Merz: The Distribution of Income of Self-employed, Entrepreneurs and Professions
12
Summarizing: the German Income Tax Statistics tables deliver tax and income data for
the above 14 selected professions, which together with the large group of “other
professions” make up all professions, according to persons liable to pay tax with income
derived exclusively from professional work.
With the intention of depicting the predominant life situation (with the additional
possibility of historical comparisons to earlier data) we will concentrate to those persons
liable to pay tax where the predominant source of income (and not just any professional
income) defines the status of self-employed work as professions.
With regard to the self-employed definition above: when we refer to professions, then
they are exactly defined and built far more than 90% of the ‘independent’ workers
within the entire self-employed group consisting of sources of agriculture and forestry,
business practise and ‘independent’ work.
5.3.3 Dependently Employed, Employees
The group of dependently employed (employees), consisting of (blue-collar) workers,
salaried employees (white-collar) and civil servants can be differentiated in the income
tax statistic directly from those persons liable to pay tax with income derived from non
self-employed work (nichtselbstandiger Arbeit). Again, we will concentrate here on the
predominant source concept, too, when employees are regarded.
5.4 Finally: Our Microdata Base of Individual Income and Tax
Information
Our specific microdata base of individual income and tax information of the actual
German Income and Tax Statistic 1992 allows for the first time a microanalysis of the
distribution of income with special emphasis on the income situation of the self-
employed and the professions.
The German Income Tax Statistic is a population wide statistic with wages and income
tax (Lohn- und Einkommensteuer) information of overall ca. 30 Mio. cases, 450 items.
Our final microdatabase is a stratified random sample of 100.000 tax payers (finally
with 80.007 as 'working' people) out of a 10% tax microsample drawn by the Federal
Statistical Office. A description of the 100.000 sample is given in Zwick 1998.
Because of the data anonymization and data protection rules, special arrangements were
necessary for the income and tax calculations inside the Federal Statistical Office.
Based on our inequality program package including the new decompostion of
generalized entropy measures written in SPSS (based on program parts of the Sfb3-
group at the University of Frankfurt under Prof. Dr. R. Hauser)4 within the Federal
Statistical Offcie all microdata calculations were done with the described microdata
base.which was delivered to the Office for the further calculations.5
4 Many thanks to Dr. Irene Becker for her helpul comments and support.
5 We are very grateful to Markus Zwick from the Federal Statistical Office for his excellent and engaged cooperation
and his efforts in preparing, selecting and (re-)running all the jobs.