Gender discrimination has been an important socio-economic and cultural issue in Indian backdrop; though India is one of the fastest-growing nations in the world but she has ranked 108th (out of 144 countries) in 2016 on the global gender gap index. In order to explore the determinants of crimes (such as dowry death, rape, molestation and torture) in India, we consider 24 states and five time points (over 1995–2015). Rank correlation over time clearly suggests that states exhibiting higher incidence of molestation experience higher incidence of rape; we get similar results in case of torture and dowry death as well. We employ both global and local Moran’s indices for dowry death, which has the least reporting bias, and observe that space as well as neighbourhood effect does matter; on the contrary, we find that other types of crimes vary randomly. Specifically, spatial panel regression results on dowry death clearly show that space along with neighbourhood effect does matter towards variations of dowry death; only two variables namely female work force participation and police station significantly reduce the incidence of dowry death.

在印度背景下，性别歧视一直是一个重要的社会经济和文化问题；尽管印度是世界上增长最快的国家之一，但她在 2016 年的全球性别差距指数中排名第 108 位（在 144 个国家/地区中）。为了探索印度犯罪（如嫁妆、强奸、猥亵和酷刑）的决定因素，我们考虑了 24 个州和 5 个时间点（1995-2015 年）。随时间推移的等级相关性清楚地表明，性骚扰发生率较高的州强奸发生率较高；在酷刑和嫁妆死亡的情况下，我们也会得到类似的结果。我们对嫁妆死亡采用全球和本地 Moran 指数，其报告偏差最小，并观察到空间和邻里效应确实很重要；相反，我们发现其他类型的犯罪随机变化。具体来说，关于嫁妆死亡的空间面板回归结果清楚地表明，空间和邻里效应对嫁妆死亡的变化确实很重要；只有两个变量，即女性劳动力参与和警察局，显着降低了嫁妆死亡的发生率。