We outline a procedure for combining six cross-sections of the German Sample Survey of Income and Expenditure, and discuss potential pitfalls of such a venture. Particularly, we investigate the consequences of a major break in the survey design for inter-temporal comparisons of expenditure...