Citation Context

...vidence on the relationship between country size and national corporate tax rates is largely absent from the literature. Empirical papers considering the determinants of corporate tax rates focus mainly on two aspects of tax competition. Some look at the strategic interactions between countries, resulting in interdependent corporate tax rates (Devereux et al., 2008; Overesch and Rincke, 2011), while others examine the effect of globalisation (proxied by measures of economic and financial openness or capital mobility), which is expected to increase competitive pressures on corporate tax rates (Rodrik, 1997; Bretschger and Hettich, 2002; Swank and Steinmo, 2002; Slemrod, 2004; Winner, 2005). Country size is not the main focus of these empirical analyses, although it is generally included in the econometric models as a control variable. The coefficient estimated on country size exhibits a positive sign, but its statistical significance is not consistently established. Given the various ambiguities present in the existing literature, the purpose of this paper is to analyse the theoretical and empirical effects of country size on the level of corporate tax rates and to determine whether the data ca...

Citation Context

...Therefore, if capital is less sensitive to tax change in the large country - as shown by Bucovetsky (1991) 22By definition, these fixed effects include the averages of omitted time-varying variables which influence corporate tax rates. 20 and Wilson (1991) - the large country will compete less vigorously for capital through tax rate reductions and therefore will set a higher tax rate than the small country. Empirically, the size of a country, or the extent to which a country is price-maker in the world capital market, can be measured by its position in the international financial integration (Lane and Milesi-Ferretti, 2007). Following Azémar and Hubbard (forthcoming),23 we use a measure of external financial assets position provided by Lane and Milesi-Ferretti (2007) updated and extended version of External Wealth of Nations Database to capture a country’s ability to influence the world return to capital. The Financial assets/world financial assets variable corresponds to the ratio of the stock of financial assets of a country i to the stock of the world financial assets. Financial assets are composed by outward portfolio investment (subdivided into equity securities and debt securities), foreign direct investme...

Citation Context

...; F23; H25. ∗We are grateful to Michael Devereux, Andreas Haufler, and Andreas Hoefele for excellent comments and suggestions. We would like to acknowledge the valuable feedback provided by participants at the European Trade Study Group conference in Birmingham, at the joint conference of ZEW and Oxford Centre for Business Taxation on Taxing Multinational Firms in Mannheim, at the CESifo Area Conference on Global Economy, and at the Munich International Economics seminar. 1 1 Introduction Tax competition models suggest that national corporate tax rates are strongly influenced by country size (Bucovetsky, 1991; Wilson, 1991; Haufler and Wooton, 1999). While these models share the prediction that larger countries set higher taxes than smaller countries, they propose different theoretical explanations for this positive relationship. Bucovetsky (1991) and Wilson (1991) consider that small countries face a more elastic capital supply with respect to changes in corporate tax rates (as they are price takers in the world capital market) than larger countries, leading them to set a lower corporate tax rate. Devereux et al. (2008) argue that larger countries have a higher proportion of domestic and conseque...

Citation Context

... allow identification of the spatial lag term, time dummies are replaced with country-specific time trends. 19Similar results are obtained when omitting the average of other countries’ GDP as an instrument. 20This transformation gives the same estimates as a fixed effects estimator when using OLS and, contrary to the within transformation, allows us to use lagged values of our explanatory variables as potential instruments in an IV context. For more information, see Arellano (2003). 21We construct our instruments “collapsed GMM-style” in the sense that missing values are replaced by zero. See Roodman (2009). 17 Table 3: Market Size and Corporate Income Taxes Dependent variable: Statutory tax rate Baseline Size and Size and Size and Common Endogeneity equation time trend decade dummies Neighbours STR correlated effects IV [1] [2] [3] [4] [5] [6] Size 0.1750∗∗∗ 0.1951∗∗∗ 0.1926∗∗∗ 0.1419∗∗ 0.1553∗∗ 0.0914∗ (0.0330) (0.0394) (0.0412) (0.0575) (0.0715) (0.0540) Population proportion young 0.0113∗∗∗ 0.0114∗∗∗ 0.0109∗∗∗ -0.0078∗∗ -0.0246∗∗ 0.0085∗∗∗ (0.0026) (0.0026) (0.0026) (0.0037) (0.0097) (0.0032) Population proportion old 0.0335∗∗∗ 0.0332∗∗∗ 0.0332∗∗∗ -0.0061 -0.0119 0.0278∗∗∗ (0.0050) (0.0051) ...

