Akademik Veri Yönetim Sistemi
The 32nd Conference on Economics of Turkish Economic Association , Gazimagusa, Kıbrıs (Kktc), 13 - 15 Kasım 2025, ss.99, (Özet Bildiri)
This study analyzes the multifaceted relationship between women's employment status and their experience of domestic violence in Turkey using Bayesian Network (BN) modeling. Based on micro-data from the Turkish Statistical Institute's 2014 Survey on Domestic Violence Against Women , the analysis was conducted on a final dataset of 5,843 complete cases. Following a comparison of multiple structure learning algorithms and model validation techniques, the most statistically and theoretically appropriate network structure was identified. The study's most crucial finding is the absence of a direct and statistically significant relationship between being employed and the risk of physical or sexual violence. The model identifies "education level" as a fundamental root cause that influences employment and other socio-demographic variables. Policy simulations revealed that a scenario of full female employment would not alter violence rates (from 31.6% to 31.8%) but would dramatically increase women's control over their own income from 27.4% to 76.9%. These results suggest that employment policies, rather than directly reducing violence, may provide longterm, indirect protection by enhancing women's economic autonomy. To combat violence, holistic approaches that target education and economic empowerment together are needed, rather than policies focused solely on employment.