Are Reporting Errors a Problem for Factual Information and, If so, for Whom? an Analysis of the Deviance between Self-Reported and Administrative Data and Its Potential Determinants

Self-reported survey data are used for a wide variety of research in a broad range of fields, often with far-reaching conclusions and policy implications. Particularly self-reports on factual information, like citizenship, work history, or income, are commonly taken at face value despite plausible concerns regarding their validity, e.g., due to social desirability, careless responding, or satisficing. Typically, exploring this issue is also complicated by the fact that reference data for validating self-reports are usually not available to researchers either due data protection concerns or simply the non-existence of suitable data to compare self-reports to on a population level.

The Austrian Socio-Economic Panel (ASEP) offers an exceptional opportunity for research into this matter, as it allows for a linkage of self-reported survey data of its household panel of the Austrian general population with administrative data of the Austrian federal statistics office (Statistics Austria) on the household and individual level, which were cleaned, processed, and edited for research purposes.

To contribute to the field of response quality and measurement error in factual self-reports, we will quantify deviances between self-reports and administrative data regarding the same information on a variety of topics ranging from demography, work, income, and education to housing. Further, we will look into potential determinants of such deviances. Lastly, we will critically discuss our results regarding potential biases that might exist in the administrative records as well.