Why are black incarceration rates falling?

So it may come as a surprise to learn that for the last 15 years, racial disparities in the American prison system have actually been on the decline, according to a Marshall Project analysis of yearly reports by the federal Bureau of Justice Statistics and the FBI’s Uniform Crime Reporting system.

This story was published in collaboration with Washington Post’s Wonkblog

Between 2000 and 2015, the imprisonment rate of black men dropped by more than 24 percent. At the same time, the white male rate increased slightly, the BJS numbers indicate.

Among women, the trend is even more dramatic. From 2000 to 2015, the black female imprisonment rate dropped by nearly 50 percent; during the same period, the white female rate shot upward by 53 percent. As the nonprofit Sentencing Project has pointed out, the racial disparity between black and white women’s incarceration was once 6 to 1. Now it’s 2 to 1.

As for potential answers:

1) Crime, arrests and incarceration are declining overall.

Those decreases benefit the most incarcerated group: African Americans. Crime rates have been on the decline since just after 1990, as have arrests. Given that both measures disproportionately affect the black community, one theory goes, the overall drop should shrink the racial gap in incarceration, too…

2) The war on drugs has shifted its focus from crack and marijuana to meth and opioids…

3) White people have also faced declining socioeconomic prospects, leading to more criminal justice involvement…

4) Criminal justice reform has been happening in cities, where more black people live, but not in rural areas.

Here is the full piece by Eli Hager, via Anecdotal.


A lot of crime statistics are reported from a beginning point about 1991. But that was when black behavior bottomed out due to crack. It had been notably better in the mid-1980s.

What about the lead poisoning hypothesis, as popularized by Kevin Drum? It would seem that blacks were likelier than whites to live near freeways, so deleading gasoline should have had a disproportionate impact on black criminality.

What about the lead poisoning hypothesis, as popularized by Kevin Drum? I

Drum never attempted an ecological study and hasn't the skills to conduct one.

Yes, he was merely reporting on the hypotheses and studies performed by others and bringing notice of them to a lay audience, possibly with exaggeration and simplification along the way. That is, I think, what "popularizing" means, which is what P Burgos said. It's not like the lead and crime hypothesis is something he invented out of whole cloth-- there were a number of studies long before he wrote about them.

Biggest problem for me was that some asian countries also removed lead from their gas at specific dates. So, where was their collapse in crime?

In terms of crime, Asian developed nations are seeing declining crime rates but their crime high points (some point of 1970 - 1990) were not as high as it was in the European or US crime rates at the time. So Asian nations don't disprove the Lead Hypothesis. (Although I tend believe more of the huge fall in birth rates in 1970s was the primary cause and generational urban populations. I believe that sudden increase in poor population in urban areas will lead to a crime wave that will diminish after a generation survives.)

Living in LA area during the Rodney King riots, it is amazing to think that we are starting a historical drop in crime over the next 25 years and nobody really has a good answer why it happened.


Changing demographics? Fewer young men means less crime.

Better policing?

The high incarceration numbers, extremely high where in some communities the exception is not to have either been or have a close acquaintance who has been imprisoned eventually has an effect.

"Living in LA area during the Rodney King riots, it is amazing to think that we are starting a historical drop in crime over the next 25 years and nobody really has a good answer why it happened."

One possibility is that when things can't get worse, they sometimes don't. The April 1992 LA riots marked a spectacular nadir of self-destructive black bad behavior, a trend downward following the relatively well-behaved years of c. 1983-1985. That the bottom happened around the time of the LA riots is apparent from all sorts of different social statistics on which c. 1991 marks the worst year. My guess is that more than a few black people started to decide to behave better after that disgraceful event.

"It would seem that blacks were likelier than whites to live near freeways"

Is that actually true? For example, I went to school from K-12 in Sherman Oaks, CA, home to the busiest freeway interchange in America in the 1960s-1970s (the 405/San Diego Freeway & 101/Ventura). I later lived and went to B-school at UCLA near the interchange of the 405 and 10.

Similarly, Drum's examples from Chicago, such as attributing the high murder rate in the 1970s in the Austin neighborhood on the west side to the Eisenhower Expressway being routed not all that far from the neighborhood seem curious to somebody who know Chicago. Most of the children who grew up in Austin in the 1950s-1960s were driven out by violence by newcomers who grew up elsewhere, such as in the freeway-free state of Mississippi.

The lead leads to crime hypothesis is interesting, but the attempt to explain high black homicide rates using it is based on cherrypicked examples that aren't even persuasive if you actually know the maps of the cities involved is eye-rollingly dumb.

