Modeller för internprissättning: Re-analys av experimentella tester

Abstract

Dejong et.al. (1989) and Avila & Ronen (1999 conduct experimental tests of various transfer pricing mechanisms. The resulting quantity data is transformed to efficiency measures which the authors analyse primarily through t-tests and analyses of variance. The problem in both experiments is that the data cannot be seen as samples from approximate normal distributions. This means that non-parametric methods should be used to analyse the data in order to get valid conclusions. In this paper, the data sets from both experiments are re-analysed with nonparametric methods. Of course, there would be no story if the conclusions from these analyses did not differ with respect to the original sources. But they do. The overall picture is that there are significant differences between the tested mechanisms in more ways than the original authors found.