Experience." International Tax and Public Finance, 15, 2008, 712-51.

Krogstrup, S. "A Synthesis of Recent Developments in the Theory of Capital Tax Competition." EPRU Working Paper Series 04-02, 2004.

--. Are Corporate Taxes Racing to the Bottom in the European Union, Mimeo. Geneva: Graduate Institute of International Studies, 2005.

Leibrecht, M., M. Klien, and O. Onaran. Forthcoming. "Globalization, Welfare Regimes and Social Protection Expenditures in Western and Eastern European Countries." Public Choice. doi: 10.1007/s 11127-010-9685-7.

Lelkes, O. "A Great Leap Towards Liberalism? The Hungarian Welfare State." International Journal of Social Welfare, 9, 2000, 92-102.

Loretz, S. "The Condition of Corporate Taxation in the OECD in a Wider Context." Oxford Review of Economic Policy, 24, 2008, 639-60.

Mendoza, E. G., A. Razin, and L. L. Tesar. "Effective Tax Rates in Macroeconomics Cross-Country Estimates of Tax Rates on Factor Incomes and Consumption." Journal of Monetary Economics, 34, 1994, 297-323.

Newey, W. K., and K. D. West. "A Simple, Positive Semi-Definite, Heteroscedasticity and Autocorrelation Consistent Covariance Matrix." Econometrica, 55, 1987, 703-08.

Nichols, A., and M. Schaffer. "Cluster Errors in Stata," Paper presented at the 2007 UK Stata Users Group meeting. Accessed October 10, 2010. http://www. stata.com/meeting/13uk/nichols_crse.pdf

Nickell, S. J. "Biases in Dynamic Models with Fixed Effects." Econometrica, 49, 1981, 1417-26.

Noelke, A., and A. Vliegenthart. "Enlarging the Varieties of Capitalism: The Emergence of Dependent Market Economies in East Central Europe." World Politics, 61, 2009, 600-702.

Oates, W. E. Fiscal Federalis. New York: Harcourt Brace Jovanovich, 1972.

Orenstein, M. A., and M. R. Haas. Globalization and the Future of the Welfare State. Pittsburgh: University of Pittsburgh Press, 2005, 130-52.

Potolidis P., M. Gerber, and P. Leimgruber. Comparative Political Data Set 1960-2008. Institute of Political Science, University of Berne, 2010.

Potrafke, N. "Social Expenditures as a Political Cue Ball? OECD Countries under Examination." DIW discussion papers, Berlin, 2007.

--. "Did Globalization Restrict Partisan Politics? An Empirical Evaluation of Social Expenditures in a Panel of OECD Countries." Public Choice, 140, 2009, 105-24.

--. "Does Government Ideology Influence Deregulation of Product Markets? Empirical Evidence from OECD Countries." Public Choice, 143, 2010, 135-55.

Quinn, D. "The Correlates of Change in International Financial Regulation." The American Political Science Review, 91, 1997, 531-51.

Roodman, D. "How to Do xtabond2: An Introduction to Difference and System GMM in Stata." Stata Journal, StataCorp LP, 9(1), 2009, 86-136.

Schaffer, M. E. "xtivreg2: Stata Module to Perform Extended IV/2SLS, GMM and AC/HAC, LIML and k-class Regression for Panel Data Models." Accessed http://ideas.repec.org/c/boc/bocode/s456501.html, 2010.

Scharpf, F. W., and V. A. Schmidt. "Introduction," in Welfare and Work in the Open Economy, Vol. I: From Vulnerability to Competitiveness, edited by F. W. Scharpf and V. A. Schmidt. Oxford, UK: Oxford University Press, 2000, 1-20.

Schmidt, M.G. Wohlfahrtsstaatliche Politik unter Biirgerlichen und Sozialdemokratischen Regierungen. Ein Internationaler Vergleich. Frankfurt a. M., Germany: Campus, 1982.

--. "Regierungen--parteipolitische Zusammensetzung," in Die westlichen Laender, edited by M. G. Schmidt. Munich, Germany: C.H. Beck, 1992, 393-400.

Slemrod, J. "Are Corporate Tax Rates, or Countries, Converging?" Journal of Public Economics, 88, 2004, 1169-86.

Standing, G. "Social Protection in Central and Eastern Europe: A Tale of Slipping Anchors and Torn Safety Nets," in Welfare States in Transition, edited by G. Esping-Andersen. London: Sage, 1996.

Swank, D. "Funding the Welfare State: Globatization and the Taxation of Business in Advanced Economies." Political Studies, 46, 1998, 671-92.

--. "Mobile Capital, Democratic Institutions, and the Public Economy in Advanced Industrial Societies." Journal of Comparative Policy Analysis, 3, 2001, 133-62.

--. "Tax Policy in an Era of Internationalization, Expanding the Spread of Neoliberalism." International Organization, 60, 2006, 847-82.

Swank, D., and S. Steinmo. "The New Political Economy of Taxation in Advanced Capitalist Democracies." American Journal of Political Science, 46, 2002, 642-55.

Tanzi, V. Taxation in an Integrating World. Washington, DC: The Brookings Institution, 1995.

Thuronyi, V. "Adjusting Taxes for Inflation," in Tax Law Design and Drafting, Volume, Chapter 13, edited by V. Thuronyi. Washington, DC: IMF, 1996.

