[Bernard Harcourt, guest-blogging, May 2, 2007 at 7:15am] Trackbacks
Institutionalization vs. Imprisonment: Are There Massive Implications for Existing Research?

Andrew Gelman at Columbia University writes on his statistics blog here that the findings I discussed yesterday seem correct and don't surprise him in the least. (He also has some entertaining reactions to some of the comments from yesterday).

I'll confess that I not only find the results surprising, I also think that, if they are indeed right, they have farreaching implications for our existing (and future) research on prisons and their effect on unemployment, crime, education, and poverty, as well as our research on gun laws (think of the right-to-carry debates here, here, and here), the effect of abortions (think of the Donohue/Levitt thesis), the deterrent effect of the death penalty (think of the recent debates here), social control and disorganization theories, collective efficacy -- and the list goes on.

In practically all those studies, we have used the imprisonment rate to measure society's level of incapacitation. But the prison rate alone may not capture what we were trying to measure. The most straightforward interpretation of my findings is that neither the rate of imprisonment alone, nor the rate of mental hospitalization alone are good predictors of serious violent crime over the period 1934-2001. In contrast, the aggregated institutionalization rate (aggregating the mental hospitalization and prison rates) is a strong predictor of homicides. This suggests that there is something going on in the relationship between mental hospitalization and prison — perhaps a form of substitution — that should make us rethink entirely how we measure social control and incapacitation.

But since practically none of our studies on prisons, guns, abortion, education, unemployment, capital punishment, etc., controls for institutionalization writ large, most of what we claim to know about these effects may be on shaky ground.

Here's a good example. My colleague Steve Levitt at the University of Chicago has a great paper on the crime decline of the 1990s published in the Journal of Economic Perspectives called Understanding Why Crime Fell in the 1990s: Four Factors that Explain the Decline and Six that Do Not. In the paper, Levitt identifies the prison-population build up as one of the four factors that explains the crime drop of the 1990s.

Levitt estimates that the increased prison population over the 1990s accounted for a 12% reduction of homicide and violent crime, and an 8% reduction in property crime — for a total of about one-third of the overall drop in crime in the 1990s (see pages 178-79). The paper and its progeny have given rise to fascinating debates over the role of the police (Malcolm Gladwell takes Levitt to task in an interesting post here), the abortion thesis, and the role of broken-windows policing vs. the crack epidemic.

What interests me here, though, is that when Levitt extends his analysis to discuss the period 1973--1991, he sticks to the prison population exclusively and does not consider the contribution of the declining mental hospital population (see pages 183-86). As a result, Levitt is surprised that the drop in crime did not start sooner (see page 186). Regarding the period 1973--1991, Levitt writes:

"The one factor that dominates all others in terms of predicted impact on crime in this earlier [1973--1991] period is the growth in the prison population. Between 1973 and 1991, the incarceration rate more than tripled, rising from 96 to 313 inmates per 100,000 residents. By my estimates, that should have reduced violent crime and homicide by over 30 percent and property crime by more than 20 percent. Note that this predicted impact of incarceration is much larger than for the latter [1990s] period." (page 184)

Based on prison data alone, Levitt is left with a significant gap between projected and actual crime rates for the period 1973--1991. Levitt concludes: "The real puzzle in my opinion, therefore, is not why crime fell in the 1990s, but why it did not start falling sooner" (see page 186).

The unexplained difference, though, vanishes if we include mental hospitalization with the prison rate in an aggregated institutionalization variable. I do the math in this paper here at page 1775. The increase in confinement from 1973 to 1991 would have been smaller (because of deinstitutionalization) and, based on Levitt's estimates, this would have translated into a 12% decrease in homicides, not a 35% decrease. Levitt's revised estimate for the total effect of his ten factors on homicide during the 1973--1991 period would be an increase in homicides of 3%, which is not far from the actual reported change in the UCR of a positive 5%.

In other words, using aggregated institutionalization data rather than prison data would eliminate Levitt's disparity regarding the change in homicides. This is just one example that explains a gap. But think of all the other areas where the difference might undermine the results.

