Incarceration and the Labor Market

by Kieran Healy on August 21, 2003

Devah Pager has won this year’s Dissertation Award from the American Sociological Association. (I wrote about her work last year. It’s worth mentioning again.) Devah studies the effect of incarceration on labor market outcomes. Her approach was to conduct an audit study of employers, sending in applications for real jobs using vitas for matched pairs of black and white men. The abstract of a working paper from the study says, in part:

bq. With over 2 million individuals currently incarcerated, and over half a million prisoners released each year, the large and growing numbers of men being processed through the [U.S.] criminal justice system raises important questions about the consequences of this massive institutional intervention. This paper focuses on the consequences of incarceration for the employment outcomes of black and white job seekers. … By using matched pairs of individuals to apply for real entry- level jobs, it becomes possible to directly measure the extent to which a criminal record in the absence of other disqualifying characteristics serves as a barrier to employment among equally qualified applicants. I find that a criminal record is associated with a 50 percent reduction in employment opportunities for whites and a 64 percent reduction for blacks.

Pager found that blacks “are less than half as likely to receive consideration by employers relative to their white counterparts, and black non-offenders fall behind even whites with prior felony convictions.” In other words, even though being black and having served time both negatively affect one’s employment opportunities, controlling for education and skills you are better off being a white male with a felony conviction than a black male with no criminal record.

To put Devah’s work in context, take a look at Bruce Western’s research in this area. “How Unregulated Is the U.S. Labor Market? The Penal System as a Labor Market Institution” is a good place to start. It’s a comparative macro-sociological account of what’s been happening to the U.S. penal system. A recent working paper co-authored with Becky Pettit, “Inequality in Lifetime Risks of Incarceration” estimates risk of imprisonment by race and education. Here’s the abstract:

bq. Although growth in the U.S. prison population over the last 25 years has been widely discussed, few studies examine changes in inequality in imprisonment. We study penal inequality by estimating lifetime risks of imprisonment for black and white men at di erent levels of education. Combining administrative, survey, and census data, we estimate that among men born 1965–69, 3 percent of whites and 20 percent of blacks will have served time in prison by their early thirties. The risks of incarceration are highly stratified by education. Among black men born 1965–69, 30 percent of those without college education and nearly 60 percent of high school dropouts went to prison by 1999. The novel pervasiveness of imprisonment indicates the emergence of incarceration as a new stage in the life course of young low-skill black men.



Phil P 08.21.03 at 8:08 pm

I like the phrase “risk of incarceration”. Rather like catching the flu.


Ophelia Benson 08.21.03 at 9:00 pm

Yeah, Phil has a point. I often have a similar thought when people say things like ‘More than 50% of the people (or whatever the figure is) in US prisons are black or Hispanic!!’ Um…could that be because they commit more crimes?

Of course it could also (or instead) be because certain crimes, to wit those committed by blacks and Hispanics, are more likely to get prison time than other crimes more likely to be committed by whites. And it is also worth studying the whole subject, researching why certain groups commit more crimes or more crimes of a certain kind, the consequences of mass-incarceration, etc. All that. But the rhetoric does tend to be truncated, so that the connection between committing crimes and winding up in prison gets…elided. I think this is a mistake.


Kieran Healy 08.22.03 at 2:26 am

There’s a big gap between commiting a crime — ie, doing something that is on the books as being illegal — and getting sent to prison for it. Saying “risk of incarceration” is not a bleeding-heart lefty way of denying that people are responsible for their actions. It’s an acknowledgement of the empirical fact that people with different characteristics (sex, race, education, whatever) can do the same illegal thing and have very different chances of getting sent to prison for it. This is so for all kinds of crimes. See Matt Wlech’s recent article in Reason for a libertarian version of the same point.

Whether to send someone to jail for a crime is not a choice that the criminal makes. Criminal justice policy and the practices of the people in the criminal justice system decide that. It is not a simple function of how much crime there is. Young men get into trouble, and do illegal things, all the time. What happens to them when they do is very much a consequence of whether they are well-educated and white or unskilled and black.


back40 08.22.03 at 4:30 am

Welch’s article made the very different point that felons can’t vote, and that since large numbers of black males have committed felonies and been caught that elections are affected.

His issue isn’t about whether one should go to prison for doing crimes, it’s about whether a given crime should be classified as a felony and whether felons should be disenfranchised forever though they have done their time. Historically, there is a racial component in felon voting restrictions which undermines arguments that felons should be denied the vote.


Kieran Healy 08.22.03 at 4:53 am

Welch’s relevance to this discussion is that he makes the plausible claim that many people have committed felonies. Yet not nearly everyone who commits a felony spends time in prison. The selection mechanism that closes this gap — that turns people who commit felonies into convicts who serve time — is what’s of interest here. Its skewness by race and education is what allows us to speak of risks of incarceration.


Doug 08.22.03 at 2:26 pm

If we get further into racial disparities, please bear in mind that something like 85% of the crimes committed by black people are also black. Letting perpetrators off more lightly when they plead the racism of the system does a terrible disservice to their victims.

