Political biases despite external expert participation? An empirical analysis of tax revenue forecasts in Germany.

This paper explores whether and how political biases arise in an institutional setting where revenue forecasting is delegated to a body that includes experts from institutions neither part of nor controlled by the government. The empirical analysis focuses on the performance of German federal tax revenue forecasts, in the preparation of which the advice of external experts has a long tradition as an institutional safeguard. While, on average, revenue forecasts turn out to be unbiased, the results show that the government exerts an influence. In particular, optimism/pessimism in the government's GDP forecast helps to explain why the revenue forecast turns out too optimistic/pessimistic. In addition, governmental estimates of the revenue effects of tax-law changes are found to contribute to forecast errors. [ABSTRACT FROM AUTHOR]