Here's another example from the death penalty deterrence debates. The fact is that none of the existing extensive research on the deterrent effect of capital punishment has included mental hospitalization within an aggregated institutionalization rate. Instead, all the studies use prison rates only to get at a measure of incapacitation.

My study includes, as a control variable in the regressions, the execution rate for each state over the period 1934 to 2001. So we can get some idea of what happens when you use aggregated institutionalization rather than the prison rate. The results are interesting: in my fourth model (Model 4 of Table III.1 at page 33), the execution rate is positively related to homicide and statistically significant at .05, suggesting that, controlling for aggregated institutionalization, there may be evidence of a brutalization effect from executions: more executions, more homicide. The statistical significance does not withstand the introduction of demographic and urban variables, and in my most complete model (Model 6 same page) the coefficient is positive but unreliable.

Much has been written recently about the deterrent effects of capital punishment. John Donohue and Justin Wolfers have reviewed the recent studies, including state-level panel data analyses, and conclude that "none of these approaches suggested that the death penalty has large effects on the murder date" (page 841). When I include mental hospitalization, my findings are consistent with these conclusions, but in the process they undermine a lot of other research.

Practically all our criminology has failed to connect the prison to the asylum. For instance, Alfred Blumstein and Joel Wallman, in their account of crime trends in the introduction to The Crime Drop in America — generally perceived as an authoritative compilation — never address aggregated institutionalization. With regard to the sharp increase in crime in the 1960s, Blumstein and Wallman hit on all the usual suspects — the baby-boom generation, political legitimacy, economics — and include later the usual explanations for the 1990s crime drop — changing drug use patterns, decreased gun violence, New York-style policing, the federal COPS program, and increased incarceration. Notably absent in all of this, though, is the relationship between mental health and prison populations.

With the exception of research that specifically explores the interdependence of the mental hospital and prison populations, including some public health studies and some empirical research into the causes of the prison explosion (for instance here, here and here) published empirical research does not conceptualize the level of confinement in society through the lens of aggregated institutionalization (asylum + prison) but rather simply through imprisonment rates.

Even the most rigorous, recent analyses of the prison-crime relationship use only imprisonment data. Though a tremendous amount of empirical work has been done on long-term crime trends, structural covariates of homicide, unemployment, and the prison expansion, none of this literature conceptualizes confinement through the larger prism of institutionalization, and none of it aggregates mental hospitalization data with prison rates.

So in contrast to Andrew Gelman, I'm not only surprised by the results of the regression, I'm also extremely concerned about the implications regarding the state of our current knowledge and existing research. And, unhappily, in contrast to Gelman's just-so post, I expect a huge amount of resistance.

A classic example of confusing correlation with cause. The Royal Road to truth is the controlled experiment. When you can't do those, you had best be modest in your claims about "proof". Global Warming zealots please note.
5.2.2007 8:40am
Zubon (www):
I would be interested in your reasons for including execution rate as an independent variable contributing to homicide rates as a dependent variable. It seems intuitive to me that states with more homicides would have more executions; there are not many other crimes for which you execute people. One could also argue that "brutality" in the populace is a separate variable that leads to both homicides and policies of capital punishment. What is your reasoning behind that direction of causality? Is the effect of controlling for aggregated institutionalization what makes it compelling to you?
5.2.2007 9:04am
Bruce Hayden (mail) (www):
I find the thesis of your studies quite plausible and am just surprised that it hadn't been pursued before. The problem to some extent is that whether someone is in prison or in some other institution is somewhat arbitrary/ fungible/ transparent. Not quite sure of the right word here. But the idea is that at one point, we institutionalized a lot of people who had trouble dealing with society. But that was deemed inhumane, so many there were released, just to ultimately turn back up in prison. And, country to country, even today, there seems to be a significant difference in this mix.

I do think that if your findings hold, that maybe we need to rethink that policy of minimizing those otherwise institutionalized, if for no other reason than mental healty facilities may be a better fit for some of those who ultimately end up in prison. But from a civil libertarian point of view, I don't want it made too easy to institutionalize people.