It would be better to research the question using the 90 or so sites on the EPA's Superfund list of lead pollution.

There are other possible sources of lead exposure besides living near highways. Residing in older housing stock with lead based paint chipping off walls is probably a greater source of environmental lead than living near a highway.

The lead=crime hypothesis folks tend to be slippery about specifying their theory enough to be falsifiable. They talk about lead in gasoline a lot, until you start suggesting ways to empirically test their theory, at which point they switch to lead in paint.

There ought to be ways to statistically test the lead in paint theory as well. For example, big post-WWII housing projects were promoted on the grounds that they would get poor children out of tenements where they would eat chips of lead paint off the wall. Did children raised in modern urban housing projects, such as Cabrini-Green in Chicago, have lower crime rates than children raised next door in decaying slum buildings?

I don't know, but it seems like something that could be investigated.

A lot of crime statistics are reported from a beginning point about 1991.

The disparity in the frequency of homicide has remained about the same over more than 35 years. It's just that there are comparatively fewer homicides across the board.

The black-white homicide rate ratio has gone up and down somewhat over time, depending apparently upon whether or not there was a drug war going on. The mid-1980s had a relatively low black to white homicide ratio, presumably because there wasn't much new going on involving fighting over drug dealing turf between the peak of the powder cocaine wars in 1980 and the immense crack wars of 1988-1994.

By the way, the historically high homicide-offending rate of black youths born in the half-decade _after_ Roe v. Wade ought to raise doubts about the Levitt-Freakonomics theory that legalizing abortion in 1969-1973 lowered crime rates in the 1990s.

In contrast, the lead-->crime theory isn't as easily falsified. Unfortunately, that's because it's poorly specified relative to the legalized abortion theory. When and why the effects happening are supposed to happen (e.g., at what age) are slippery, as is even what is the mechanism: lead in gasoline? Lead in paint?

For example, the Chicago Tribune editorialized for the construction of the notorious Cabrini-Green housing project to get poor children away from peeling lead paint in the tenements. That accomplishment didn't seem to cut the homicide rate in the vicinity, however. (Tearing down Cabrini-Green in the 21st Century, however, has done wonders for the homicide rate on the Near North side of Chicago, however, although not necessarily for that of the city of Chicago as a whole.)

Of course, the homicide-offending rate among blacks has skyrocketed since Ferguson in August, 2014, with the big increases in the homicide rate coming largely where BLM enjoyed its greatest political victories (e.g., St. Louis, Baltimore, Chicago), and often not happening where BLM lost politically (e.g., in NYC, where the BLM-sympathizing mayor kowtowed to the NYPD's soft mutiny in early 2015).

The total number of homicides in the U.S. was 20.3% higher in 2016 than in 2014, one of the biggest 2-year jumps in recorded history.

DOn't know about the black offending rate, but the white offending rate has also increased by 25% over the same time period. I would suspect that is correlated to the white offending rate.

Is the decline steady over that period, or does it change with the administration in 2009?

Steady. See for yourself, the graph is in the article.

Racial relations in America are getting better since late 19th century, the problem is its pace.

Day to day in meatworld, that's true. In the civic and political sphere, that was true up until about 1964. Since then, there's been improvement in some respects and decay in others.

good stuff for virtual marketing and earn some money.

Considering the quality of the comments that this post is inevitably going to generate today, I give this comment a big fat +1.

Not all the news is good. From the article: "Even with all of these factors at work, the racial inequity of the American prison system remains vast and continues to wreak devastation on black and Latino communities nationwide. At the current rate, the disparities would not fully disappear for many decades. Even more troubling, racial divides in the juvenile justice system are getting significantly worse. In 2003, black youth were incarcerated at 3.7 times the rate of white youth; by 2013, that number had grown to 4.3."

Here's my anecdote: I grew up in a small southern town, where blacks were routinely arrested for minor offenses. Whites, at least whites that I knew, were never arrested. White teens, if they were arrested for a minor offense, were treated more like the children they were and sent home to their parents for the parents to discipline them. Fast forward to the small southern town I reside in now. White teens are routinely arrested and incarcerated, not sent home to their parents. I am shocked by the number of young whites, mostly boys, who are arrested for what I consider minor offenses, teens being the children they are not criminals. This is partly due to the zero tolerance policies adopted in many small towns, but it's also a reflection of the number of young whites who seem aimless, not willing or able to stay in school or to find and keep a job. The divide between the ambitious white children and the aimless white children is vast, the latter group at high risk of arrest and incarceration. There's always been a divide, but nothing like what I see today. This may come across as the ramblings of an old man, but these are young whites from successful families, the kind of families that one would least expect to have children who would be arrested and incarcerated.

