After the systematic disadvantage of educational opportunities for children from rural areas in comparison to children from urban areas was a central research topic in the 1960s and 1970s, the role of regional disparities for educational opportunities at the transition to secondary level of the education system in Germany is now only irregularly discussed. There is still a relative lack of educational infrastructure in rural areas and regional labor market structures have hardly been considered as possible further determinants of regional educational disparities. Moreover, in the few current studies, important individual factors such as competencies, grades, school recommendations and educational aspirations could not be taken into account or they refer to individual regions. This paper picks up the question and tries to overcome these restrictions via combining data from the National Educational Panel Study Starting Cohort 3 (NEPS SC3) and regional information at the individual level. Once enriched with information on regional infrastructure, this data set not only contains necessary information on educational decisions, but also on important control variables at the individual level as mentioned above. Using multiple imputation to deal with missing values, hierarchical logistic regressions allow for investigating the impact of regional infrastructure on the probability of attending a high school in the school year 2010/2011 in Germany. The results show that school infrastructures are still significant factors in school choice even when controlling for relevant individual factors. The analyses show that a higher proportion of high schools to all secondary schools in the county a student lives in increases the probability that a high school is chosen at the transition from elementary to secondary school. For regional labor market structures no eff ects are revealed in this study.

Keywords
School choice; Regional infrastructure; Hierarchical binary logistic regression, Missing values