America’s Declining Prison Population
We've seen a massive fall from the 2009 peak.
Stanford psychiatry professor Keith Humphreys assesses why “America’s Incarceration Rate Is About to Fall Off a Cliff.”
For more than 40 years, the United States—a nation that putatively cherishes freedom—has had one of the largest prison systems in the world. Mass incarceration has been so persistent and pervasive that reform groups dedicated to reducing the prison population by half have often been derided as made up of fantasists. But the next decade could see this goal met and exceeded: After peaking at just more than 1.6 million Americans in 2009, the prison population was just more than 1.2 million at the end of 2023 (the most recent year for which data are available), and is on track to fall to about 600,000—a decline of roughly 60 percent.
Discerning the coming prison-population cliff requires understanding the relationship between crime and incarceration over generations. A city jail presents a snapshot of what happened last night (for example, the crowd’s football-victory celebration turned ugly). But a prison is a portrait of what happened five, 10, and 20 years ago. Middle-aged people who have been law-abiding their whole life until “something snapped” and they committed a terrible crime are a staple of crime novels and movies, but in real life, virtually everyone who ends up in prison starts their criminal career in their teens or young adulthood. As of 2016—the most recent year for which data are available—the average man in state prison had been arrested nine times, was currently incarcerated for his sixth time, and was serving a 16-year sentence.
So, what happened? Basically, a crime wave that started in the late 1960s led to draconian sentencing laws, which led to mass incarceration. But the wave has long since crested.
The U.S. had an extremely high-crime generation followed by a lower-crime generation, meaning that the older population is not being replaced at an equal rate. The impact of this shift on the prison population began more than a decade ago but has been little noticed because it takes so long for the huge prison population of longer provenance to clear.
But such a transformation is now well under way. One statistic vividly illustrates the change: In 2007, the imprisonment rate for 18- and 19-year-old men was more than five times that of men over the age of 64. But today, men in those normally crime-prone late-adolescent years are imprisoned at half the rate that senior citizens are today.
This is good news all around:
The benefits of a smaller prison population are not limited to those who would otherwise be locked up and the people who love them. Prisons crowd out other policy priorities that many voters would like the government to spend more money on. In all 50 states, the cost to imprison someone for a year significantly exceeds the cost of a year of K–12 education. But even greater than the financial savings would be the prosperity in human terms: Less crime and less incarceration are profound blessings for a society.
[…]
Society is going to experience the benefits of past decades of lower crime throughout its prison system. The imprisonment rate will be lower in five years and lower still in 10. Prisons will still exist then and still be needed, but the rate at which Americans are confined in them could be lower than anything in the preceding half century. This is the fruit of a lower-crime society—good in and of itself, surely, particularly for the low-income and majority-minority communities where most crime occurs. It will also, of course, be a blessing for those who avoid prison, and for the taxpayers who no longer have to pay for it. The decline in the prison population will be something everyone in our polarized society will have reason to celebrate.
Humphreys offers no explanation for why crime spiked in the late 1960s and declined in the early 2000s. The late Kevin Drum would surely point to the gradual removal of lead from our environment. While I’m sure that’s part of the explanation, my guess is that there are multiple factors involved.
This is good news. Part of the revisiting of sentencing and other forms of prison reform was due in part to the fact that mass incarceration was set to bankrupt States.
To be fair, violent crime was on the rise until the 1990’s and has since fallen off. There are numerous theories about what led to the decline in crime (and the fact that the so-called “super predators” never materialized). I think there’s a lot to the lead theory (alongside the strengthening of the social safety net).
I appreciate the call-out to Kevin Drum. This was something I will always remember about him.
Texas prisons are under pressure to close because of 1) demand and 2) inability to staff.
Texas should close prisons and jails with staffing challenges, state oversight panel says
As the Texas prison population shrinks, the state is closing two more lockups
Here’s another thought. Since immigrants cause less crime, the need for prisons are declining.
The resistance to closing prisons is high in Texas rural areas since they are a big source of employment.
The growth of mental health diversion programs have worked to keep people out of jails. Harris County (Houston) Sheriff’s dept had the reputation of being the largest provider of mental health in Texas.
They created the Harris County Mental Health Jail Diversion Program to offload cases where mental health would be a better option than jail.
But what about the American Carnage? THE AMERICAN CARNAGE???
I believe I’ve read that the private prison companies are ramping up…. only for immigrants, then?
Demographics is king. This under a minute video from Pinkerton makes the point that most crime is done by those in teens to early twenties. And the pattern is consistent over generations varying in amplitude
In 1993, a late 20-something guy who worked for me opined, “The Baby Boomers grew up to make everything they did illegal for their kids”.
Not strictly true, the “tough on crime” were those who came of age in the 1960s, of Joe Biden’s generation, with some help of early Boomers.
