The impact of education reform on women's crime exposure in Türkiye

Abstract

This study examines the effects of an educational reform on middle school completion, exposure to crime, and perceptions of security and trust in security forces. Using data from the Turkish Statistical Institute's Life Satisfaction Survey (2018-2022), the analysis focuses on female individuals born within a 72-month bandwidth around the January 1987 cutoff. This study applies a regression discontinuity design and Two Stage Least Squares (2SLS) to assess the impact of the reform, while also considering the inclusion and exclusion of the COVID-19 years to account for any pandemic related irregularities. The decision to exclude the COVID-19 years (2020 and 2021) from the analysis was made due to the potential disruption in data patterns during the pandemic, which may have affected the stability of our outcomes. Such disruptions could include changes in response behavior or alterations in the broader socio-economic context, making it difficult to isolate the true effects of the reform during those years. The exclusion ensures that the results more accurately reflect the reform's impact rather than any pandemic-related anomalies. The findings reveal that the education reform led to significant increases in middle school completion rates. We also find suggestive evidence of reductions in exposure to crime, although this result is not statistically significant. Additionally, the reform improved individuals' perceptions of personal security, particularly in terms of feeling safe at home while alone. Trust in security forces also increased, with women showing improved satisfaction with police forces and a higher perception of timely responses to security issues. However, some variables, such as the feeling of safety when walking alone at night, did not change significantly. Considering the reduction in the number of observations due to the exclusion of the COVID-19 years, we observed minor changes in the magnitude of the results, while the standard errors increased in the regressions when 2020 and 2021 were excluded. The decrease in the number of data points, resulting from the exclusion of the COVID 19 years, likely contributed to these changes. While the sizes of the coefficients remained largely unchanged, the statistical significance of some variables weakened with the reduced sample size. Despite this, the overall findings remained consistent. These results enhance our understanding of how educational reforms can influence not only academic outcomes but also broader societal dimensions, such as public confidence in institutions.