Foreign connections and the difference they make: how migrant ties influence political interest and attitudes in Mexico

Comparative Migration Studies

Table 2 Doubly Robust Estimations of Migration Treatments on Political Interest, Talk and Efficacy of Mexican Elections, Panel Waves 1, 2 & 3

	Wave 1				Wave 2				Wave 3
	Unweighted Average Outcome		Weighted Average Outcome for Treated if in Control		Unweighted Average Outcome		Weighted Average Outcome for Treated if in Control		Unweighted Average Outcome		Weighted Average Outcome for Treated if in Control
	Treatment Group	Control Group	Weighted Average Outcome for Treated if in Control	DRϕ	Treatment Group	Control Group	Weighted Average Outcome for Treated if in Control	DRϕ	Treatment Group	Control Group	Weighted Average Outcome for Treated if in Control	DRϕ
Interest in Politics
Treatment 1	72.65	69.22	78.46	–5.81	82.95	77.53	83.02	–0.077	84.74	79.64	84.79	–0.058
Treatment 2	66.67	71.03	74.07	–7.41^.	84.7	79.76	78.14	6.52^.	83.45	81.79	75.66	7.79*
Talk about Politics
Treatment 1	43.01	36.67	38.61	4.41**	55.36	46.13	56.18	–0.82	49.82	41.36	44.86	4.96*
Treatment 2	32.02	40.86	37.07	–5.05	48.88	51.25	46.52	2.36	36.55	47.01	42.92	–6.36
Treated Sample Size	847	Effective Sample Size			208
Elections are Free and Fair
Treatment 1	−	−	−	−	−	−	−	−	76.54	71.14	79.35	–2.81
Treatment 2	−	−	−	−	−	−	−	−	75.91	74	83.99	–8.08*
Treated Sample Size	115	Effective Sample Size			497

Source: Lawson et al. (2007). Authors' calculations using doubly robust estimation with propensity score weighting in R. ϕ Doubly robust estimator is the unweighted treatment group average minus the weighted control group average. Note: Effective sample size is after weighting. Signif. codes: p < 0.10 *p < 0.05 **p < 0.01 ***p < 0.001