Citation Context

... suggest that national corporate tax rates are strongly influenced by country size (Bucovetsky, 1991; Wilson, 1991; Haufler and Wooton, 1999). While these models share the prediction that larger countries set higher taxes than smaller countries, they propose different theoretical explanations for this positive relationship. Bucovetsky (1991) and Wilson (1991) consider that small countries face a more elastic capital supply with respect to changes in corporate tax rates (as they are price takers in the world capital market) than larger countries, leading them to set a lower corporate tax rate. Devereux et al. (2008) argue that larger countries have a higher proportion of domestic and consequently less mobile activity that governments prefer to tax at higher rates than foreign activity. Lastly, Haufler and Wooton (1999) highlight that a large country is a more attractive location for investment than a smaller country, allowing the former to set a higher corporate tax rate than the latter. Empirically, robust evidence on the relationship between country size and national corporate tax rates is largely absent from the literature. Empirical papers considering the determinants of corporate tax rates focus mai...

Citation Context

...Michael Devereux, Andreas Haufler, and Andreas Hoefele for excellent comments and suggestions. We would like to acknowledge the valuable feedback provided by participants at the European Trade Study Group conference in Birmingham, at the joint conference of ZEW and Oxford Centre for Business Taxation on Taxing Multinational Firms in Mannheim, at the CESifo Area Conference on Global Economy, and at the Munich International Economics seminar. 1 1 Introduction Tax competition models suggest that national corporate tax rates are strongly influenced by country size (Bucovetsky, 1991; Wilson, 1991; Haufler and Wooton, 1999). While these models share the prediction that larger countries set higher taxes than smaller countries, they propose different theoretical explanations for this positive relationship. Bucovetsky (1991) and Wilson (1991) consider that small countries face a more elastic capital supply with respect to changes in corporate tax rates (as they are price takers in the world capital market) than larger countries, leading them to set a lower corporate tax rate. Devereux et al. (2008) argue that larger countries have a higher proportion of domestic and consequently less mobile activity that government...

Citation Context

...e grateful to Michael Devereux, Andreas Haufler, and Andreas Hoefele for excellent comments and suggestions. We would like to acknowledge the valuable feedback provided by participants at the European Trade Study Group conference in Birmingham, at the joint conference of ZEW and Oxford Centre for Business Taxation on Taxing Multinational Firms in Mannheim, at the CESifo Area Conference on Global Economy, and at the Munich International Economics seminar. 1 1 Introduction Tax competition models suggest that national corporate tax rates are strongly influenced by country size (Bucovetsky, 1991; Wilson, 1991; Haufler and Wooton, 1999). While these models share the prediction that larger countries set higher taxes than smaller countries, they propose different theoretical explanations for this positive relationship. Bucovetsky (1991) and Wilson (1991) consider that small countries face a more elastic capital supply with respect to changes in corporate tax rates (as they are price takers in the world capital market) than larger countries, leading them to set a lower corporate tax rate. Devereux et al. (2008) argue that larger countries have a higher proportion of domestic and consequently less mobi...

Citation Context

...ing the presence of multiple exchange rates (k1); a variable indicating restrictions on current account transactions (k2); a variable indicating restrictions on capital account transactions (k3);14 and a variable indicating the requirement of the surrender of export proceeds (k4). This variable is a de jure measure of capital openness as it is coded on the basis of regulatory restrictions on capital. 12 A country which is located close to large foreign markets should be able to set a higher tax rate than other countries. We test two measures of foreign market potential. The first proxy is the Harris (1954) market potential. This variable is the sum of foreign GDPs weighted by the inverse of bilateral distance. The second proxy is the Redding and Venables (2004) foreign market access. It can be interpreted as the sum of countries’ propensities to demand imports weighted by the inverse of bilateral trade costs. We also include the sum of the inverse of bilateral trade costs, because greater trade accessibility can induce a country to lower its tax rate. This measure of trade accessibility varies over time. It includes, in addition to the effects of time-invariant geographic characteristics (e.g. ...