Wilson, J. D. "Theories of Tax Competition." National Tax Journal, 52, 1999, 269-303.

Winner, H. "Has Tax Competition Emerged in OECD Countries? Evidence from Panel Data." International Tax and Public Finance, 12, 2005, 667-87.

Wooldridge, J. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press, 2002.

World Bank. Special Economic Zones: Performance, Lessons Learned, and Implications for Zone Development, April 2008, Washington, 2008.

Zodrow, G. R., and P. Mieszkowski. "Pigou, Tiebout, Property Taxation, and the Underprovision of Local Public Goods." Journal of Urban Economics, 19, 1986, 356-70.

OZLEM ONARAN, VALERIE BOESCH and MARKUS LEIBRECHT *

* The authors are grateful to Ilker Atac, Oliver Prausmuller, Claudia Hochgatterer, and two anonymous referees for very helpful comments. Support from the FWF Project Nr. F2008 is acknowledged. The usual disclaimer applies.

Onaran: Senior Lecturer in Economics, Department of Economics and Quantitative Methods, Westminster University, 35 Marylebone Road, London NW1 5LS, UK. Phone 20 7911 5000, Fax 20 7911 5703, E-mail o.onaran@westminster@ac.uk

Boesch: Research Assistant, Vienna University of Economics and Business, Research Institute International Taxation, Augasse 2-6, A-1090 Vienna, Austria. Phone 1 31336-5932, Fax 1 31336-728, E-mail valerie.boesch@wu.ac.at

Leibrecht: Senior Economist, Department of Economics and Research Institute International Taxation, Vienna University of Economics and Business, Augasse 2-6, A-1090 Vienna, Austria. Phone 1 31336-5833, E-mail markus.leibrecht@wu.ac.at

(1.) ITRs are also referred as average effective tax rates (AETRs).

(2.) This is based on Garretsen and Peeters (2009), who use a dummy that is one if the statutory tax rates of the country are higher than the distance-weighted average of the statutory tax rate of the other countries.

(3.) The EMTR is based on the costs of capital of an investment project at the break-even point. The EATR measures the tax burden on a project earning a positive economic rent.

(4.) A widely used method to calculate ITRs using National Accounts data has been developed by Mendoza, Razin, and Tesar (1994).

(5.) For data availability, see Table A1.

(6.) For the ITR on capital income tax revenue on capital income is divided by operating surplus (data source: Eurostat).

(7.) Potrafke (2007, 2009) also develops an index for government's ideology; however the data available in these papers end in 2003 and only cover OECD countries. Bjornskov and Potrafke (2010) calculate a similar index that also covers the CEE NMS; however their data is not available publicly to the best of our knowledge. Therefore we use the Schmidt index. The index of Potrafke (2007, 2009) for the Western countries for the period until 2003 is broadly similar to the Schmidt index (the correlation coefficient between the two indices is very high [0.82] for the common sample); it seems to vary less in time and group more countries as left wing parties. Potrafke (2010) cites another ideology index developed by Bjornskov (2008, cited in Potrafke 2010); and writes that the correlation coefficient between this index and that in Potrafke is 0.70.

(8.) We did not include Cyprus and Malta to this pool, as we wanted to keep the group of EU15 comparable to the former estimations, and it would not be correct to group Cyprus and Malta together with the transition economies in CEE.

(9.) The cross-sectional dimensions are 15 (EU15) and 25 (EU15 and CEE NMS aggregate pool), respectively.

(10.) In addition to the LSDV estimator, we applied the GMM-estimator of Arellano and Bond (1991) using Roodman's xtabond2 command in Stata (also see Roodman 2009). However, we did not end up with a specification that meets minimum requirements for reliable estimates: The Hansen-J test signals extreme overfitting (p value of 1) and the test on second-order autocorrelation leads to a low p value. We also applied the bias-corrected LSDV estimator developed by Bruno (2004), which is available as xtlsdvc command in Stata. However, to derive the bias-correction term this estimator also relies on Arellano and Bond type estimates, which, in our case, cannot be meaningfully retrieved. Note that the Bruno type estimator also does not solve the endogeneity issue with respect to right-hand side regressor other than the lagged-dependent variable. For these reasons, we stick to the LSDV estimator.

(11.) Newey--West-HAC robust standard errors are chosen as the alternative cluster-robust standard errors need a rather large number of clusters (here countries) for reliable interence. Typically a minimum cluster dimension of about 50 is required (see Nichols and Schaffer 2007). Estimations are carried out with Schaffer's xtivreg2 Stata command (see Schaffer 2010).

(12.) The total economic effect of globalization on the ITRs is calculated as the coefficient of the KOF index multiplied by the average change of the KOF index. The average change of KOFecon in the EUI5 in the common period 1980-2007 is 23,375.

(13.) The data are taken from Eurostat's New Cronos database and OECD. Eurostat provides rather short series on bilateral FDI stock and trade for most countries. Therefore we have supplemented the Eurostat data with data from the OECD. Owing to data limitations we did not add trade and FDI simultaneously to the estimations, since then we would lose even more observations due to differences in the dates of missing values for both variables. The estimation results are available upon request.

