Incarceration and Unemployment: U.S. and Europe
Ever since Bryan offered this bet on future unemployment rates in the U.S. and Europe, I’ve been wondering: how do incarceration rates affect those numbers?
Europe has consistently higher unemployment than the U.S., but the U.S. has far and away the highest incarceration rate in the world — .75% of the population. (World Prison Population List [PDF], compiled since 1992 by Roy Walmsley of the International Centre for Prison Studies.)
Only Russia comes even close, at .63%. (Canada: .12%. Australia: .13%. China .18%. Germany .09%.) Our rate is four to eight times that of most other countries.
Prisoners aren’t part of the unemployment calculation. They’re not counted as part of the work force, and they’re not counted as unemployed. There are various arguments about whether that makes sense (feel free to comment), but if we include them in the calculations, what do unemployment rates look like? In particular, Bryan’s bet makes me curious: How does U.S. unemployment compare to the EU15?*
Here are the numbers for 2008. Calculations based on labor force and unemployment figures from Eurostat.
Percent of (total work force + incarcerated population)
| Incarceration | Unemployment | Total | |
| EU15 | 0.2% | 7.1% | 7.3% |
| U.S. | 1.5% | 5.8% | 7.3% |
In 2008, all of the difference between EU15 and U.S. unemployment rates is accounted for by the prison population. The cynical view would say that we just imprison our unemployed, which doesn’t strike me as the most economically efficient arrangement. (Here putting aside any foolish notions of Christian charity or the like.)
I would have liked to do a fever graph comparing unemployment rate, incarceration rate, and combined rate over the years, but the data’s only available in fairly intractable country-by-country form, and I didn’t have time or energy for all the cutting and pasting. (I wrote to Mr. Walmsley and he was nice enough to reply, but he was unable to provide the data in a more usable form.)
Note: Eyeballing the data, I do not think incarceration accounts for the (significantly larger) differences in previous years. (I would suggest that the additional difference is mostly the result of labor-market and other market rigidities imposed by unions and government regulation — not the result of redistribution. But that’s another post.)
Perhaps one of my gentle readers might have the time and inclination to compile those years? Or, write to Roy Walmsley and ask him to provide a simple table of the data that only he has: prison population by country and year. All the other data, at least for OECD countries, is easily available.
* Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, United Kingdom.
I should amend:
Europe has *had* consistently higher unemployment than the U.S.
In 2009 and 2010, by some measures they were the same:
http://www.asymptosis.com/wp-content/uploads/2012/01/Screen-shot-2012-01-16-at-8.11.52-AM.png
Ooops: http://www.asymptosis.com/?attachment_id=4777
What portion of the US prison population is employed though?
How extensive are things like PIECP (PIECP is not that extensive, fewer than 10k, I mean how extensive are “things like” PIECP).
Do prisoners on “work release” count as incarcerated? Do they count as employed?
There is also a big difference in the statistics counting the unemeployed. Many, if I am not mistaken, more than half, of unemployed Americans don’t qualify for unemployment compensation, how are they counted? Employers can and do oppose payments, and the length of time is much less than in Europe.
How do we compare?
Keeping students in college would keep them off the list too.
@Lysistrata: These are OECD “harmonised” unemployment rates. I haven’t gone deep on what that means, so here we’re pretty much relying on them to do the best job possible of resolving those discrepancies. FWIW…
@J.Goodwin: I have no idea. Also: do European countries have similar programs? I would guess that the numbers are very small in terms of affecting the numbers shown here.
http://www.youtube.com/watch?v=NuvxTvHhqsc&feature=player_embedded
Steve:
Unemployment is calculated from the Noninstitutional Civilian Population which would exclude the military, prison population, all institutionalized, etc. Hence the population as calucalted by the BLS is ~254 million and not the 300+ million of the normal population.
Lys:
You have to remember, Unemployment is calculated by doing a survey of people and U3 is calculated from the survey. It is within that survey, the questions you are asking about are discovered. Who is or is not on Unemployment Compensation is not a determinant of the resulting numeric. The same would also apply to those who do not qualify for Unemployment Compensation such as Chrysler White Collar who took the buy-out and could never find a job.
Michigan U3 was much greater if we were to take into account the buyout people.
@run:
Actually, U3 unemployment % is unemployed/*work force*, not /NCP.
The whole point here is adding incarcerated pop to both the numerator and the denominator.
I remember during the Reagan years the government decided to add the military personell to the labor statistics as employed, the goal was to manipulate unemployment rates down. I have no idea how they were counted before, it was still cold war and the army was huge. I realize, that kind of manipulation is common, but it is interesting.
An additional consideration is that the high incarcerated populations in the US has a sort of multiplier effect, producing a larger number of police, legal, judicial, and penal jobs. In addition, there must be continuing prison construction and improvement, which means construction jobs. I would guess that US incarceration is not only taking people out of the workforce, but also putting people into the workforce. The fair assessment would have to measure the impact of high incarceration on this type of unemployment reduction as well.