What are the white teens commonly being arrested for?

This is the bad side of "Criminal justice reform". My father in law police lieutenant said he had pity for the kids today (20 years ago). Cops could just give the kids a whack with the nightstick and send them on their way, no record. They could also take the kid home to the parents, but instead of the parents disciplining the kids, they abused the cop for discrimination, ect. Easier for the cop just to arrest them now.

+1 I like your theory. I once saw a cop car purposefully hit a bad youth he was going slow enough to not kill the kid but he send him sprawling on the payment. The kid probably got some serious road burn. No need to arrest him and do all that paper work. I wonder if they would be a away with that today.

It's possible that equalizing the disparity will incentivize whites to address the problem and cut back on overzealous policing in general. As long as cops can get away with arresting ONLY the black kids while letting the white ones off the hook, there will not be enough political pressure to reform, since blacks are too small of a minority to influence law enforcement policies on their own. In other words, right now, more white kids are getting arrested needlessly, which draws the attention of white parents to the issue.

The problem isn't that kids are getting arrested needlessly, they are really doing illegal stuff. Before they could get a course correction, now it's juvi, which makes it worse.

It is also the parents. The cops could expect the parents would discipline the kid instead of making up excuses. My own sister in law is no better. She had two foster kids, both black. Kid A was a nice kid, and B got into trouble. S-I-L thinks it might be racial that B got questioned for something (when A never does). I asked her maybe it's because he's already been caught stealing 3 times already.

Or it took longer for lead exposure to decline in predominantly black areas and the impact of lead reduction in predominantly white areas has already fully played out...

Amusing how a completely speculative hypothesis offered by a journalist with zero background in sociology or psychology is taken for granted as true.

The hypothesis predated Kevin Drum and has been studied by others. I grant that it's interesting what people latch on (similarly, how some people take IQ points as important and meaningful when the effect of lead is being discussed, but shun them in other contexts), but it's not like the lead crime hypothesis is his other than, as mentioned above, for his responsibility in popularizing it.

Likewise the abortion argument, which came from Freakonomics.

I like how people insist IQ is 99% genetic and then proceed to discuss lead poisoning as if that has nothing to do with the subject.

Once again, where is the evidence that blacks on average were exposed to more lead?

I can cite anecdotal evidence to the contrary until the cows come home. For example, my cousins grew up in Arcadia, CA between the 10 and 210 freeways. Arcadia HS now typically has a few dozen national merit semifinalists in each class. (Of course, many of the semifinalists grew up not in Arcadia but in highly polluted Chinese cities.)


Here is a link to a Kevin Drum post that claims to cite CDC data on leaf poisoning levels in white and black children going back to the 70’s, in case you are interested. I myself didn’t go back to the CDC source data to verify, because, well, I am just posting comments on a blog for shits and giggles, and I don’t make my living writing. Obviously there is more going on to higher rates of black criminality than just lead, but lead does seem to be part of the picture.

If you read the blog Second City Cop, http://secondcitycop.blogspot.com, they would seem to argue that many urban police forces have gone "fetal" i.e. less enforcement, fewer arrest. Fewer crimes are solved.

The growing hostility to police and a culture of non-cooperation has increased. "No snitching" leads to underreporting of crime and fewer convictions.

In addition, the length of sentences seems to have been shortened with many crimes being pleaded down or even dismissed. Courts are less likely to impose stiff sentences. In addition, criminals are less likely to serve their full sentence, quicker parole.

Many blacks are moving south away from northern urban areas. Perhaps the way these communities respond to criminal activity is different.

No Snitch culture in black community





Rather than simply criticizing 'black community' as an out-group, if you want to understand the self-preservational incentives that drive 'no-snitch culture' and retaliatory shooting cycles, and gain insight into how police can pro-actively change those dynamics and dramatically reduce violence, I highly recommend David Kennedy's book Don't Shoot. A close second would be Jill Leovy's Ghettoside.

Where did I criticize the "black community"?

The black community is in many ways a separate sub-culture, as you yourself claim when you say that I need "to understand the self-preservational incentives that drive ‘no-snitch culture’....." I identified a pattern of behavior in a self-identifying group.

For example, the fact that only about 17% of homicides in Chicago are solved is, in part, because of the anti-snitch culture that exists in the black community. That can, in part, explain why people are spending less time in jail for these killings.

BTW the conviction rate on homicides in Chicago is at historic lows.