(14.) We are grateful to the referee who pointed out this aspect.

(15.) The total economic effect of globalization on the ITRs is calculated as the sum of the coefficient of the KOF index and the coefficient of the interaction of the KOF index and regime dummy multiplied by the average change of the KOF index in the respective regime. The average change in the common period 1980-2007 in the conservative regime of KOFglobal is 18,243 and of KOFecon 19,872; the average change in the social-democratic regime of KOFglobal is 18,525 and of KOFecon 27,895; the average change of KOFecon in the liberal regime is 10,780.

(16.) As explicated above, the results presented in this section should be seen as indicative rather than conclusive due to the limited time range the estimates are based on.

TABLE 1
The Effects of Globalization on the Implicit Tax Rate on Capital
Income, Labor Income, and Consumption in the EU 15, 1970-2007

 (1) (2) (3)
 ITR cap ITR cap ITR lab

ITR [cap.sub.(t-1])] 0.745 *** 0.744 ***
 (0.000) (0.000)
ITR [lab.sub.t-1]) 0.718 ***
 (0.000)
ITR [con.sub.(t-1)]

KOF [global.sub.(t-1)] -0.035 0.028
 (0.530 (0.220
KOF [econ.sub.(t-1)] 0.035
 (0.561)
[size.sub.(t-1)] 1.007 ** 1.009 ** 0.076
 (0.025) (0.028) (0.585)
[growth.sub.(t-1)] 0.162 0.147 -0.002
 (0.162) (0.214) (0.964)
[inflation.sub.(t-1)] -0.072 -0.072 0.034
 (0.418) (0.398) (0.108)
[expenditures.sub.(t-1)] 0.012 -0.008 0.100 ***
 (0.844) (0.900) (0.000)
[debt.sub.(t-1)] 0.002 -0.002 0.009
 (0.900) (0.902) (0.121)
[oldage.sub.(t-1)] 0.368 0.314 0.197 **
 (0.315) (0.398) (0.031)
[govparty.sub.(t)] -0.043 -0.043 -0.017
 (0.712) (0.713) (0.659)
KOFgl x high [cap.sub.(t-1)]

KOFec x high [cap.sub.(t-1)]

KOFgl x high [lab.sub.(t-1)]

KOFec x high [lab.sub.(t-1)]

KOFgl x high [con.sub.(t-1)]

KOFec x high [con.sub.(t-1)]

Trend x high_cap

Trend x high_lab

trend x high_con

high cap

high lab

high con

[R.sup.2] 0.702 0.702 0.931
N 380 380 398
F-test TD 0.001 0.002 0.224
F-test high ITR

 (4) (5) (6)
 ITR lab ITR con ITR con

ITR [cap.sub.(t-1])]

ITR [lab.sub.t-1]) 0.715 ***
 (0.000)
ITR [con.sub.(t-1)] 0.833 *** 0.834 ***
 (0.000) (0.000)
KOF [global.sub.(t-1)] -0.018
 (0.197)
KOF [econ.sub.(t-1)] 0.032 ** -0.011
 (0.045) (0.319)
[size.sub.(t-1)] 0.148 0.042 0.004
 (0.253) (0.649) (0.959)
[growth.sub.(t-1)] -0.013 0.039 0.043
 (0.705) (0.152) (0.115)
[inflation.sub.(t-1)] 0.028 0.013 0.016
 (0.189) (0.510) (0.430)
[expenditures.sub.(t-1)] 0.094 *** 0.011 0.012
 (0.000) (0.453) (0.427)
[debt.sub.(t-1)] 0.011 * -0.006 -0.007 *
 (0.056) (0.136) (0.076)
[oldage.sub.(t-1)] 0.225 ** 0.068 0.045
 (0.011) (0.296) (0.475)
[govparty.sub.(t)] -0.004 -0.004 -0.011
 (0.918) (0.878) (0.655)
KOFgl x high [cap.sub.(t-1)]

KOFec x high [cap.sub.(t-1)]

KOFgl x high [lab.sub.(t-1)]

KOFec x high [lab.sub.(t-1)]

KOFgl x high [con.sub.(t-1)]

KOFec x high [con.sub.(t-1)]

Trend x high_cap

Trend x high_lab

trend x high_con

high cap

high lab

high con

[R.sup.2] 0.931 0.837 0.837
N 398 398 398
F-test TD 0.049 0.000 0.000
F-test high ITR

 (7) (7) (9)
 ITR cap ITR cap ITR lab

ITR [cap.sub.(t-1])] 0.623 *** 0.623 ***
 (0.000) (0.000)
ITR [lab.sub.t-1]) 0.713 ***
 (0.000)
ITR [con.sub.(t-1)]

