Table S3.

Falsification test, claims

Experimental arm	Diff in Diff: pre-post	Falsification: 2008–Pre	Falsification: 2007/8–Pre
Main effects
Placebo	−0.014	0.018	0.017*
	(0.010)	(0.011)	(0.009)
Any treatment sticker	−0.017***	0.010	0.012*
	(0.008)	(0.009)	(0.007)
Individual vs. collective action
Individual	−0.012	0.010	0.015*
	(0.008)	(0.010)	(0.008)
Collective action	−0.022***	0.010	0.010
	(0.008)	(0.010)	(0.008)
Message types
Text message only	0.001	−0.003	0.002
	(0.010)	(0.011)	(0.009)
Supportive: Voice	−0.024**	0.017	0.019**
	(0.010)	(0.012)	(0.010)
Event aversion: Crash	−0.028***	0.019	0.023**
	(0.010)	(0.012)	(0.010)
Consequence aversion: Injury	−0.016*	0.008	0.006
	(0.010)	(0.012)	(0.010)
Observations	23,474	23,474	23,474

The pre period is 2009/10, and the post period is 2011/13, as in Table 4, from which the first column is copied. The second column is a falsification test in which a double-difference estimation is conducted using data from 2008 and the pre period. The third column combines data from 2007 and 2008 and compares this with the pre period. All reported coefficients are on the interaction between the independent variable and the latter of the two periods considered. SEs in parentheses. Significant differences at *10%, **5%, and ***1% levels.