Measures for savings and saving rates in the Ger-man SAVE data set

Saving is frequently measured using a one-shot question for total annual saving dur-ing the preceding year. This type of one-shot recall question might cause severe measurement errors since saving is a complicated concept which consists of various components, many of which respondents might not be fully aware of. This paper uses the SAVE data to analyze poten-tial errors generated by this kind of questioning and provides remedies in order to construct the most of reliable saving measure given the information at hand.