KOF [global.sub.(t-1)] 0.057 0.030
 (0.361) (0.225)
KOF [econ.sub.(t-1)] 0.030
 (0.555)
[size.sub.(t-1)] 0.905 ** 1.020 ** 0.216
 (0.047) (0.026) (0.121)
[growth.sub.(t-1)] 0.133 0.116 0.009
 (0.220) (0.302) (0.804)
[inflation.sub.(t-1)] -0.028 -0.042 0.045 **
 (0.743) (0.596) (0.034)
[expenditures.sub.(t-1)] -0.026 -0.029 0.110 ***
 (0.617) (0.601) (0.000)
[debt.sub.(t-1)] 0.008 0.008 0.012 **
 (0.544) (0.513) (0.046)
[oldage.sub.(t-1)] 0.170 0.264 0.118
 (0.624) (0.448) (0.189)
[govparty.sub.(t)] -0.130 -0.115 -0.040
 (0.229) (0.283) (0.294)
KOFgl x high [cap.sub.(t-1)] -0.035
 (0.450)
KOFec x high [cap.sub.(t-1)] -0.003
 (0.922)
KOFgl x high [lab.sub.(t-1)] -0.026
 (0.367)
KOFec x high [lab.sub.(t-1)]

KOFgl x high [con.sub.(t-1)]

KOFec x high [con.sub.(t-1)]

Trend x high_cap 0.111 ** 0.087 **
 (0.011) (0.037)
Trend x high_lab -0.021
 (0.416)
trend x high_con

high cap 3.250 1.315
 (0.292) (0.515)
high lab 2.610
 (0.138)
high con

[R.sup.2] 759 758 932
N 380 380 398
F-test TD 0.000 0.000 0.147
F-test high ITR 0.705 0.685 0.887

 (10) (11) (12)
 ITR lab ITR con ITR con

ITR [cap.sub.(t-1])]

ITR [lab.sub.t-1]) 0.694 ***
 (0.000)
ITR [con.sub.(t-1)] 0.818 *** 0.823 ***
 (0.000) (0.000)
KOF [global.sub.(t-1)] -0.009
 (0.623)
KOF [econ.sub.(t-1)] 0.054 *** 0.006
 (0.002) (0.666)
[size.sub.(t-1)] 0.344 *** -0.004 -0.030
 (0.008) (0.969) (0.744)
[growth.sub.(t-1)] -0.007 0.041 0.044
 (0.822) (0.130) (0.100
[inflation.sub.(t-1)] 0.041* 0.028 0.028
 (0.063) (0.202) (0.212)
[expenditures.sub.(t-1)] 0.103 *** 0.012 0.014
 (0.000) (0.404) (0.336)
[debt.sub.(t-1)] 0.017 *** -0.003 -0.004
 (0.003) (0.374) (0.354)
[oldage.sub.(t-1)] 0.102 0.038 -0.009
 (0.263) (0.585) (0.891)
[govparty.sub.(t)] -0.036 -0.014 -0.020
 (0.329) (0.582) (0.426)
KOFgl x high [cap.sub.(t-1)]

KOFec x high [cap.sub.(t-1)]

KOFgl x high [lab.sub.(t-1)]

KOFec x high [lab.sub.(t-1)] -0.018
 (0.202)
KOFgl x high [con.sub.(t-1)] -0.018
 (0.180)
KOFec x high [con.sub.(t-1)] -0.019 **
 (0.040)
Trend x high_cap

Trend x high_lab -0.035 **
 (0.024)
trend x high_con -0.006 -0.005
 (0.766) (0.789)
high cap

high lab 2.379 ***
 (0.009)
high con 1.809 ** 1.725 ***
 (0.025) (0.001)
[R.sup.2] 933 0.840 0.840
N 398 398 398
F-test TD 0.012 0.000 0.000
F-test high ITR 0.034 0.091 0.218

Note: Newey-West-HAC robust p values in parentheses; estimates based
on Schaffer's xtivreg2 command with bw(3) robust option; F-test TD = p
values of test of significance of time dummies, F-test high ITR = p
value of joint significance of KOF and KOF x high ITR dummy; ITR cap:
implicit tax rate on capital income, ITR lab: ITR on labor income, ITR
con: ITR on consumption.

*** p value <0.01; ** p value <0.05; * p value <0.10.

TABLE 2
The Effects of Globalization on the Share of Taxes in Total Tax
Revenues in the EU15, 1970-2007

 (1) Captax (2) Captax (3) Labtax

ITR [cap.sub.(t-1)] 0.830 *** 0.827 ***
 (0.000) (0.000)
ITR [lab.sub.(t-1)] 0.846 ***
 (0.000)
ITR [con.sub.(t-1)]

[KOFglobal.sub.(t-1)] 0.017 0.004
 (0.449) (0.887)
[KOFecon.sub.(t-1)] 0.008
 (0.728)
[size.sub.(t-1)] 0.133 0.169 0.101
 (0.528) (0.430) (0.581)
[growth.sub.(t-1)] 0.114 *** 0.111 ** -0.049
 (0.007) (0.010) (0.166)
[inflation.sub.(t-1)] -0.007 -0.009 -0.009
 (0.843) (0.782) (0.787)
[expenditures.sub.(t-1)] -0.002 -0.002 0.024
 (0.940) (0.927) (0.336)
[debt.sub.(t-1)] -0.001 0.001 0.006
 (0.927) (0.932) (0.251)
[oldage.sub.(t)] -0.062 -0.040 0.020
 (0.618) (0.747) (0.850)
[govparty.sub.(t)] -0.008 -0.003 0.023
 (0.844) (0.950) (0.568)
[R.sup.2] 0.814 0.814 0.848
N 398 398 398
F-test TD 0 0.000 0.000

 (4) Labtax (5) Contax (6) Contax

ITR [cap.sub.(t-1)]

