Does education affect religiosity? Causal evidence from a conservative emerging economy
Özer, Mustafa; Fidrmuc, Jan; Mentzakis, Emmanouil; Özkan, Özcan (22.02.2024)
Numero
2/2024Julkaisija
Bank of Finland
2024
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe202402228380Tiivistelmä
Does education make people more or less religious? The previous literature offers mixed findings on the relationship between education and religiosity. This may be due to endogeneity bias: education and religiosity can be caused by a third variable such as culture or upbringing. We instrument education by exposure to the 1997 education reform in Türkiye which increased mandatory schooling from 5 to 8 years. The schooling reform increased the probability that young girls would complete 8 years of schooling and report lower religiosity later in life. The reform apparently did not influence such outcomes for boys. These effects are observed primarily for females growing up in strongly religious or poor areas.
Julkaisuhuomautus
NON-TECHNICAL SUMMARY
FOCUS
Our paper seeks to shed new light on the causal impact of education on religiosity. A possible reason earlier literature offers mixed results in this respect is endogeneity bias. Education and religiosity can be affected by other variables such as culture and upbringing. We address this problem by utilizing the 1997 reform of compulsory education in Türkiye. The reform increased mandatory schooling from 5 to 8 years.
CONTRIBUTION
We use the exposure to the 1997 education reform to identify as good as random variation in educational outcomes between birth cohorts affected by the reform and those slightly older who were unaffected by it. Specifically, individuals born between 1981 and 1985 are used as the control group, as they were obliged to remain in school for only five years. Those born between 1987 and 1991 were affected by the reform and thus were required to complete eight years of schooling. Given that there was little change in the nature or quality of education changed after the reform was implemented, the sole difference between these two groups is the amount of schooling that they were required to complete.
FINDINGS
The reform increased the probability that young girls complete 8 years of schooling, but had no similar impact on such outcome for boys. Girls subject to the reform, in turn, displayed lower self reported religiosity when surveyed in 2013. These effects are primarily observed for females who grew up in poor or strongly religious areas.
FOCUS
Our paper seeks to shed new light on the causal impact of education on religiosity. A possible reason earlier literature offers mixed results in this respect is endogeneity bias. Education and religiosity can be affected by other variables such as culture and upbringing. We address this problem by utilizing the 1997 reform of compulsory education in Türkiye. The reform increased mandatory schooling from 5 to 8 years.
CONTRIBUTION
We use the exposure to the 1997 education reform to identify as good as random variation in educational outcomes between birth cohorts affected by the reform and those slightly older who were unaffected by it. Specifically, individuals born between 1981 and 1985 are used as the control group, as they were obliged to remain in school for only five years. Those born between 1987 and 1991 were affected by the reform and thus were required to complete eight years of schooling. Given that there was little change in the nature or quality of education changed after the reform was implemented, the sole difference between these two groups is the amount of schooling that they were required to complete.
FINDINGS
The reform increased the probability that young girls complete 8 years of schooling, but had no similar impact on such outcome for boys. Girls subject to the reform, in turn, displayed lower self reported religiosity when surveyed in 2013. These effects are primarily observed for females who grew up in poor or strongly religious areas.