Luckily, I don't have to make these decisions.
5.2.2007 9:35am
MikeM (mail):
I'm concerned about the use of crime rates. A grad student of mine, for example, wanted to understand why, as the population dropped in Hammond IN over some decades, the crime count increased. I suggested that it might have been (1) the loss of jobs that led to lower tax base that led to smaller police force, or (2) the loss of the middle class that (a la WJ Wilson) had a moderating influence on behavior. In either case, one would have to look at the facts on the ground to determine what actually occurred.

Similarly, the crime rate in a city with an increasing population has its own dynamic: fields may be converted to housing tracts and annexed, and these tracts usually have low crime counts, leading to a lower crime rate.

In other words, one shouldn't assume that the ratio crime/population is meaningful by itself.
5.2.2007 9:54am
cirby (mail):
I'm just a bit confused.

Why are some people so surprised that crazy people commit crimes?

It's not exactly a hard stretch.

A number of years ago, an FBI agent put it a different way: about one percent of murders in the US are committed by a consistent population of about one hundred mentally unbalanced people - who seldom get caught.

They're called "serial killers."
5.2.2007 10:34am
TDPerkins (mail):
It seems intuitive to me that states with more homicides would have more executions; there are not many other crimes for which you execute people.

The time period investigated includes periods where the death penalty was variably prohibited nationally, by state, and when permitted pursued with varying vigor.

That has to be controlled for.

Yours, TDP, ml, msl, &pfpp
5.2.2007 10:44am
Steven Vickers:

The Royal Road to truth is the controlled experiment. When you can't do those, you had best be modest in your claims about "proof".

1) When you can't do controlled experiments, that doesn't mean we should throw up our hands and say "nothing to say here." The entire social sciences, along with a good chunk of natural sciences, would evaporate in that case.

2) Who exactly said this was "proof" of anything? The words "proof" or "prove" aren't in the post that I can see, and Harcourt has been relatively modest in his claims.
5.2.2007 10:49am
TJIT (mail):
Bernard Harcourt,

Interesting post, I enjoyed reading it
5.2.2007 11:28am
John_Pfaff (mail):
I think one fact that makes Bernard's results so interesting and perhaps surprising is the population differences (which he noted on Monday): the mental hospital population was older, more white, and more female than the prison population is today. Yet I believe that those who make up the perpetrators of homicide remain relatively constant over the sample--young men (but that is a guess). So, simplifying crudely, locking up younger black men or older white women seems to influence homicide rates the same.

This disconnect could suggest that the correlation does not reflect a true underlying causal connection. But if the correlation is causal, then I agree with Bernard that the results tell a surprising story. What would be the causal mechanism that would result in such different commitment strategies exerting a common influence? It is a fascinating puzzle.

Relatedly, I would be interested to know how the pools of offenders and victims in the high-hospitalization period compare to those in the high-incarceration period. Since most crime is intraracial, for example, are the offenders and victims relatively more black in the high-hospitalization period--since relatively more whites are committed--and vice versa in the high-incarceration period (controlling for factors such as crack in the 1980s)?
5.2.2007 11:52am
JohnK (mail):
Criby is right when he wonders above how it is that people can be surprised that crazy people commit crimes. Prison reform advocates for years have argued that many inmates in prison, especially problem inmates we are now housing in very inhumane "Supermax" prisons, are in fact mentally ill and in need of treatment. If a significant portion of the prison population are in fact mentally ill, it follows that deinstitutionalizing large numbers of the mentally ill in the 1960s would have increased the crime rate.

Bernard is beating around the bush looking at general statistics. It is an interesting and intuitively correct correlation. That said, I would like to see real studies on de-institutionalized mental patients. Of the ones who were released in the 1960s, how many of them ended up committing crimes? Of the people in prison today, how many of them would be classified as mentally ill to such an extent warranting commitment had they not ended up in prison? Answer those two questions and you might have a real case that deinstitutionalization resulted in higher crime rates. Until then, it is just an interesting graph.
5.2.2007 12:20pm
Bernard E. Harcourt (mail) (www):
In response to Zubon here, the reason I originally included the execution rate in my analyses is that some researchers argue that the number or rate of executions in a state has a deterrent effect on homicides (see generally here versus here). As a result, it seemed important to take account of the possible effect of executions.