Citation Context

...the country GDP 11 3.3 Control Variables The choice of the control variables is based on the existing literature, e.g. Slemrod (2004), Devereux et al. (2008). The age and the geographical distribution of the population could influence national tax policies. We therefore include as control variables the proportion of young, corresponding to the proportion of the population below 14 years old, the proportion of old, corresponding to the proportion of the population above 65 years old, and the proportion urban, corresponding to the proportion of the population living in urban areas. As argued by Gordon and Slemrod (2000) and Slemrod (2004), the corporate income tax serves as a “backstop” for the personal income tax, since it ensures that individuals do not have an incentive to escape personal income taxes by reclassifying their income as corporate income. We take into account this backstop role of the corporate income tax by including the top personal income tax rate among the control variables. Finally, we control for the revenue needs of a government by including a measure of government spending, corresponding to the general government consumption expenditure in percentage of GDP. 3.4 Additional Variables I...

Citation Context

... of a positive relationship between country size and the statutory tax rate. In Column [1], the coefficient of 0.175 indicates that doubling market size increases the statutory tax rate by about ln(2) × 0.175 = 12 percentage points.17 This is a large effect. The 16We have not used clustered standard errors for three reasons. First, we did not find evidence of serial correlation. Second, our number of clusters is too small. Finally, our time period appears long enough for not being concerned by the bias in the conventional heteroskedasticity-robust (HR) variance matrix estimator highlighted by Stock and Watson (2008). Similar estimates for the standard errors are obtained when their bias-adjusted HR estimator is applied. 17For illustration purposes, in our sample, Belgium has twice the GDP of Ireland, Portugal has twice the GDP of Hungary, and Germany almost twice the GDP of Italy, in 2005. Between 1981 and 2005, the United Kingdom has doubled its market size and Ireland has more than tripled its market size. 15 coefficient estimated is an average obtained with three decades of data, 1981-2005, which have witnessed an increase in international pressures for tax competition and a decrease in corporate tax ...

Citation Context

...ed by the statutory tax rate (STR). This measure presents a number of advantages. As emphasised by Overesch and Rincke (2011), it is the simplest indicator of expected tax payments for firms and it is readily available across countries and years. However, while the statutory tax rate is the most relevant variable for the determination of income shifting, it neglects the tax base, and thus, it is an imperfect indicator for investment and the location of firms. More complex measures of effective tax rates take into account the role of the base of the corporation income tax but, as emphasised by Swenson (1994) and Slemrod (2004), these measures rely on a number of arbitrary assumptions about economic parameters and do not consider some characteristics of the tax system that can influence the decision of the firms such as the degree of enforcement of the tax system. The existing empirical literature provides some evidence about how meaningful the statutory tax rate is for the decision of the firms and for government strategies to compete for firms and investment. Most papers working on the relationships between the statutory tax rates and FDI find a negative and statistically significant elasticity ...

Citation Context

... relationship between country size and national corporate tax rates is largely absent from the literature. Empirical papers considering the determinants of corporate tax rates focus mainly on two aspects of tax competition. Some look at the strategic interactions between countries, resulting in interdependent corporate tax rates (Devereux et al., 2008; Overesch and Rincke, 2011), while others examine the effect of globalisation (proxied by measures of economic and financial openness or capital mobility), which is expected to increase competitive pressures on corporate tax rates (Rodrik, 1997; Bretschger and Hettich, 2002; Swank and Steinmo, 2002; Slemrod, 2004; Winner, 2005). Country size is not the main focus of these empirical analyses, although it is generally included in the econometric models as a control variable. The coefficient estimated on country size exhibits a positive sign, but its statistical significance is not consistently established. Given the various ambiguities present in the existing literature, the purpose of this paper is to analyse the theoretical and empirical effects of country size on the level of corporate tax rates and to determine whether the data can distinguish between competin...

