Table 5.

Polynomial distributed lag model

Polynomial	Quadratic		Quartic
Number of lags	5	10	5	10
Impact effect (same day)	$-$33.405***	$-$22.260***	$-$32.660***	$-$29.892***
	(8.461)	(6.058)	(10.384)	(9.056)
Cumulative effect	$-$79.886***	$-$61.903***	$-$138.392***	$-$97.132***
	(11.653)	(11.391)	(14.319)	(13.018)
Polynomial specification test
test-stat ($\chi (1)$)	0.68	2.10	0.88	1.58
P-value	0.409	0.147	0.349	0.208
Observations	35,190	35,190	35,190	35,190
Number of workers	635	635	635	635
Worker FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Month FE	Yes	Yes	Yes	Yes
Other weather	Yes	Yes	Yes	Yes

Robust standard errors clustered at the worker level are presented in parentheses, and ***$p<0.01$, **$p<0.05$, *$p<0.1$. P-values of the polynomial specification tests indicate none of the polynomial distributed lagged model specifications can be rejected. The sample comprises all day shifts from May to September