A theoretical and empirical evaluation of the corporate tax base elasticity in OECD countries from a Norwegian viewpoint

Abstract

For the past 30 years, the corporate income tax rates in the OECD countries have decreased from 45 percent in 1982 to below 27 percent in 2012. At the same time, the corporate income tax revenue as share of GDP has increased. Tax competition is seen as the main source of this development. Firms respond behaviourally to the tax rate and the tax systems deductions and allowances in order to maximise their after-tax returns. The larger the behavioural effect is, the greater the revenue costs and dead weight loss for society becomes. Firms decide in which country to locate based on comparing their expected after-tax profits between countries. Further, a multinational firm can shift their profits to be taxed in countries with low taxes. A firm, or an entrepreneur, will also choose legal form based on tax incentives. Investments and the financial structure of the firm are also influenced, determining how much to invest and the composition of debt and equity. I use panel data of 19 OECD countries in the period of 1982 to 2012 to estimate the long run effects of the domestic and neighbouring tax rates on the corporate tax base. The regression controls for various macro data variables such as income per capita, GDP growth, cost of labour and EU-membership as well as country specific, time invariant and country trend effects. The results of this analysis is that the tax base is elastic with regards to changes in the domestic tax rate, specifically, the elasticity of -1.17 in the tax base with a one percent change in domestic tax rate. The effects of neighbouring countries changes in tax rates are inelastic and of low significance. A comparison of my results to the closely related literature of similar methodology shows a higher elasticity of the tax base and less effect of the neighbouring tax rates then previous findings. Extensions show that replacing the data for Norway with oil-revenue corrected data increases the estimated elasticity of the domestic tax rate and lowers the effect of the foreign tax rate. The dynamic model estimations show that in the short run the tax base is inelastic, where half the effect of a change in the domestic tax rate occurs in the immediate year. The resulting long run semi-elasticity of -2.9 % is close to the aggregate effect estimated in a meta-analysis by De Mooij and Ederveen (2008) of -3.1 % of a one percentage point change in the tax rate.