Citation Context

...derivatives’ contracts). To avoid an endogeneity bias, we lag this control variable by two years. In that way, given the absence of serial correlation of the errors, it should not be correlated with the contemporaneous error term. Devereux et al. (2008) assume that governments might prefer to tax domestic activities at high corporate tax rates, probably because these activities are less mobile. This relates to the debate of whether all profits should be taxed at the same corporate tax rate or whether some type of capital, such as the most mobile, should benefit from a favorable tax treatment (Keen, 2001; Haupt and Peters, 2005; Janeba and Smart, 2003; Marceau et al., 2010; Wilson, 2005). Since large countries have a higher proportion of economic activity that is domestic, they might have a stronger incentive to tax this activity at a high rate. However, as emphasised by Devereux et al. (2008), if domestic activities are taxed more heavily than foreign activities, large countries will be constrained by pressures from competitive countries.24 Thus, they assume that the higher the domestic part of the economy, the higher the country would set the “overall” corporate tax rate. Different measures...

Citation Context

...size and national corporate tax rates is largely absent from the literature. Empirical papers considering the determinants of corporate tax rates focus mainly on two aspects of tax competition. Some look at the strategic interactions between countries, resulting in interdependent corporate tax rates (Devereux et al., 2008; Overesch and Rincke, 2011), while others examine the effect of globalisation (proxied by measures of economic and financial openness or capital mobility), which is expected to increase competitive pressures on corporate tax rates (Rodrik, 1997; Bretschger and Hettich, 2002; Swank and Steinmo, 2002; Slemrod, 2004; Winner, 2005). Country size is not the main focus of these empirical analyses, although it is generally included in the econometric models as a control variable. The coefficient estimated on country size exhibits a positive sign, but its statistical significance is not consistently established. Given the various ambiguities present in the existing literature, the purpose of this paper is to analyse the theoretical and empirical effects of country size on the level of corporate tax rates and to determine whether the data can distinguish between competing theoretical explanation...

Citation Context

... largely absent from the literature. Empirical papers considering the determinants of corporate tax rates focus mainly on two aspects of tax competition. Some look at the strategic interactions between countries, resulting in interdependent corporate tax rates (Devereux et al., 2008; Overesch and Rincke, 2011), while others examine the effect of globalisation (proxied by measures of economic and financial openness or capital mobility), which is expected to increase competitive pressures on corporate tax rates (Rodrik, 1997; Bretschger and Hettich, 2002; Swank and Steinmo, 2002; Slemrod, 2004; Winner, 2005). Country size is not the main focus of these empirical analyses, although it is generally included in the econometric models as a control variable. The coefficient estimated on country size exhibits a positive sign, but its statistical significance is not consistently established. Given the various ambiguities present in the existing literature, the purpose of this paper is to analyse the theoretical and empirical effects of country size on the level of corporate tax rates and to determine whether the data can distinguish between competing theoretical explanations for a positive relationship ...

Citation Context

... contracts). To avoid an endogeneity bias, we lag this control variable by two years. In that way, given the absence of serial correlation of the errors, it should not be correlated with the contemporaneous error term. Devereux et al. (2008) assume that governments might prefer to tax domestic activities at high corporate tax rates, probably because these activities are less mobile. This relates to the debate of whether all profits should be taxed at the same corporate tax rate or whether some type of capital, such as the most mobile, should benefit from a favorable tax treatment (Keen, 2001; Haupt and Peters, 2005; Janeba and Smart, 2003; Marceau et al., 2010; Wilson, 2005). Since large countries have a higher proportion of economic activity that is domestic, they might have a stronger incentive to tax this activity at a high rate. However, as emphasised by Devereux et al. (2008), if domestic activities are taxed more heavily than foreign activities, large countries will be constrained by pressures from competitive countries.24 Thus, they assume that the higher the domestic part of the economy, the higher the country would set the “overall” corporate tax rate. Different measures of the share of domesti...

Citation Context

... endogeneity bias, we lag this control variable by two years. In that way, given the absence of serial correlation of the errors, it should not be correlated with the contemporaneous error term. Devereux et al. (2008) assume that governments might prefer to tax domestic activities at high corporate tax rates, probably because these activities are less mobile. This relates to the debate of whether all profits should be taxed at the same corporate tax rate or whether some type of capital, such as the most mobile, should benefit from a favorable tax treatment (Keen, 2001; Haupt and Peters, 2005; Janeba and Smart, 2003; Marceau et al., 2010; Wilson, 2005). Since large countries have a higher proportion of economic activity that is domestic, they might have a stronger incentive to tax this activity at a high rate. However, as emphasised by Devereux et al. (2008), if domestic activities are taxed more heavily than foreign activities, large countries will be constrained by pressures from competitive countries.24 Thus, they assume that the higher the domestic part of the economy, the higher the country would set the “overall” corporate tax rate. Different measures of the share of domestic or foreign activities ...