You are right, though, that executions and homicides may be related in more than one way: increased executions may deter (reduce) homicides, but increased homicides may promt (increase) executions. John Donohue and Justin Wolfers address this question at length in their review of the death penalty deterrence literature here.

This may have a confounding effect on my results on executions, but my focus was on institutionalization. Note that in the institutionalization context, the dual relationship between institutionalization and homicide works in the opposite direction: more homicides should cause more incarcerations, but more institutionalization should cause less homicide. The possible bias here works to understate my findings. In other words, the negative effect of institutionalization on homicides is probably even bigger than what I report (because of the confounding effect of incarceration on homicides).
5.2.2007 12:37pm
Bernard E. Harcourt (mail) (www):
John Pfaff raises one of the central questions in his comment here: what about the race, gender and age differences of the two populations? I've been meaning to get back to this all week -- and hope to elaborate on it in a post. No doubt, there were important differences as I discuss in my paper here at pages 1781 to 1784.

Not only were there difference as between the two groups (asylums and prisons), there were also shifts over time within the groups. So, for instance, Henry Steadman and John Monahan show in a study published in 1984 that in their mental hospitalization sample, the proportion of non-whites increased from 18.3% in 1968 to 31.7% in 1978: "Across the six states studied, the mean age at hospital admission decreased from 39.1 in 1968 to 33.3 by 1978. The percentage of whites among admitted patients also decreased, from 81.7% in 1968 to 68.3% in 1978."

There are important difference and shifts over time in the institutionalized and incarcerated populations. Naturally, as John Pfaff suggests, we need to explore whether there were corresponding shifts -- one way or the other -- among the victims of crime and among the perpetrators.

Once we have nailed all that down, we then need to think about and test possible causal hypotheses. So there is indeed lots of work to do. It's the topic of my next paper in this series and I hope to discuss it more on this blog soon.
5.2.2007 12:57pm
Shannon Love (mail) (www):
Bernard Harcourt,

Perhaps what your study actually tracks are the broad social, political and legal changes in the way society views non-conforming behavior in general? In eras when non-conforming behavior is tolerated there is less political support for all forms of incarceration. In eras when non-conforming behavior is frowned on there is more political support for incarceration.

Both criminals and the mentally ill do not conform, albeit from wholly different reasons. The mechanisms of dealing with both populations share strong legal and political similarities even though there is relatively little overlap between the two. A political consensus on how to deal with one issue would be the same consensus used to deal with the other.

This would explain why two populations seem statistically linked when very little about them i.e. age, sex, ethnicity, income etc overlap.
5.2.2007 1:59pm
JohnK wrote:

That said, I would like to see real studies on de-institutionalized mental patients. Of the ones who were released in the 1960s, how many of them ended up committing crimes?

I agree. It's hard to see how we would know that prior analysis was wrong without such a study.

I'll add one more confusing factor: the insanity defense to a murder charge. In any state sponsored mental institution, we can find people who simply aren't mentally ill. They never were. Most of them got there by pleading insanity to a murder charge. After they gain such privilege, they tend to hang out with each other, because the company of sane people is almost always more rewarding than that of the insane (Cuckoo's Nest not withstanding).

The institution authorities are presented with a quandary. They're supposed to hold the patient until the patient recovers, but the patient was never ill. They make such authorities nervous, because, ironically, they're much harder to control than the truly ill. They don't take the controlling drugs.

And so, the institution kicks them out "before their time", but the real problem is, they belonged in prison, and now they go on to commit new murders. I suggest that this class of people were largely unaffected by the sweeping policy changes which occurred around 1970, although there are far fewer such people in mental institutions today than there were in 1970. The rate of sane people "getting off" by pleading insanity has been significantly lowered due to a raised legal bar.