ITR [lab.sub.(t-1)] 0.847 ***
 (0.000)
ITR [con.sub.(t-1)] 0.768 *** 0.772 ***
 (0.000) (0.000)
[KOFglobal.sub.(t-1)] -0.035
 (0.174)
[KOFecon.sub.(t-1)] 0.001 -0.025
 (0.952) (0.134)
[size.sub.(t-1)] 0.108 -0.304 ** -0.374 **
 (0.525) (0.037) (0.013)
[growth.sub.(t-1)] -0.049 -0.055 -0.046
 (0.170) (0.117) (0.189)
[inflation.sub.(t-1)] -0.010 0.015 0.020
 (0.778) (0.678) (0.574)
[expenditures.sub.(t-1)] 0.024 -0.031 -0.028
 (0.364) (0.153) (0.220)
[debt.sub.(t-1)] 0.006 -0.009* -0.011 **
 (0.228) (0.094) (0.032)
[oldage.sub.(t)] 0.025 0.114 0.072
 (0.805) (0.232) (0.438)
[govparty.sub.(t)] 0.025 -0.007 -0.022
 (0.539) (0.865) (0.572)
[R.sup.2] 0.848 0.822 0.822
N 398 398 398
F-test TD 0.000 0.000 0.000

Note: Newey-West-HAC robust p values in parentheses; estimates
based on Schaffer's xtivreg2 command with bw(3) robust option;
F-test TD = p values of test of significance of time dummies;
Captaxes: tax revenue on capital income as a share of total
tax revenues, Labtax: tax revenue on labor income as a share
of total tax revenues, Contax: tax revenue on consumption as a
share of total tax revenues.

TABLE 3
The Effects of Globalization on the Implicit Tax Rate on Capital
Income, Labor Income, and Consumption in Different Welfare
Regimes, 1970-2007

 (1) ITR (2) ITR (3) ITR
 cap cap lab

ITR [cap.sub.(t-1)] 0.690 *** 0.708
 (0.000) (0.000)
ITR [lab.sub.(t-1)] 0.655 ***
 0.000
ITR [con.sub.(t-1)]

[KOFglobal.sub.(t-1)] 0.046 0.053 *
 (0.576) (0.057)
KOFglobal x social-dem -0.413 ** -0.017
 (0.028) (0.732)
KOFglobal x [souther.sub.(t-1)] -0.089 **
 (0.130) (0.021)
KOFglobal x [liberal.sub.(t-1)] -0.159 -0.063
 (0.322) (0.165)
[KOFecon.sub.(t-1)] -0.010
 (0.897)
KOFecon x [social-dem.sub.(t-1)] -0.194 *
 (0.091)
KOFecon x [southern.sub.(t-1)] 0.047
 (0.692)
KOFecon x [liberal.sub.(t-1)] 0.291
 (0.162)
[size.sub.(t-1)] 1.092 ** 0.842 * 0.066
 (0.031) (0.072) (0.577)
[growth.sub.(t-1)] 0.199 * 0.160 0.016
 (0.075) (0.167) (0.614)
[inflation.sub.(t-1)] 0.021 -0.009 0.067 ***
 (0.837) (0.926) (0.004)
[expenditures.sub.(t-1)] 0.003 -0.003 0.117 ***
 (0.963) (0.966) (0.000)
[debt.sub.(t-1)] 0.004 0.010 0.020 ***
 (0.812) (0.599) (0.001)
[oldage.sub.(t)] 0.224 -0.174 -0.109
 (0.614) (0.700) (0.275)
[govparty.sub.(r)] -0.023 -0.008 -0.048
 (0.836) (0.943) (0.170)
Trend x social-dem 0.441 *** 0.314 ** -0.020
 (0.009) (0.029) (0.663)
trend x southern 0.349 ** 0.135 0.228 ***
 (0.016) (0.412) (0.000)
trend x liberal 0.104 -0.206 0.053
 (0.370) (0.219) (0.170)
[R.sup.2] 0.719 0.720 0.937
N 380 380 398
F-test TD 0.000 0.000 0.049
F-test social-dem 0.027 0.057 0.422
F-test southern 0.216 0.689 0.296
F-test liberal 0.399 0.107 0.802

 (4) ITR (5) ITR (6) ITR
 lab con con

ITR [cap.sub.(t-1)]

