Table 2.

Estimated effects on labor-force status, EI⁺ samples of Union veterans and IPUMS⁺⁺ sample of men ages 35–60

Status and covariate	EI veterans	IPUMS males
Unemployed 1880
Never married	.048*	.094*
African American	.010*	− .042*
“Crippled”	.083*	.040
Older than 40	.003	.008
Unemployed 1870, employed 1880
Never married	.013*	.033*
African American	.003	.003
“Crippled”	.002	.011
Older than 40	− .014*	− .015*
N	17,794	5201

Effects estimated by weighted multinomial logistic regression. Cell entries are estimated changes in probability of each outcome (as opposed to employment at both censuses) if all sample members had each characteristic in turn, versus those coded 0

^*p < .05

⁺Early Indicators of Later Work Levels, Disease, and Death project; University of Chicago Center for Population Economics

⁺⁺Integrated Public Use Microdata Series, University of Minnesota