Table 2

Difference-in-Differences-in-Discontinuities Results for Individual Outcomes

	70 y Mas	Earnings	Income	Employment	Hours Worked	Hours Worked (if >0)	Seguro Popular
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Panel A: Pooled
1(Age ≥ 70) * Eligible * Post	26.163***	−20.483	5.639	−0.218**	-6.351*	−0.900	0.173***
1(Age ≥ 70) * Eligible * Post	(3.129)	(18.687)	(42.024)	(0.086)	(3.343)	(4.765)	(0.065)
R-W p-value	0.010	0.525	0.970	0.050	0.099	0.970	0.050
R-squared	0.270	0.051	0.050	0.063	0.055	0.020	0.171
Mean dep. var.	4.90	54.41	255.05	0.39	17.54	41.29	0.22
Observations	15,860	15,860	15,867	15,860	16,017	6,804	12,457
Panel B: Male
1(Age ≥ 70) * Eligible * Post	25.901***	−31.936	22.865	−0.240**	−7.068	1.243	0.215***
1(Age ≥ 70) * Eligible * Post	(4.371)	(40.058)	(76.151)	(0.099)	(5.720)	(6.339)	(0.069)
R-W p-value	0.010	0.554	0.693	0.069	0.515	0.554	0.010
R-squared	0.214	0.072	0.094	0.121	0.103	0.031	0.169
Mean dep. var.	5.62	89.33	353.29	0.58	27.19	44.49	0.22
Observations	7,719	7,719	7,728	7,719	7,853	4,799	6,068
Panel C: Female
1(Age ≥ 70) * Eligible * Post	26.593***	−6.910	0.729	−0.247**	−7.598*	−5.293	0.156
1(Age ≥ 70) * Eligible * Post	(3.888)	(6.183)	(37.592)	(0.122)	(4.263)	(7.820)	(0.133)
R-W p-value	0.010	0.545	0.960	0.149	0.208	0.713	0.545
R-squared	0.389	0.045	0.029	0.038	0.030	0.023	0.177
Mean dep. var.	4.22	21.29	161.78	0.22	8.26	33.63	0.22
Observations	8,141	8,141	8,139	8,141	8,164	2,005	6,389
Weights	Y	Y	Y	Y	Y	Y	Y
Year-of-survey FE	Y	Y	Y	Y	Y	Y	Y

Notes: This table shows the results obtained from estimating a difference-in-difference-in-discontinuities regression (RD-DD) using individual-level outcomes as the dependent variable. The 1(Age ≥ 70) ∗ Eligible ∗ Post estimates show results after the 70 y Más program started between eligible and ineligible localities and before and after for those above the cutoff. 70 y Más is the self-reported amount of cash transfer received. Hours Worked (if >0) only includes non-zero hours worked. Income includes wages from job, wages from business, and secondary jobs, wages from previous months, all transfers (private and government, including 70 y Más), income from business and property, and other income not from work. Means are weighted using survey weights. Standard errors are clustered at the household level. The period of analysis before the program started is 2005 and 2006 surveys and after the program started is the 2008 and 2010 surveys. All regressions use survey weights and year-of-survey fixed effects. Observations across columns differ because we drop the upper 1 percent outliers for each outcome of interest. *p < 0.1, **p < 0.05, ***p < 0.01.