In recent years, small and/or rural schools have attracted renewed interest from educational researchers (Roberts & Fuqua, 2021). One of the key reasons is the need to understand the growing polarization between the values and opportunities available in metropolitan areas and other parts of the country. In this study, we are looking for an answer to the question of when small, often rural schools are also an attractive choice considered by urban parents. More specifically, we analyze available data on all cases where students transfer to small and/or rural schools for various reasons during primary school attendance in three years (2016-2019).

Between school mobility might be result of ordinary residential mobility in the population. It may also be a sign of repeated school choices correcting a prior choice of school, in some cases indicating system-wide failure, e.g. schools underserving particular groups of students (Fox & Buchanan, 2017; Koinzer et al., 2017). School choice and between school mobility are two related phenomena that are often studied for urban areas only. Systematic review has also shown that most research on student mobility comes from North America (e.g. Welsh, 2017; Calibuso & Winsler, 2021). In this study, we analyse the complete data on horizontal mobility for all small rural schools in Czechia as a post-socialist Central European country.

Context of the study. Around 95 % of Czech students in compulsory education age (typically 6 to 15) attend the municipal comprehensive school called základní škola, literally “basic school”, somewhat resembling Nordic comprehensive schooling (Åberg-Bengtsson, 2009; Kvalsund, 2000). Out of the total number of municipal schools is approx. 2100 rural schools and 1500 urban schools. Czech basic comprehensives, however, exist in two forms with different grade span: either relatively large “full” schools that combine both primary and lower secondary grades (attended by children aged 6–15), or smaller “incomplete”’ schools (some of them with multi-grade classes). While 93% of urban schools are combined (grade span 1-9), about 57% rural schools have primary classes (5 grades or less) only. As expected, rural school are generally smaller than urban ones: average urban school has 435 students, while average primary rural school has only 42 students and “full” rural school 193 students. Thus, approx. 75% students attend an urban school.

The migration of urban students to rural schools can be an expression of a new appreciation of the values that are important in small communities and the community school concept, or just a search for a dream of a safe and understandable micro-world for their children (Pospěch, 2021). Rural schools are, on the other hand, surprisingly resourceful in offering alternative and innovative educational services from "traditional" alternative models (as Montessori or Waldorf) to hybrid and online teaching or de facto homeschooling. Despite the recent emergence of some private schools in rural sites, most events of school choice in Czechia happen within public education systems (cf. Asadolahi et al., 2021).

Our scope of analysis is limited by the available administrative data of national registers. Thus, we address the following research questions: How does the student mobility patterns differ for different types of original and target schools (urban/rural, public/private)? What is the ratio of residential and non-residential (strategic/reactive) transfers? Can some typical reasons for school change be inferred from the available data?

This study is based on administrative microdata from the Czech educational registers merged with databases containing geographical information. The administrative data on all transfers of Czech comprehensive school students between September 2016 to September 2019 were obtained from the data administrator. They had been fully anonymised to comply with the ethical and legal requirements of personal data protection. Before analyzing the entire national sample, we studied the data for the selected region as a pilot case to better understand the data. (The results of the pilot study were presented at the ECER 2021 conference.) In full-scale study, we used 97315 records about the transfers of 79918 unique students. Due to the settlement structure, most of the events concerned transfers between city schools. In the next step, we selected students who once transferred to a small, i.e. primary only school (7416 cases). Out of this subset, we were able to identify both original and destination school data for 5413 events. The cases we have not succeed to link were of two types: enrollment in a new school in the first year under review, for which we did not received data on the school the student attended at the end of the previous school year. The second group are students with a different nationality transferring from a school abroad, where the number of unassigned transfers is approximately constant in each year. We used exploratory data analysis, mainly data tabulation and visualisation in space to spot patterns and outliers, and examine possible clusters and relationships.

In our set (n = 5413) of transfers, there was roughly the same share of residential mobility (the student changed both school and the municipality of residence) and non-residential mobility, where we can assume as a reason primarily a new choice of school. Somewhat surprisingly, the distribution of residential and non-residential mobility by grade does not differ. In the Grade 1, it can be assumed that the parents coordinate the start of school with a possible move. Within the non-residential mobility, however, we see a difference in distinguishing the target school. First, transfers to schools for students with special needs mainly happen in the Grade 1; we can interpret them as failed cases of inclusion. In absolute volumes, however, these are not big numbers. Second, transfers to private schools are most pronounced in the Grade 2, probably a response to the experience of public schools. Transfers to church schools follow similar pattern, but again these are small numbers burdened by a random error. Students from private schools transfer to private schools again. About three times more students move from public to private school, but the number of students moving in the opposite direction is also significant given the number of both types of schools. Non-residential transfer from a municipal school is almost three times more likely to go to a private school than a residential transfer. Non-residential transfers are clearly reactive or strategic moves looking for a "better" school. At the same time, 47% of our sample consisted of cases where a student transferred from a school in the city to a school in a rural settlement (whether with or without a change of residence). As we assumed, there is mobility of Students from the city to small rural schools. This subset of events will undergo further detailed analysis.

Åberg-Bengtsson, L. (2009). The smaller the better? A review of research on small rural schools in Sweden. International Journal of Educational Research, 48(2),100-108. https://doi.org/10.1016/j.ijer.2009.02.007. Asadolahi, S., Farney, J., Triadafilopoulos, T., & White, L. (2021). Charting the Rise of School Choice across Canadian Provinces: A Policy Index. Canadian Journal of Political Science, 1-20. doi:10.1017/S0008423921000901 Calibuso, E., & Winsler, A. (2021). Who switches schools? Child-level predictors of school mobility in middle school students. Journal of Child and Family Studies, 30, 263–275. Fox, R. A., & Buchanan, N. K. (Eds.). (2017). The Wiley handbook of school choice. Wiley Blackwell. 10.1002/9781119082361 Koinzer, T., Nikolai, R., & Waldow, F. (2017). Private schools and school choice in compulsory education: Global change and national challenge. Springer Fachmedien. Kvalsund, R. (2000). The transition from primary to secondary level in smaller and larger rural schools in Norway: comparing differences in context and social meaning. International Journal of Educational Research, 33(4), 401-423. Pospěch, P. (2021). Neznámá společnost. Pohledy na současné Česko. Host. Roberts, Ph., & Fuqua, M. (2021). Ruraling education research: Connections between rurality and the disciplines of educational research. Springer. Welsh, R. O. (2017). School hopscotch: A comprehensive review of K–12 student mobility in the United States. Review of Educational Research, 87(3), 475-511.