So any study showing new murders by old mental patients needs to control for the raising of the insanity defense bar which occurred over 1970..1990. To do this would require categorizing some mental patients as "beat murder charge, belonged in prison the first time". You simply can't count such people as among the violent which were released by the policy changes of the early 70s.
5.2.2007 1:59pm
Kent Scheidegger (mail) (www):
Responses to the Donohue and Wolfers article are available at the links below. The impact of deinstitutionalization does require further investigation. I have not seen it considered in the death penalty deterrence literature to date.

Mocan &Gittings

5.2.2007 2:05pm
I don't see how the execution rate (ER) in the last seven decades can be used to study anything.

The ER is totally a construct of public opinion - though statute - and constitutional law. Indeed the SCUSA invalidated all capital punishment laws at one point and it was about fifteen years before executions resumed under new statutes.

And, excluding the laws themselves, public opinion also changes and makes juries less willing to impose a death sentence at times.

I doubt there is much to these correlations. But Shannon Love made an interesting observation when she said:

"Perhaps what your study actually tracks are the broad social, political and legal changes in the way society views non-conforming behavior in general?"
5.2.2007 3:18pm
M. Simon (mail) (www):
I'd like to know how much incarcerating pot smokers reduces crime.

Perhaps one reason the crime drops were not as big as expected is that we are locking up people who do not committ crimes against other persons.
5.2.2007 3:23pm
M. Simon (mail) (www):
We know that (roughly) 80% of all crimes are done by 20% of all criminals.

We could probably get a greater than 12% decline in crime if we were locking up the right people.
5.2.2007 3:26pm
M. Simon (mail) (www):
The decline of crack explanation implies that crime is fueled in part by drug prohibition.

I wonder if the UC guys ever heard of alcohol prohibition and its crime inducing properties.

I note that there is no check on how the War On Drugs has affected crime.

We saw this in my town. The FBI predicted an increase in the murder rate locally after a big gang bust. They were right.

We haven't seen a similar bust in almost 20 years. Perhaps the FBI and others have reduced their enforcement efforts since they know that increased enforcement of prohibition laws increases the murder rate.

So why does the FBI know something that smart UChicago (my alma mater) guys do not?
5.2.2007 3:36pm
Paul A'Barge (mail):
blah blah blah ... put the nut jobs in the mental pokey and let's all move on.
5.2.2007 3:44pm
M. Simon (mail) (www):
And then the brilliant UC guys say we need to look at the "crack baby" thing.

Except Dr. Ira Chasnoff, who invented the "crack baby" syndrome, says he was in error.

You know I'm beginning to believe that these brilliant UC guys don't know what they are talking about.

Crack babies - not all they are cracked up to be.
5.2.2007 3:46pm
M. Simon (mail) (www):
BTW the "crack baby" bit was debunked in 1992.

Not exactly news.
5.2.2007 3:59pm
M. Simon (mail) (www):
OK so now it is not crack babies (the brilliant UC guys admit somewhat). It is crack neighborhoods.

Perhaps turning some neighborhoods into war zones adversely affects mental health AND crime.

Perhaps "crack neighborhoods" are just a proxy for war zones.

Did I mention alcohol prohibition?
5.2.2007 4:12pm
Bernard E. Harcourt (mail) (www):
A reader asks "how much [does] incarcerating pot smokers reduce crime" here. That's a great question. Surprisingly, there is some new empirical evidence on the question. It's a paper I just finished with Jens Ludwig called Reefer Madness and it looks at data from New York City. In the 1990s, the NYPD dramatically increased arrests for smoking marijuana in public view -- up from about 2,000 arrests per year to over 50,000. The effect on crime? We don't see it.
5.2.2007 5:19pm
Kent Scheidegger (mail) (www):
Arrests do not equal incarceration, at least not for any substantial length of time. How many people are actually in jail for possession for personal use? Very few, I suspect.
5.2.2007 5:49pm