Table 4.

Extreme temperature, mortality & interaction with SES

	(1) Interaction with SES
≤ 5th percentile	0.001
	(0.002)
≤ 5th percentile # Low SES	0.007***
	(0.001)
≤ 5th percentile # Medium SES	0.007***
	(0.002)
> 5th and ≤ 10th percentile	0.010
	(0.005)
> 5th and ≤ 10th percentile # Low SES	−0.008***
	(0.002)
> 5th and ≤ 10th percentile # Medium SES	−0.013**
	(0.004)
> 10th and ≤ 25th percentile	0.001
	(0.001)
> 10th and ≤ 25th percentile # Low SES	0.002*
	(0.001)
> 10th and ≤ 25th percentile # Medium SES	0.001
	(0.001)
≥ 75th and < 90th percentile	0.004
	(0.003)
≥ 75th and < 90th percentile # Low SES	−0.005***
	(0.001)
≥ 75th and < 90th percentile # Medium SES	−0.006***
	(0.001)
≥ 90th and < 95th percentile	−0.000
	(0.001)
≥ 90th and < 95th percentile # Low SES	0.000
	(0.002)
≥ 90th and < 95th percentile # Medium SES	0.003
	(0.002)
≥ 95th percentile	0.002
	(0.002)
≥ 95th percentile # Low SES	0.006***
	(0.001)
≥ 95th percentile # Medium SES	0.006***
	(0.001)
Observations	25,200
Controls	YES
Province-Month FE	YES
Month-Year FE	YES

Results are obtained estimating Eq. (1) adding an interaction between the temperature bins and SES groups. The High SES group is at the baseline level. Standard errors are clustered at the provincial level. Constant present but not reported. Control variables are precipitation, solar radiation, wind speed, relative humidity, GDP per capita, PM2.5, gender and population density. Significance levels: ***p < 0.001, **p < 0.01, *p < 0.05