. Author manuscript; available in PMC: 2013 Dec 1.

Published in final edited form as: Int J Health Care Finance Econ. 2012 Sep 15;12(4):253–267. doi: 10.1007/s10754-012-9113-2

Table 4.

Probability remain employed,^a first differences and difference-in-difference^b from linear probability models, initially-employed married men under age 65 with employer-provided health insurance

	Cancer	Lung disease	Hospitalized	New diagnosis	New diagnosis only
ECHI×health shock	(1) −0.005	(2) 0.048	(3) −0.084^*	(4) −0.015	(5) 0.198^***
Non-ECHI×health shock	(0.085) 0.103	(0.098) −0.239	(0.049) −0.050	(0.064) −0.034	(0.059) −0.098
Diff-in-diff	(0.081) −0.108	(0.216) 0.287	(0.084) −0.034	(0.098) 0.019	(0.144) 0.296^**
N	(0.093) 1,212	(0.223) 1,175	(0.076) 1,385	(0.093) 1,265	(0.145) 1,192
Treated	72	35	245	125	52

Notes: ECHI=employment contingent health insurance; non-ECHI=health insurance through spouse’s employer.

Significance: p<.1

^**

p<.05

^***

p<.01.

“Employed” is defined as working for pay.

First difference is (health shock employment – healthy employment) for the specified group (ECHI or non-ECHI). Diff-in-diff is the difference between these effects.

Controls for first interview age under 40 or 40-59, nonwhite, some college, college degree or more, income under $20k, income more than $75k, physical job, stressful job, physical job×health shock, stressful job×health shock, firm size (25-100 or >100 employees), part-time work (fewer than 35 hours), spouse not working, spouse part time, spouse retired, spouse had bad health (fair/poor versus excellent/very good/good), spouse over 65, spouse covered by respondent’s employer-based insurance, dependents covered by respondent’s employer-based insurance, and year dummies.