ITR [lab.sub.(t-1)] 0.627 ***
 (0.000)
ITR [con.sub.(t-1)] 0.806 *** 0.799
 (0.000) (0.000)
[KOFglobal.sub.(t-1)] -0.030
 (0.132)
KOFglobal x social-dem -0.004
 (0.906)
KOFglobal x [souther.sub.(t-1)] 0.000
 (1.000)
KOFglobal x [liberal.sub.(t-1)] -0.021
 (0.535)
[KOFecon.sub.(t-1)] 0.048 ** -0.032 **
 (0.011) (0.042)
KOFecon x [social-dem.sub.(t-1)] 0.026 -0.003
 (0.381) (0.910)
KOFecon x [southern.sub.(t-1)] -0.010 0.026
 (0.805) (0.385)
KOFecon x [liberal.sub.(t-1)] 0.017 0.055
 (0.677) (0.100)
[size.sub.(t-1)] 0.054 -0.008 -0.064
 (0.618) (0.938) (0.493)
[growth.sub.(t-1)] 0.002 0.048 * 0.049 *
 (0.954) (0.082) (0.084)
[inflation.sub.(t-1)] 0.056 ** 0.025 0.026
 (0.014) (0.294) (0.281)
[expenditures.sub.(t-1)] 0.108 *** 0.012 0.015
 (0.000) (0.405) (0.330)
[debt.sub.(t-1)] 0.023 *** -0.007 -0.007
 (0.000) (0.129) (0.133)
[oldage.sub.(t)] -0.164 -0.038 -0.070
 (0.118) (0.675) (0.445)
[govparty.sub.(r)] -0.038 0.006 -0.004
 (0.257) (0.820) (0.869)
Trend x social-dem -0.080 ** 0.008 0.018
 (0.041) (0.805) (0.576)
trend x southern 0.146 *** 0.048 0.016
 (0.006) (0.267) (0.661)
trend x liberal -0.011 -0.004 -0.047 *
 (0.745) (0.890) (0.085)
[R.sup.2] 0.938 0.841 0.841
N 398 398 398
F-test TD 0.011 0.000 0.000
F-test social-dem 0.003 0.309 0.091
F-test southern 0.300 0.350 0.813
F-test liberal 0.066 0.102 0.405

Note: Newey-West-HAC robust p values in parentheses; estimates based
on Schaffer's xtivreg2 command with bw(3) robust option; F-test TD = p
values of test of significance of time dummies; ITR cap: implicit tax
rate on capital income, ITR lab: ITR on labor income, ITR con: ITR on
consumption.

TABLE 4
The Effects of Globalization on the Implicit Tax Rate on Capital
Income, Labor Income, and Consumption in the CEE NMS and EU15,
1995-2007

 (1) (2) (3)
 ITR cap ITR cap ITR lab

ITR [cap.sub.(t-1)] 0.690 *** 0.703 ***
 (0.000) (0.000)
ITR [lab.sub.(t-1)] 0.521 ***
 (0.000
ITR [con.sub.(t-1)]

[KOFglobal.sub.(t-1)] -0.172 0.028
 (0.148) (0.432)
KOFglobal x [ceec.sub.(t-1)] 0.214 -0.103
 (0.173) (0.171)
[KOFecon.sub.(t-1)] -0.163 *
 (0.095)
KOFecon x [ceec.sub.(t-1)] 0.219 *
 (0.072)
[size.sub.(t-1)] 1.025 0.722 0.965 ***
 (0.164) (0.314) (0.000
[growth.sub.(t-1)] 0.197 ** 0.227 *** -0.034
 (0.010 (0.003) (0.422)
[inflation.sub.(t-1)] 0.040 0.050 * -0.059 ***
 (0.121) (0.081) (0.001)
[expenditures.sub.(t-1)] 0.087 0.073 0.062 **
 (0.271) (0.356) (0.037)
[debt.sub.(t-1)] -0.002 0.001 0.039 ***
 (0.921) (0.975) (0.000)
[oldage.sub.(t)] 0.045 0.060 0.045
 (0.916) (0.884) (0.777)
[govparty.sub.(r)] 0.145 0.122 0.040
 (0.201) (0.288) (0.391)
KOFgl x high [cap.sub.t-1])

KOFec x high [cap.sub.t-1])

KOFgl x high [lab.sub.t-1)]

KOFec x high [lab.sub.t-1)]

KOFgl x high [con.sub.t-1)]

KOFec x high [con.sub.t-1)]

trend x high_cap

trend x high_lab

trend x high_con

trend x ceec -0.144 -0.166 -0.184 *
 (0.477) (0.393) (0.095)
high cap

high lab

high con

[R.sup.2] 0.623 0.626 0.694
N 279 279 300
F-test TD 0.004 0.006 0.659
F-test CEE NMS 0.718 0.413 0.244
F-test high ITR EU15
F-test high ITR CEE

 (4) (5) (6)
 ITR lab ITR con ITR con

ITR [cap.sub.(t-1)]

ITR [lab.sub.(t-1)] 0.510 ***
 (0.000)
ITR [con.sub.(t-1)] 0.641 *** 0.634 ***
 (0.000) -0
[KOFglobal.sub.(t-1)] 0.016
 (0.595)
KOFglobal x [ceec.sub.(t-1)] -0.114 **
 (0.050
[KOFecon.sub.(t-1)] 0.048 * 0.034
 (0.058) -0.127
KOFecon x [ceec.sub.(t-1)] -0.029 -0.094 ***
 (0.508) -0.009
[size.sub.(t-1)] 1.012 *** -0.520 ** -0.543 ***
 (0.000 (0.021) -0.01
[growth.sub.(t-1)] -0.039 0.002 -0.009
 (0.382) (0.939) -0.78
[inflation.sub.(t-1)] -0.059 *** 0.002 -0
 (0.001) (0.888) -0.975
[expenditures.sub.(t-1)] 0.070 ** 0.045 ** 0.047 **
 (0.027) (0.049) -0.044
[debt.sub.(t-1)] 0.037 *** -0.011 -0.012
 (0.000) (0.172) -0.153
[oldage.sub.(t)] 0.025 -0.290 * -0.370 **
 (0.875) (0.065) -0.015
[govparty.sub.(r)] 0.046 -0.071 * -0.068 *
 (0.289) (0.060 -0.061
KOFgl x high [cap.sub.t-1])