Kevin Drum made, as I recall, a pretty good case for his lead hypothesis. The effect of lead on impulse control is well known. The temporal correlation was strong. The lag from banning leaded gas to the drop in crime neatly matched the lag from childhood to the crime prone late teens. The spatial correlation was strong with crime rates matching traffic density where the urban freeways converged. And there was a temporal-spatial correlation, with the decline in crime in other countries in step with the timing of their banning leaded gasoline.
Correlation is not causation, but I was struck by the resistance to the hypothesis. Was it because people who study crime are political scientists, sociologists, legal scholars, political scientists, etc., and lead was outside their ken. Was it because we’d already banned leaded gas and paint, so there was no opportunity for new programs, or expansion of police departments, or profitable lobbying?
Or was it because it made inner city crime not something minorities did, but something we did to them?
Finally, good news!
Yes, it’s an overdetermind outcome. I suppose also some are not ready to admit the “Biden Crime Bill” worked.
@gVOR10:
I’m not aware of significant resistance to that hypothesis. A lot of economists involved in Criminology have been receptive to it. The biggest push back I saw was “we don’t yet have the data to back that up. That’s been changing in the last decade or so as more studies and meta-studies are being published. Honestly it’s still a relatively new theory–only really starting to gain traction in the mid 2000’s.
I think there was more significant pushback to the argument that the legalization of abortion played a major role in declining crime rates. And much of that was tied to the overall “Freakamnomics” approach.
@Scott:
Texas was the first major state to realize the budget issue being caused by prisons.
For the record, closing a prison is a logistical nightmare–even closing down a section (or pod) of a prison because of issues related to staffing and the fact that if you drop below a certain threshold you can’t simply close a section because there may not be enough beds in the remaining sections of the prison to absorb the displaced incarcerated folks.
Also, it’s worth noting that corrections unions are almost as powerful as police unions and that creates a set of perverse incentives to keep incarceration up. And the sad reality is many of the communities prisons are in have based their local economies on the presence of the prison–not to mention the entire topic of prison gerrymandering, which Texas engages in–which is typically used to maintain political power in rural areas: https://www.dallasnews.com/news/investigations/2021/12/15/prison-gerrymandering-how-inmates-are-helping-the-texas-gop-maintain-its-power/
@Scott:
Diversion programs (in which people are required to complete different forms of treatment as an alternative to prison–if they fail to accomplish those tasks they are sent to prison) have been a great tool for prosecutors to help keep prison populations down (and prevent recidivism).
@JKB:
Man, thank goodness no GOPers came of age in the 1960’s and contributed to the mass incarceration problem. And that the spry current occupant of the White House, who is just a few years younger than Biden neither falls into that category or has a long-established record of being “tough on crime.”
But partisan brainworms for the win, AMIRIGHT?!
@DK:
Count me among those people. While I think some negative aspects of its impact have been overrated–and a lot of people have left out context about the support it had within Black communities at the time–it still was responsible for a lot of our mass incarceration issues (and the harsh sentencing of non-violent drug offenders). The three-strike rule and mandatory minimums were quite bad. Ditto the enhanced sentencing for crack. It’s telling that most of those had to be revisited or reversed by future administrations.
Ultimately, I don’t think any legit rigorous criminologist will say that it played a major role in reducing crime rates.
Here’s a pretty good rundown on it: https://19thnews.org/2024/09/the-complicated-legacy-of-the-1994-crime-bill/
That said, most of the people–especially those on the right–who use this to bludgen Biden or the Democrats are full of bullshit. Almost all Senate Republicans voted for it. Fewer did in the House. And the rational for voting against it was that it *wasn’t tough enough* on crime.
I think lead removal is a strong contender as an impact, but not the only impacting factor. I’m sure there are a slew of other things that could be playing a role here, including affirmative action.
From the tone of my NextDoor feed, you’d think crime is worse than it’s ever been.
@Matt Bernius:
They can try, I just don’t know who the audience for it is. Left-wing activists? They already don’t show up for Democrats. Rank-and-file black voters didn’t abandon Hillary and Biden over the crime bill despite the best efforts of of their critics, because black voters and politicians supported those policies.
Do legit rigorous criminologists agree that lead removal played a major role in reducing crime? There seems to be a correlation. But then there’s also a correlation between locking up criminals, the passage of VAWA, and less crime.
@DK:
There is growing acknowledgement of a strong correlation and many are open to causation. Again, a lot of this is based on recent studies. The Wikipedia page for the topic has a pretty solid primer on where things currently stand:
https://en.wikipedia.org/wiki/Lead%E2%80%93crime_hypothesis#Correlation_between_lead_exposure_and_crime
Yes/no. Honestly this is going to go into an academic analysis discussion. Hell, if you are talking about things this broadly, we could also suggest that there was a correlation between mass incarceration and a drop in crime. That doesn’t mean mass incarceration was a good policy.
I hate to play expert here, but I’ve been working in and around the criminal legal system policy and reform space for about a decade. So I’m more familiar with the dialogue and research around the VAWA. I think you will be hard pressed to find any person who studies the space that is going to link VAWA with the current drop in crime.