I quick skim of the Kennedy book shows how quickly technology advances. Open drug markets are in decline in many communities in part because of the growth of "Obama phones." Drug dealers and buyers can now arrange meetings through texts or calls with dealers driving to the customer.

Maybe Obama phones have reduced violence?

With this post

Is Tyler getting soft on crime?

Is Tyler getting weak on borders also?

We need Jeff Session's War on Legal Marijuana to

Make America Great Again,

And get those crime statistics going in the right direction.

How can we have successful private prisons if we can't fill them?

Blacks still commit a highly disproportionate amount of crime(resulting in more incarceration) but the disproportion has declined since the 2000s. A lot of it is simply that violent crimes, where the disproportion is higher, have fallen faster than property crimes, though the disproportions have fallen within crime categories as well.


That's a hypothesis based on statistics, with some explanatory power. The other question is what fraction of latino convicts were reported as white at the various time periods. This also might skew the ratios beyond behavioral changes.

I also wouldn't be surprised if Latino behavior is changing over time. Nobody pays much attention to Latinos, but I wouldn't be surprised if old-time Chicanos in Los Angeles with roots in northern and central Mexico tended to be more violent than newer Mexican immigrants from central and southern Mexico. In Mexico, the northern half has the reputation of being the more violent half of the country. It's cowboy country while the southern half is more peasant culture.

For one thing, old time Chicanos in Los Angeles in, say, 1975 were bigger than more recent immigrants from Mexico.

3) White people have also faced declining socioeconomic prospects, leading to more criminal justice involvement.

I would like for people to always admit that this could go either way. Declining socioeconomic prospects can lead to healthier living with less drug, alcohol and gun use, because those things become less affordable.

Though this is true: "Proverbs 6:30 People do not despise a thief if he steals to satisfy his hunger when he is starving." how much of crime is theft?

The poor tend to be obese.

But maybe those are the ones who just gorge on food and don't have a drug problem.

In New York, authorities invested in both boots on the ground and prison cells, but put less emphasis on the latter than did other states. New York trebled its incarcerated population, whereas other states increased theirs 6 or 7 fold.

Suggest the reason for the phenomenon you're seeing is that ca. 1990 bang for the buck was most intense in investing in manpower to deter slum crime.

I agree with conservative Art Deco here, I do think that more and better police would help. IMHO more funding for schooling has not proven helpful and money should be shifted from schooling to more better policing which may even improve the schools as life becomes safer.

Better to live on the corner of a housetop that amid violence.

There are declining returns to incarceration, and the police state known as New York City is a good example. They did do better than a lot of other places in the crime drop, but not that much better.

If I remember correctly, the best incarceration did was contribute to 25% of the crime drop.

3) White people have also faced declining socioeconomic prospects, leading to more criminal justice involvement…

They haven't, except in certain loci and in certain industries.

So, they haven't, except when they have

Part of the decline in numbers is two fold. Overcrowding and the enforcement of grand theft penalties. With the Black Lives Matter movement I would expect the trends to accelerate. Why risk your job and pension on a criminal? Enforcement is just not worth it for most nonviolent crimes.

Isn't 1) the obvious reason here that the African-American communities are so much better off today than in 1991? Of with less crime there is less incarceration of the population and hopefully this continues.

2) In terms of modern opoids, it is amazing that it is the first drug epidemic to center on white populations but it is also very cheap in general to produce and sell so it limits a lot of the secondary 'drug' crime such as gang murders and property theft.

3) Probably should mention the rural/semi-rural white populations are in the most economic stagnant economies. A lot of white people are doing well today just not all populations.

Sam Quinones's book Dreamland says that Mexican heroin retailers in the U.S. avoid black cities, too much violence and drama, and focus on white areas.

I suppose I could wait around all day for you guys to figure out that incarceration rates are down because the freaking prisons are full and nobody wants to build any more, but I'm just going to post it myself.

And I do so at the risk of raining on your idea that when fewer criminals actually spend time in prison, the world must necessarily be getting better!

"nobody wants to build any more"

That's the real story. Perhaps it's a good thing and perhaps not, but there's no shortage of concrete, steel, or construction workers.

I guess you all wanted to ignore the fact that crime has been dropping since 1994? The Republicans just took 15 more years to realize that it was no longer worth it to invest in prisons.

I promise you, prisons are not overcrowded. They are losing population every day because people aren't committing as much crime anymore.

People are making the lead in gasoline argument for African Americans, and trying to explain that they were closer to interstates (which is a good argument), but there are better arguments for higher exposure to lead, such as their concentration in dilapidated housing with lead still in the paint literally peeling off of the walls, or higher exposure to lead in their water.

Comments for this post are closed