KOFec x high [cap.sub.t-1])

KOFgl x high [lab.sub.t-1)]

KOFec x high [lab.sub.t-1)]

KOFgl x high [con.sub.t-1)]

KOFec x high [con.sub.t-1)]

trend x high_cap

trend x high_lab

trend x high_con

trend x ceec -0.316 *** 0.282 *** 0.283 ***
 (0.001) (0.002) -0.001
high cap

high lab

high con

[R.sup.2] 0.695 0.639 0.644
N 300 300 300
F-test TD 0.622 0.000 0
F-test CEE NMS 0.561 0.059 0.042
F-test high ITR EU15
F-test high ITR CEE

 (7) (7) (9)
 ITR cap ITR cap ITR lab

ITR [cap.sub.(t-1)] 0.597 *** 0.607 ***
 (0.000) (0.000)
ITR [lab.sub.(t-1)] 0.513 ***
 (0.000)
ITR [con.sub.(t-1)]

[KOFglobal.sub.(t-1)] -0.186 0.015
 (0.122) (0.665)
KOFglobal x [ceec.sub.(t-1)] 0.286 * -0.097
 (0.053) (0.180)
[KOFecon.sub.(t-1)] -0.143 *
 (0.091)
KOFecon x [ceec.sub.(t-1)] 0.237 **
 (0.030)
[size.sub.(t-1)] 0.325 -0.208 0.992 ***
 (0.629) (0.772) (0.000)
[growth.sub.(t-1)] 0.142 * 0.171 ** -0.041
 (0.053) (0.017) (0.355)
[inflation.sub.(t-1)] 0.053 ** 0.067 ** -0.062 ***
 (0.039) (0.028) (0.000
[expenditures.sub.(t-1)] 0.065 0.045 0.056 *
 (0.363) (0.519) (0.069)
[debt.sub.(t-1)] -0.002 0.004 0.038 ***
 (0.907) (0.825) (0.000)
[oldage.sub.(t)] 0.084 -0.006 0.036
 (0.843) (0.988) (0.824)
[govparty.sub.(r)] 0.172 0.150 0.039
 (-0.107) (-0.162) (0.378)
KOFgl x high [cap.sub.t-1]) 0.003
 (0.971)
KOFec x high [cap.sub.t-1]) -0.054
 (0.330)
KOFgl x high [lab.sub.t-1)] 0.034
 (0.150)
KOFec x high [lab.sub.t-1)]

KOFgl x high [con.sub.t-1)]

KOFec x high [con.sub.t-1)]

trend x high_cap 0.002 0.030
 (0.985) (0.783)
trend x high_lab -0.052
 (0.188)
trend x high_con

trend x ceec -0.279 -0.249 -0.204 *
 (0.166) (0.192) (0.057)
high cap 3.092 6.687
 (0.477) (0.109)
high lab -0.842
 (0.704)
high con

[R.sup.2] 0.666 0.671 0.698
N 278 278 299
F-test TD 0.006 0.009 0.474
F-test CEE NMS 0.341 0.127 0.176
F-test high ITR EU15 0.129 0.088 0.218
F-test high ITR CEE 0.366 0.608 0.458

 (10) (11) (12)
 ITR lab ITR con ITR con

ITR [cap.sub.(t-1)]

ITR [lab.sub.(t-1)] 0.496 ***
 (0.000)
ITR [con.sub.(t-1)] 0.591 *** 0.579 ***
 (0.000 (0.000)
[KOFglobal.sub.(t-1)] 0.004
 (0.917)
KOFglobal x [ceec.sub.(t-1)] -0.127 **
 (0.019)
[KOFecon.sub.(t-1)] 0.046* 0.072 ***
 (0.062) (0.006)
KOFecon x [ceec.sub.(t-1)] -0.021 -0.106 ***
 (0.645) (0.001)
[size.sub.(t-1)] 1.051 *** -1.133 *** -1.212 ***
 (0.000) (0.000) (0.000)
[growth.sub.(t-1)] -0.051 0.008 -0.009

 (0.258) (0.783) (0.778)
[inflation.sub.(t-1)] -0.062 *** 0.006 0.005
 (0.001) (0.637) (0.674)
[expenditures.sub.(t-1)] 0.059 * 0.029 0.035
 (0.069) (0.187) (0.121)
[debt.sub.(t-1)] 0.039 *** -0.010 -0.009
 (0.000 (0.176) (0.262)
[oldage.sub.(t)] 0.013 -0.405 ** -0.625 ***
 (0.939) (0.015) (0.000)
[govparty.sub.(r)] 0.044 -0.083 ** -0.060 *
 (0.299) (0.021) (0.088)
KOFgl x high [cap.sub.t-1])

KOFec x high [cap.sub.t-1])

KOFgl x high [lab.sub.t-1)]

