Do Imputed Educational Histories Provide Satisfactory Results in Fertility Analysis in the West

This paper investigates how well imputed educational histories perform in the analysis of first birth rates in the West German context. The focus here is on the quality of estimates when only rudimentary information on the timing of education is available. In many surveys, information on respondents’ educational histories is restricted to the highest level of educational attained by the time of interview and the date at which this highest degree was attained. Skeleton educational histories can be imputed simply from such rudimentary information. The German Life History Study has complete educational histories. We use these to compare estimates based on the complete histories with estimates based on corresponding imputed histories. We find that the imputed histories produce relatively reliable estimates of the effect on first-birth rates of having a university degree vs. having a vocational certificate. Estimating corresponding rates for women who have no such education proved to cause greater difficulties.