Where Did 2.5 Million Native Working Men Go?

I worked on the revision to my paper on the labor supply of immigrants earlier this month. I also played around with the data some, and was struck by the substantial rise in the number of prime-age native men who do not work at all during an entire calendar year.


The graph uses March CPS data and shows the trend in the percent of men who did not work at all during the calendar year prior to the survey. It is obvious that there was a substantial increase in the fraction of “permanent” joblessness after the Great Recession. Although there has been some recovery, the situation is still dismal for many. Just compare the data for 2000 and 2016: In 2000, 8 percent of prime-age men didn’t work at all. Today, that number is 13 percent. There are about 50 million men in this age group, so that around 2.5 million prime-age men have joined the rank of permanent joblessness since 2000.

It would be a mistake to interpret the graph as suggesting a causal link between immigration and the dramatic rise in permanent joblessness among native men. But the data clearly suggest that if one wishes to understand the economic dissatisfaction that motivated a lot of the political action this year, this graph would not be a bad place to start. It may be that: “It’s all jobs, stupid.”

The very simple STATA code that generates the graph follows, so reproduce to your heart’s content.

*The cps.dta file is downloaded from IPUMS

use cps.dta
replace uhrsworkly=0 if uhrsworkly==999
generate annhrs=wkswork1*uhrsworkly
keep if age>=25 & age<=54
drop if wtsupp<0
generate imm=(citizen==2 | citizen==3)
keep if year>=1994
keep if sex==1
generate olf=(annhrs==0)
replace wtsupp=wtsupp/2 if year==2014
collapse (mean) olf (rawsum) wsum=wtsupp [aw=wtsupp], by(year imm)
reshape wide olf wsum, i(year) j(imm)

scatter olf* year, connect(l l)

Author: George Borjas

I am a Professor of Economics and Social Policy at the Harvard Kennedy School.