Citation Context

...te tax rates is largely absent from the literature. Empirical papers considering the determinants of corporate tax rates focus mainly on two aspects of tax competition. Some look at the strategic interactions between countries, resulting in interdependent corporate tax rates (Devereux et al., 2008; Overesch and Rincke, 2011), while others examine the effect of globalisation (proxied by measures of economic and financial openness or capital mobility), which is expected to increase competitive pressures on corporate tax rates (Rodrik, 1997; Bretschger and Hettich, 2002; Swank and Steinmo, 2002; Slemrod, 2004; Winner, 2005). Country size is not the main focus of these empirical analyses, although it is generally included in the econometric models as a control variable. The coefficient estimated on country size exhibits a positive sign, but its statistical significance is not consistently established. Given the various ambiguities present in the existing literature, the purpose of this paper is to analyse the theoretical and empirical effects of country size on the level of corporate tax rates and to determine whether the data can distinguish between competing theoretical explanations for a positiv...

Citation Context

...ount the tax rates of other countries when setting their own corporate tax rate. The instrument diagnostics indicate that our instruments are relevant and exogenous.19 The coefficient on country size is not qualitatively affected by the inclusion of this variable. It remains large, positive and statistically significant, indicating that the setting of corporate income tax rate is truly related to country size. As an alternative to control for a potential omitted variable bias related to spatial interdependence or unobserved common factors, we employ the Common Correlated Effects (CCE) method (Chudik et al., 2011; Pesaran and Tosetti, 2011) in Column [5]. This method, which consists in including the annual cross-section averages of each variable of the model, presents the advantage that we can control for cross-section dependence of any kind (e.g. global shocks affecting countries with different degrees or spatial spillovers) without having to model this process explicitly. Column [5] shows that using this approach does not change our previous results. We also consider a potential endogeneity bias between the level of the statutory tax rate and country size. In line with basic economic theory, empiric...

Citation Context

...e by two years. In that way, given the absence of serial correlation of the errors, it should not be correlated with the contemporaneous error term. Devereux et al. (2008) assume that governments might prefer to tax domestic activities at high corporate tax rates, probably because these activities are less mobile. This relates to the debate of whether all profits should be taxed at the same corporate tax rate or whether some type of capital, such as the most mobile, should benefit from a favorable tax treatment (Keen, 2001; Haupt and Peters, 2005; Janeba and Smart, 2003; Marceau et al., 2010; Wilson, 2005). Since large countries have a higher proportion of economic activity that is domestic, they might have a stronger incentive to tax this activity at a high rate. However, as emphasised by Devereux et al. (2008), if domestic activities are taxed more heavily than foreign activities, large countries will be constrained by pressures from competitive countries.24 Thus, they assume that the higher the domestic part of the economy, the higher the country would set the “overall” corporate tax rate. Different measures of the share of domestic or foreign activities in total economic activity can be use...

Citation Context

...ty. Lastly, Haufler and Wooton (1999) highlight that a large country is a more attractive location for investment than a smaller country, allowing the former to set a higher corporate tax rate than the latter. Empirically, robust evidence on the relationship between country size and national corporate tax rates is largely absent from the literature. Empirical papers considering the determinants of corporate tax rates focus mainly on two aspects of tax competition. Some look at the strategic interactions between countries, resulting in interdependent corporate tax rates (Devereux et al., 2008; Overesch and Rincke, 2011), while others examine the effect of globalisation (proxied by measures of economic and financial openness or capital mobility), which is expected to increase competitive pressures on corporate tax rates (Rodrik, 1997; Bretschger and Hettich, 2002; Swank and Steinmo, 2002; Slemrod, 2004; Winner, 2005). Country size is not the main focus of these empirical analyses, although it is generally included in the econometric models as a control variable. The coefficient estimated on country size exhibits a positive sign, but its statistical significance is not consistently established. Given the vario...

