Brief Methodological Report on Tax Estimation: Influence of Wage Spread and Household Composition on the Forecast Error in the Wage Tax

Tasks

The derivation of the wage tax in the context of the tax projection is based on the forecasts for employment and per capita wages according to the System of National Accounts (SNA). Although the results of this wage tax projection are quite accurate, there have been recurrent deviations in recent years and a tendency to underestimate the actual amount of revenue. Causes for these differentials can be manifold. The main focus of the analysis, which is based on micro data from the Socio-Economic Panel (SOEP) and a special evaluation of wage tax statistics, is on the effects of changes in the distribution of wages and the relative occupation of wage tax classes or changes in societal and demographic structures. Because of the progressive tax rate, changes in the relative distribution of wages can affect tax revenues beyond the development of macroeconomic aggregates. The change in the composition of the payroll tax classes mainly concerns the extent of the use of joint taxation of married couples.

Methods

Descriptive evaluation and comparison with official statistics, microsimulation of tax revenue, comparison of the deviation with the one of the Working Group on Tax Revenue Forecasting.
Data and other sources:

The Socio-Economic Panel (SOEP), a special analysis on wage and income tax statistics, tax and labor market data from the  Federal Statistical Office of Germany and estimations from the Working Group on Tax Revenue Forecasting serve as the data basis.

Results

The hypothesis that an increase in wage inequality has led to a systematic underestimation of wage tax revenue can be discarded. On the basis of the SOEP, however, three observations can be made that support the thesis that changes in the distribution of wage tax classes over time lead to an increase in tax revenue. First, the share of single households has risen steadily since the beginning of the 1980s. Second, women's labor market participation has been steadily increasing over this period. Third, the labor income of second earners has risen steadily relative to the labor income of first earners over the past ten years. These developments diminish the role of joint taxation of married couples.

The following conclusions can be drawn for the derivation of wage tax revenues. The bias towards positive differences between realization and estimation of the wage tax is predominantly due to the failure to follow social and demographic trends in the derivation, which results in the relative distribution of tax classes developing in favor of tax class I and at the expense of tax class III. Although the distribution of wages also has an impact, over time this has probably worked in different directions. For the derivation of the wage tax, the overall positive difference can be justified in the continuation of the current social and demographic trends