But if you can find one, then I’m happy to revisit my position.
And BTW, I think the entire reason Republicans went so hard on that issue was to try and peel off younger Black and minority voters. I agree with your overall analysis on how effective that really was.
@gVOR10: I think the pushback came from the fact we don’t have any idea about what the biological/anatomical mechanism by which lead would produce these behaviors might be.
And yes, that is definitely an important part of this puzzle. We may still see something like, “It isn’t just lead, but lead AND something something”.
The most memorable pushback I recall seeing Kevin get is “there is no such thing as a criminal brain cell”. Which I understand to be a sloganized version of what I said above.
(I think the lead hypothesis has some pretty darn good support. I think demographics, as @JKB brings up, might also be a factor, but I don’t think it’s strong enough on its own to explain all the data.)
I thought lead exposure was a pretty widely accepted theory? As recently as 2021 a study showed that lead exposure in childhood can result in “mean, cranky adults.”
@Matt Bernius:
I mean, that is precisely what I am pondering. Is there a correlation between what people call mass incarceration and a drop in crime?
The answer to the question of whether or not “mass incarceration” was/is good policy — which involves subjective value judgments — is separate from the answer to the above question — which involves quantifiable outcomes. I can understand why liberal academics might be biased against even daring to study it. The answer might be inconvenient and uncomfortable for our preferred ideological orthodoxy.
According to Democrats, VAWA is associated with a decline in intimate partner violence, rape, and assault. In academic study of crime, this is considered wholly separate from other tough-on-crime policies, from the same bill? That seems neatly compartmentalized.
(I swear I’m not trying to bait you into writing a long, top-level blog post. But I wouldn’t be sad if you wrote one. Hehe.)
You need not on my account. I’m not threatened by folks having viewpoints separate from mine. I don’t need an Amen corner.
On lead, there were other countries that banned leaded gasoline. Were their crime rates affected as well?
You can’t do a controlled study where lead and lead compounds in the atmosphere are the only variable, but if crime trends down as leaded gas is phased out, one can then compare crime rates along with other factors.
Ruminations of stealing something in the mind of a teenager today are perhaps, to some degree anyway, nipped the bud by knowledge that these days there are cameras everywhere.
@DK:
Thanks for the discussion on this. I have a few more comments based on my reflecting on things and doing a bit more research.
On the Violent Crime Control and Law Enforcement Act of 1994:
1. It’s a huge bill, and as such, it has many parts to it. So I’m not sure there is any way to access the full impact of the bill. As such, we often have to look at its parts and what they were intended to address.
2. As with any Federal Law, it’s important to call out that it only directly affects federal law enforcement. So in terms of mass incarceration, its primary effects were at the federal level. Given that the majority of people are incarcerated at the state level, it’s direct impact on the mass incarceration issue was limited.
3. That said, there are a lot of indirect ways that federal legislation impacts state legislation–both in terms of carrot/stick funding provisions and setting standards that states can adopt. The act was definitely in line with and influenced criminal legal system legislation nationwide.
4. As a number of scholars have pointed out, the contents of the bill are a great time capsule of the complex debates going on within the criminal legal system space of the time.
5. In the end there were parts of the bill that were good (including increasing the amount of policy impact research) and there were parts of the bill that were not good (including “3 strikes” and “creating the crack scheduling at a significantly higher rate than cocaine, and other mandatory minimums.”).
6. Would we still have had mass incarceration without the law? Yes. And that doesn’t negate how the law contributed to that issue.
On the topic of “did mass incarceration contribute to a drop in crime?” and “are academics willing to ask that question?”:
1. The field of criminology (which draws people from a wide range of disciplines) is an example of a field where researchers have always demonstrated a pretty wide range of social/political ideologies. There are a lot of “conservative” scholars in the field. In my experience, there is little fear of running studies and analyses that could lead to “controversial” conclusions. In part that’s because of the realities of the data sets–we know that there is a lot of racial disproportionality in the data. This gets to:
2. Are people afraid to study “did mass incarceration lead to a reduction in crime?” Last I checked this is something that a lot of people in the field are interested in understanding. And yes, an argument can be made that it definitely correlated with a drop in crime. However, it’s also true that we have seen crime rates continue to drop as (1) mass incarceration has been undone, and (2) sentencing reform (and things like diversion programs) have led to shorter sentences (at least from an American POV). So that suggests causality isn’t strong.
3. All that said, there are still people who will defend mass incarceration as a thought experiment. I find that akin to Bill Bennet’s infamous thought experiment:
4. Finally, as noted above, one positive aspect of the ’94 law was that it significantly increased funding for impact research. As a result, it has also become an incredibly closely studied bill. Given that it turned 30 just last year and played a role in the 2020 Democratic Party primaries, there has been considerable coverage of studies on its impact in the last five years.
That’s about it for me. Just wanted to add some additional context and nuance on the topic.