Also, when I worked for the Tennesseee Sentencing Commission back in the late 1980s, we found that the biggest source of disparities by far was the jurisdiction where the crime was committed. Crimes that drew short sentences or probation in Memphis (where most criminals and victims were black) could well draw years of incarceration in rural counties that were overwhelmingly white.


Ophelia Benson 08.22.03 at 5:47 pm

Sure. I agree with what Kieran said – my point is just that the shorthand versions can be misleading – and can for that very reason come across as, precisely, bleeding heart or woolly denials, thus prejudicing people against their arguments in advance.


johnx 08.25.03 at 4:35 am

Perhaps using race isn’t a good measuring tool.

Perhaps using economic numbers is better.

Or perhaps educational levels.

Just do not use a chicken to measure it…


Anarch 08.25.03 at 9:39 am

please bear in mind that something like 85% of the crimes committed by black people are also black.

I’m just praying there’s a missing noun in there somewhere…


Anders 08.25.03 at 3:01 pm

I think it’s very interesting that Kieran blogs a study about employment discrimation that has an appalling conclusion — not only are whites with criminal records more likely (3x) to get a job interview callback than blacks w/ criminal records, but that whites w/ criminal records are more likely to get called back than blacks _without_ criminal records — and it turns into a discussion about crime.

If the people who’d been discriminated against, were Americans w/ an Irish or Jewish background, do you think we’d see the same reaction?


pj 08.25.03 at 6:55 pm

doug’s sentence should have read please bear in mind that something like 85% of the crimes committed by black people have black victims. I don’t have data handy, but I believe for murder the percentage is over 90%. It’s also the case that the juries that judge black criminals are more black than the rest of the population, because the jury is drawn from the locality where the crime was committed, and most criminals commit crimes close to home.

So, for instance, the death penalty statistics — (1) black murderers are much less likely to get the death penalty than white murderers; (2) murderers of blacks are much less likely to get the death penalty than murderers of whites; and (3) black jurors are much less likely to vote for the death penalty than white jurors — are three statistics referring to the same disparity.


kathy 08.25.03 at 10:56 pm

“Risk” is just a statistical term. When you figure probabilities and hazard models and the like, you are searching for the “risk” of event A or event B actually occurring.
If everyone who committed a crime was immediately caught and convicted then you could assume a 1:1 correspondence between # of people of a particular race committing criminal acts committed and # of people of a particular race in prison. But since there are steps between the act and the conviction (the arrest, the decision to prosecute, etc) you can’t make this leap.
One study of drivers on the New Jersey Turnpike in 1993 found that “African Americans made up 13.5 percent of the turnpike’s population and 15 percent of all speeders. But they represented 35 percent of those pulled over. In stark numbers, blacks were 4.85 percent more likely to be stopped.” A similar study in Maryland 1996-97 found that blacks were 17.5% of violators, 28.8% of those pulled and a whopping 71.3% of those searched (See John Lambeth, “Driving While Black” Washington Post Aug 16, 1998).


zizka 08.26.03 at 2:21 am

People seem to have been overeager to change the subject: Why do white felons find it easier to get jobs than black non-felons? This all-star blog doesn’t seem to have all-star commentators.

The discepancies of the US justice system are most obvious in sentencing for drug crimes. Crack cocaine offenses are notoriously sentenced more severely than other cocaine offenses, since crack is used much more by blacks. But I believe that the discrepancy is across the board at every stage of the process. Black offenders are just NOT “good people who made a mistake and need help”.

(PS. Well, people didn’y really change the subject. They just ignored the lead and focussed on the second part).


Larry Kestenbaum 08.26.03 at 1:13 pm

Very much agreed with Zizka, but I must respond to a comment above.

The disparity between Memphis and other Tennessee counties is similar to disparities in most other states, and has nothing to do with race.

The U.S. criminal justice system tends to be set up on a county level; the allocation of resources is such that urban counties get shortchanged compared to rural counties. Given the number of reported crimes or arrests, large urban counties usually have drastically fewer judges, prosecutors, and all the rest, per capita or per crime, than smaller-population counties.

The effect of this is dramatic. A person who is arrested for, say, armed robbery in an urban county is one of a great many. Such a person is likely to be offered a plea bargain, because the prosecutor’s office doesn’t have the resources to go to trial in more than a handful of armed robbery cases.

By contrast, in a smaller county, a person arrested for armed robbery is the center of attention. The police, judges, and prosecutors have ample time for the case, and have no reason to plea bargain. The sentence imposed is likely to be CONSIDERABLY greater than the usual sentence for the same crime in the large urban county.


Kenneth Hundzinski 08.28.03 at 2:42 am

My comment may be somewhat unrelated to the specific topic but it has some relevance. I am a white software specialist who is out of work and have few prospects for the future. If recism is so prevalent why do you think that over one million workers, many of them from India, have been admitted into this country to work in my profession? Are Indians considered white? Do cheap salaries mitigate feelings of racism?


zizka 08.28.03 at 4:30 am

In the US the right wing is contested between the racists and the employers of cheap labor. The solution is to admit lots of non-white workers and then treat all workers like shit. Sorry, buddy. You lose.


Meban 02.20.04 at 1:05 pm

Thee best bloggg

Comments on this entry are closed.