Citation Context

...country to lower its tax rate. This measure of trade accessibility varies over time. It includes, in addition to the effects of time-invariant geographic characteristics (e.g. distance, common language) on international trade, those of time-varying memberships in free trade and common currency associations.15 We finally consider the theoretical prediction that a less competitive market is more profitable to firms choosing to locate in the country by adding a proxy for the Intensity of local competition. This measure is based on a survey conducted by World Economic Forum (2013) and provided by Teorell et al. (2013), in the Quality of Government Standard Dataset. Randomly selected firms representing the main sectors of the economy, were asked to assess, on a scale of 1 (limited) to 7 (intense), the intensity of competition in the local markets in which they operate. We use the country score for the period 2011-2012. To deal with a potential endogeneity bias, the variable Effectiveness of anti-monopoly policy from the same source will be used as an instrumental variable. Respondents had to evaluate, to what extent anti-monopoly policy promotes competition in their country, on a scale of 1 (does not promot...

Citation Context

...le (IV) approach is necessary because corporate tax rates are expected to be jointly determined, creating a simultaneity bias. To allow identification of the spatial lag term, time dummies are replaced with country-specific time trends. 19Similar results are obtained when omitting the average of other countries’ GDP as an instrument. 20This transformation gives the same estimates as a fixed effects estimator when using OLS and, contrary to the within transformation, allows us to use lagged values of our explanatory variables as potential instruments in an IV context. For more information, see Arellano (2003). 21We construct our instruments “collapsed GMM-style” in the sense that missing values are replaced by zero. See Roodman (2009). 17 Table 3: Market Size and Corporate Income Taxes Dependent variable: Statutory tax rate Baseline Size and Size and Size and Common Endogeneity equation time trend decade dummies Neighbours STR correlated effects IV [1] [2] [3] [4] [5] [6] Size 0.1750∗∗∗ 0.1951∗∗∗ 0.1926∗∗∗ 0.1419∗∗ 0.1553∗∗ 0.0914∗ (0.0330) (0.0394) (0.0412) (0.0575) (0.0715) (0.0540) Population proportion young 0.0113∗∗∗ 0.0114∗∗∗ 0.0109∗∗∗ -0.0078∗∗ -0.0246∗∗ 0.0085∗∗∗ (0.0026) (0.0026) (0.0026)...

Citation Context

... process explicitly. Column [5] shows that using this approach does not change our previous results. We also consider a potential endogeneity bias between the level of the statutory tax rate and country size. In line with basic economic theory, empirical work such as Djankov et al. (2010) finds that corporate taxes have a large negative effect on aggregate investment, FDI, and entrepreneurial activity. This detrimental effect on investment can in turn affect economic growth. To allow for this potential endogeneity between corporate taxes and country size, we re-estimate equation (15) using an Anderson and Hsiao (1981) style IV approach. To keep consistent with our fixed effects approach, we apply the forward orthogonal deviations transformation to the data20 and use as instruments the second and third lags of GDP in Column [6].21 Values for the first-stage F - statistic and for the test of overidentifying restrictions suggest that the instruments are relevant and exogenous. Our IV results are in line with our previous results: the coefficient on market size is positive and statistically significant determinant at the 10% level. 4.2 Additional Determinants We consider additional determinants in Table 4. The...

Citation Context

...g this control variable by two years. In that way, given the absence of serial correlation of the errors, it should not be correlated with the contemporaneous error term. Devereux et al. (2008) assume that governments might prefer to tax domestic activities at high corporate tax rates, probably because these activities are less mobile. This relates to the debate of whether all profits should be taxed at the same corporate tax rate or whether some type of capital, such as the most mobile, should benefit from a favorable tax treatment (Keen, 2001; Haupt and Peters, 2005; Janeba and Smart, 2003; Marceau et al., 2010; Wilson, 2005). Since large countries have a higher proportion of economic activity that is domestic, they might have a stronger incentive to tax this activity at a high rate. However, as emphasised by Devereux et al. (2008), if domestic activities are taxed more heavily than foreign activities, large countries will be constrained by pressures from competitive countries.24 Thus, they assume that the higher the domestic part of the economy, the higher the country would set the “overall” corporate tax rate. Different measures of the share of domestic or foreign activities in total economic acti...