Profit shifting in the Norwegian and British petroleum industry: Differentiating between the real and shifting response to tax changes
Helene Vada, OFS Scholarship Recipients 2016.
In this master’s thesis, I explain the concept of profit shifting within multinational enterprises and investigate whether petroleum companies on the Norwegian and British continental shelves engage in tax motivated profit shifting, by applying ordinary least squares and company fixed effects estimation. To be able to distinguish between traditional tax distortions and profit shifting, I extend the model developed by James R. Hines and Eric M. Rice in their 1994 article “Fiscal Paradise: Foreign Tax Havens and American Business” published in The Quarterly Journal of Economics, by including the statutory tax rate that applies to the petroleum companies and the parent companies’ corporate income tax rates as separate explanatory variables in the regressions, while also performing regressions using the tax difference as explanatory variables. The majority of previous studies on profit shifting use the average tax difference between affiliates as the explanatory variable. The tax difference variable used in this thesis is a simplified version including only the difference between the tax rates that apply to the petroleum affiliate and the parent company. As companies also have the option of shifting profits to other affiliates within the group, the simplified tax difference variable is expected to capture only parts of any profit shifting. I propose that the statutory tax rate that applies to the petroleum companies also represent the tax difference between the petroleum affiliate and the affiliates located in a country with a corporate income tax rate of zero. As this is the most profitable channel for profit shifting, the coefficients that result from a regression of reported profits on the tax rate that applies to the petroleum companies can also be interpreted as the semi-elasticity of reported profits with respect to the maximum tax difference between the petroleum company in question and its affiliates. Previous studies have shown that the semi-elasticities of reported profit with respect to the tax difference between the affiliate at hand and its group members are in the range [-3,-0.5]. The corresponding semi-elasticities that result in this thesis are in the range [-3, -1.4]. The semi-elasticities of reported profit with respect to the petroleum tax rate are in the range [-5.5, -0.1], while semi-elasticities of reported profits with respect to the parent company tax rate are in the range [-12.4, 3.3]. I conclude that there is some evidence of profit shifting in the data, and that some of the results are comparable to the ones from previous studies.