Table 5.

Effect of Local Weather Shocks on Dowry Deaths by Timing of Shocks

	(1)	(2)	(3)	(4)
Dry Shock (JJAS)	0.122*** [0.0386]	0.140*** [0.0378]	0.160*** [0.0388]	0.154*** [0.0390]
Wet Shock (JJAS)	0.0184 [0.0302]	0.0192 [0.0314]	0.0190 [0.0328]	0.0186 [0.0330]
Dry Shock (OND)		0.111 [0.0819]	0.0975 [0.0852]	0.0831 [0.0800]
Wet Shock (OND)		−0.0152 [0.0453]	−0.0165 [0.0489]	−0.0175 [0.0484]
Dry Shock (AM)		0.0394 [0.0742]	0.0456 [0.0785]	0.0319 [0.0762]
Wet Shock (AM)		0.0406 [0.0500]	0.0434 [0.0486]	0.0462 [0.0481]
Dry Shock (JFM)		0.144 [0.183]	0.165 [0.177]	0.138 [0.187]
Wet Shock (JFM)		−0.103 [0.169]	−0.0877 [0.170]	−0.0514 [0.152]
Lagged Dry Shock (JJAS)			0.0895** [0.0452]	0.0621 [0.0406]
Lagged Wet Shock (JJAS)			0.0134 [0.0491]	0.00260 [0.0497]
Lagged Dry Shock (OND)			0.0603 [0.0658]	0.0341 [0.0624]
Lagged Wet Shock (OND)			−0.0215 [0.0314]	−0.0145 [0.0326]
outlier				1.449*** [0.0420]
Observations	3,138	3,138	3,138	3,138
Districts	523	523	523	523

Notes: Each column presents estimated coefficients from a separate conditional fixed effects Poisson regression for the number of dowry deaths in a district-year. “Dry Shock” is the absolute deviation of rainfall below the long run (1971–2000) mean in the specific months within the defined periods. “Wet Shock” is the analogous variable above the mean. JJAS shocks are dry and wet shocks June–September, OND are October–December, JFM are January–March, and AM are April and May. Standard errors are reported in brackets and are clustered by district. All regressions include district and year fixed effects and 4 socioeconomic controls: percent literate, percent employed, percent scheduled caste, and total population, all interacted with year indicators.

^***indicates significance at 1,

^**at 5, and

^*at 10 percent level.