KOFec x high [lab.sub.t-1)] 0.007
 (0.651)
KOFgl x high [con.sub.t-1)] -0.022
 (0.609)
KOFec x high [con.sub.t-1)] -0.063 ***
 (0.002)
trend x high_cap

trend x high_lab -0.037
 (0.382)
trend x high_con -0.118 *** -0.068 *
 (0.003) (0.079)
trend x ceec -0.352 *** 0.329 *** 0.309 ***
 (0.000) (0.000) (0.000)
high cap

high lab 0.990
 (0.541)
high con 5.549 ** 7.041 ***
 (0.031) (0.000)
[R.sup.2] 0.698 0.667 0.686
N 299 299 299
F-test TD 0.375 0.000 0.000
F-test CEE NMS 0.450 0.009 0.232
F-test high ITR EU15 0.065 0.662 0.680
F-test high ITR CEE 0.360 0.016 0.002

Note: Newey-West-HAC robust p values in parentheses; estimates based
on Schaffer's xtivreg2 command with bw(3) robust option; F-test TD = p
values of test of significance of time dummies; F-test CEE NMS = p
value of joint significance of KOF and KOF x ceec; F-test high ITR
EU15 = p value of joint significance of KOF and KOF x high ITR dummy;
F-test high ITR CEE = p value of joint significance of KOF, KOF x
CEEC, and KOF x high ITR dummy; ITR cap: implicit tax rate on capital
income, ITR lab: ITR on labor income, ITR con: ITR on consumption.

TABLE 5
The Effects of Globalization on the Share of Taxes in Total Tax
Revenues in the CEE NMS and EU15, 1995-2007

 (1) Captax (2) Captax (3) Labtax

ITR [cap.sub.(t-1)] 0.651 *** 0.660 ***
 (0.000) (0.000)
ITR [lab.sub.(t-1)] 0.616 ***
 (0.000)
ITR [con.sub.(t-1)]

[KOFglobal.sub.(t-1)] -0.068 0.029
 (0.273) (0.520)
KOFglobal*[ceec.sub.(t-1)] 0.134 -0.028
 (0.101) (0.693)
[KOFecon.sub.(t-1)] -0.103 **
 (0.029)
KOFecon*[ceec.sub.(t-1)] 0.166 ***
 (0.004)
[size.sub.(t-1)] 0.105 0.039 1.003 ***
 (0.787) (0.915) (0.004)
[growth.sub.(t-1)] 0.049 0.072 -0.003
 (0.280) (0.115) (0.93l)
[inflation.sub.(t-1)] -0.011 -0.007 0.007
 (0.285) (0.518) (0.425)
[expenditures.sub.(t-1)] -0.055 * -0.060 ** 0.088 ***
 (0.062) (0.034) (0.00l)
[debt.sub.(t-1)] -0.003 -0.002 0.024 **
 (0.802) (0.894) (0.013)
oldage(,) 0.016 0.125 0.167
 (0.934) (0.487) (0.354)
govparty),) 0.151 *** 0.147 *** -0.057
 (0.002) (0.004) (0.265)
trend*ceec -0.094 -0.148 * -0.205 *
 (0.352) (0.096) (0.074)
[R.sup.2] 0.613 0.626 0.728
N 302 302 302
F-test TD 0.001 0.003 0.000
F-test CEE NMS 0.261 0.052 0.990

 (4) Labtax (5) Contax (6) Contax

ITR [cap.sub.(t-1)]

ITR [lab.sub.(t-1)] 0.620 ***
 (0.000)
ITR [con.sub.(t-1)] 0.64l *** 0.647 ***
 (0.000) (0.000)
[KOFglobal.sub.(t-1)] 0.024
 (0.555)
KOFglobal*[ceec.sub.(t-1)] -0.078
 (0.309)
[KOFecon.sub.(t-1)] 0.026 0.063 **
 (0.483) (0.02l)
KOFecon*[ceec.sub.(t-1)] -0.028 -0.123 **
 (0.590) (0.014)
[size.sub.(t-1)] 1.045 *** -1.088 *** -1.089 ***
 (0.002) (0.000) (0.000)
[growth.sub.(t-1)] -0.005 -0.047 -0.066
 (0.882) (0.348) (0.202)
[inflation.sub.(t-1)] 0.007 0.001 -0.003
 (0.455) (0.898) (0.735)
[expenditures.sub.(t-1)] 0.090 *** -0.031 -0.028
 (0.00l) (0.330) (0.386)
[debt.sub.(t-1)] 0.024 ** -0.015 -0.015
 (0.015) (0.222) (0.217)
oldage(,) 0.166 -0.160 -0.268 *
 (0.339) (0.327) (0.083)
govparty),) -0.054 -0.108 ** -0.109 ***
 (0.286) (0.014) (0.010)
trend*ceec -0.203 ** 0.253 ** 0.323 ***
 (0.036) (0.030) (0.002)
[R.sup.2] 0.728 0.617 0.626
N 302 302 302
F-test TD 0.000 0.045 0.037
F-test CEE NMS 0.959 0.426 0.127

Note: Newey-West-HAC robust p values in parentheses; estimates based
on Schaffer's xtivreg2 command with bw(3) robust option; F-test TD = p
values of test of significance of time dummies; Captaxes: tax revenue
on capital income as a share of total tax revenues; Labtax: tax
revenue on labor income as a share of total tax revenues; Contax: tax
revenue on consumption as a share of total tax revenues.