Trade integration and corporate income tax differentials

Abstract

Building on recent contributions to the New Economic Geography literature, this paper analyses the relation between asymmetric market size, trade integration and corporate income tax differentials across countries. First, relying on Ottaviano and Van Ypersele's (2005) footloose capital model of tax competition, we illustrate that trade integration reduces the importance of relative market size for differences in the extent of corporate taxation between countries. Then, using a dataset of 26 OECD countries over the period 1982-2004, we provide supportive evidence of these theoretical predictions: i.e., market size differences are strongly positively correlated with corporate income tax differences across countries but, crucially, trade integration weakens this link. These findings are obtained controlling for the potential endogeneity of trade integration and are robust to alternative specifications. 2 "You've had a broad substantial reduction in corporate tax rates outside the US. That occurs at a time when it's much easier (...) for companies to shift investment and income to take advantage of lower tax rates overseas. (...) To have a more competitive system, you want to try to bring down the rate closer to the range of our major trading partners." (Timothy Geithner, former US Treasury Secretary, Wall Street Journal, 28-30 January 2011, 9) IntroductionLooking at the evolution of corporate tax rates in developed countries over the last three decades, two observations stand out. The first is a decline in corporate tax rates (Figure 1, (Figure 1, right-hand side). ---Figure 1 about here ---Recent studies indicate the importance of such tax differences for firm location or FDI flows (Bénassy-Quéré et al., 2005;Kammas, 2011;Brülhart et al., 2012). Empirical analyses aimed at understanding what allows some countries to sustain higher tax rates than others have, however, not followed suit. 2 Yet, such insights can have important policy implications.1 This decline is also present when looking only at countries that were OECD members throughout the 1982-2011 period. 2 While numerous studies analyse the decline in corporate tax rates (Devereux et al., 2002(Devereux et al., , 2008Davies and Voget, 2008), Gilbert et al. (2005) is the only study analysing the determinants of corporate tax rate differentials. Compared to that study, we provide a firmer theoretical rationale for the analysis, address potential endogeneity problems, control for the influence of various country-level variables, and employ a Our results support theoretical predictions from Ottaviano and Van Ypersele (2005) and Haufler and Wooton (2010). Specifically, we show that the tax gap responds positively to population size (and GDP) differences between countries and, crucially, that this relationship is significantly attenuated by trade integration. Our analysis thereby controls for the endogeneity of standard measures of trade integration by extracting the level of trade liberalisation between each country-pair from a gravity equation (Rose ...

Leaders, Followers, and Asymmetric Local Tax Policy Diffusion

“…Bucovetsky (), Haufler and Wooton (), and Peralta and van Ypersele () all conclude that, in equilibrium, larger countries optimally tax at higher rates than smaller countries. Recent work by Exbrayat () and Exbrayat and Geys () also provide interesting theoretical and empirical findings along these lines. Using a three‐country setting with quasi‐linear utility functions and a lump‐sum tax on capital invested in a country, they find that asymmetry of population implies that larger markets are more profitable for firms than smaller markets.…”

Economic Integration, Corporate Tax Incidence and Fiscal Compensation

“…The former effect confirms that larger countries with a larger market can maintain higher taxes (Ludema and Wooton, 2000;Kind et al, 2000;Baldwin and Krugman, 2004;Exbrayat and Geys, 2014 Davies and Voget, 2013, and references therein).…”

“…We furthermore include a number of control variables specific to either equation. For the TAX-equation, this supplementary control vector (Yi,t) is inspired by research in the vast tax competition literature (Devereux et al, 2008;Overesch and Rincke, 2011;Exbrayat and Geys, 2014) and includes total population size (in million people), political leaning of the government (1 if left-wing, 0 otherwise), majority status of government (dummy = 1 if majority, 0 if minority) and public consumption (i.e., public sector size as a share of GDP in PPP). For the WAGE-equation, this supplementary control vector (Zi,t) includes a measure of worker's productivity (i.e., GDP per person employed in constant 1990 US$ at PPP) and the number of working days lost due to strikes and lock-outs (obtained from the International Labor Organization).…”

Economic Integration, Corporate Tax Incidence and Fiscal Compensation

“…Starting our discussion with a quick overview of the control variables in the TAX equation, we first of all observe that larger countries with a higher union density have higher statutory corporate tax rates. The former effect confirms that larger countries with a larger market can maintain higher taxes (Kind et al., ; Ludema and Wooton, ; Baldwin and Krugman, ; Exbrayat and Geys, ). To understand the latter effect, note that the union density variable captures the direct impact of unionisation on corporate taxes, its indirect effect being captured by the wage variable.…”

“…We furthermore include a number of control variables specific to either equation. For the TAX equation, this supplementary control vector ( Y i,t ) is inspired by research in the vast tax competition literature (Devereux et al., ; Overesch and Rincke, ; Exbrayat and Geys, ) and includes total population size (in million people), political leaning of the government (1 if left‐wing, 0 otherwise), majority status of government (dummy = 1 if majority, 0 if minority) and public consumption (i.e., public sector size as a share of GDP in PPP). For the WAGE equation, this supplementary control vector ( Z i,t ) includes a measure of worker's productivity (i.e., GDP per person employed in constant 1990 US$ at PPP) and the number of working days lost due to strikes and lockouts (obtained from the International Labor Organization)…”