Tuesday, June 12, 2012

Hypothesizing reasons for continued crime declines

Reported crime including both violent and property crime continued its remarkable decline in the first half of 2011, reported the FBI:
Preliminary figures indicate that, as a whole, law enforcement agencies throughout the Nation reported a decrease of 6.4 percent in the number of violent crimes brought to their attention for the first 6 months of 2011 when compared with figures reported for the same time in 2010. The violent crime category includes murder, forcible rape, robbery, and aggravated assault. The number of property crimes in the United States from January to June of 2011 decreased 3.7 percent when compared with data from the same time period in 2010. Property crimes include burglary, larceny-theft, and motor vehicle theft. 
Here's a table including Texas cities over 100,000 people comparing reported crime rates for the first six months of 2010 and 2011. Notably, a few towns where murders had virtually bottomed out saw significant year-to-year increases, including Amarillo (3-8), Arlington (6-12), Beaumont (3-7), and El Paso (1-14). But overall murders declined because of offsetting reductions in other jurisdictions, particularly Houston (143-90) and Fort Worth (32-18), along with more modest declines in most other places.

Most Texas jurisdictions saw declines in reported violent and property crime in the first six months of 2011 compared to 2010. Despite reduced federal grant funds. Despite a faltering economy and rising wealth disparity. Despite increased gun ownership. Despite lofty drop-out rates in high schools. Despite depopulating Texas youth prisons, reducing their inmate numbers from 5,000 to 1,100 since 2007. Despite Texas releasing more than 70,000 adult inmates per year from prison back into their home communities. All that's going on and crime is still declining. So what's causing it?

Grits believes the answer lies in a confluence of several possible reasons. First, it must be acknowledged, the best econometric estimates (Spelman, Levitt) say around one-quarter to a third of crime reduction in the '90s came from locking up more criminals for longer stretches, which makes sense given the vast scope of expanded incapacitation. But those same models showed incarceration levels had long since passed the point of maximizing marginal utility (in the economists' jargon). Based purely on a cost-benefit equation, wrote Spelman, "Enormous cutbacks - reductions of 50% or more in the prison population - are not difficult to justify and would probably save the US public billions of dollars each year. Certainly there is little economic justification for continuing to build."

So those who argue most strongly for the effectiveness of incarceration at reducing crime say we've already gone to that well too often. And anyway, in Texas crime continued to decline even after we stopped building new prisons. Incarcerating ever-more people can't be the only factor. So the question remains, if expanded imprisonment accounted for a quarter to a third of the '90s crime decline, what accounted for the rest, and why does crime keep falling now that incarceration (even in Texas) has leveled off?

Our friends at the Texas Legislature might suggest that new probation and parole programming effectively reduced incarceration levels without harming public safety, and because I supported them, I'd love to give the 2007 reforms credit. But the trend is national, not Texas-specific, so even if that had a positive effect, it can't explain it all.

My own favorite theory: Young people spend a great deal of time engaged with technology like the internet, video games and cell phones that didn't exist 25 years ago. These activities occupy time of teens and young adults who are the most likely to commit crimes. The kid perfecting his skills at Grand Theft Auto V may not be preparing himself for the job market, but he isn't out stealing my car.

On property crimes, there's an hypothesis that the rise of cheap, mass-market goods has contributed to reducing theft because lots of things we own have relatively low resale value and simply aren't worth stealing anymore. The list of items most often stolen bears out that trend, consisting mainly of things that aren't cheaply produced or where brand value trumps utility among callow minds.

Some theorize that improved pharmacology reduced crime, both thanks to improved antipsychotic and anti-depression drugs for adults and potentially because of earlier diagnosis and more frequent medication of mental illness among juveniles. At a gut level, Grits remains skeptical of that explanation, but there's inarguably a time correlation.

James Q. Wilson has cited research to argue that "reduction in gasoline lead produced more than half of the decline in violent crime during the nineties" because of lower lead-levels in the blood.

Professor Steven Levitt, co-author of the book Freakonomics, has provocatively argued (to paraphrase) that legalization of abortion and expanded access to family planning services during the 1970s reduced crime in the 1990s by enhancing women's control over their reproductive cycle and thus reducing the number of unplanned, unwanted children coming of age in that period.

Our friends at the NRA might suggest that the expansion of "concealed carry" laws to three dozen states contributed to deterrence, though that explanation fails to explain parallel declines in the other states and Canada, and Grits doesn't believe the practice is widespread enough to explain the observed crime decline.

Similarly, for Texas and other states with significant Latin American immigration, legal and otherwise, immigrants tend to commit crimes at much lower rates than US citizens and their out-sized presence could explain lower crime rates. But just as with concealed carry laws, that explanation doesn't account for (larger) crime drops in New York and other parts of the country that didn't see the same immigration wave. (Notice the theme: For several of these it's difficult if not impossible to link correlation and causation.)

In Texas (Grits can't speak for elsewhere), one seldom discussed factor may be the evolving makeup, culture and revenue streams of criminal gangs. Historically, conventional wisdom held that criminal street gangs primarily sought to nail down turf in their hometown for retail drug sales and sometimes extortion, tagging neighborhoods to mark their territory, etc.. But in Texas today the most important gangs are prison-affiliated organizations, some of which have become arms of transnational drug smuggling cartels out of Mexico, sometimes even working as paid assassins south of the border. Their focus is less on controlling a piece of turf in a Texas town so much as maintaining a transit corridor through which drugs may be smuggled to the rest of the country. They're still dangerous criminals but the focus of their business (and often, their violence) has changed along with the sources of prohibition-driven profits.

Finally, there are usually a few, typically anonymous zealots/trolls who show up whenever FBI crime data are reported to claim that they far understate actual crime, much of which, the argument goes, remains forever unreported. (Often this absurd construction attributes unreported crime to the success of the "stop snitching" meme.) But Grits believes those arguments are debunked by essentially similar trends demonstrated in national crime victimization surveys, which also show crime declining. I'm willing to believe numbers are juked in some jurisdictions, but that's something that must be demonstrated, not merely alleged.

Why do you believe crime is declining? Are the reasons related to things the justice system is or isn't doing, or something else entirely? And even if the answer is "something else entirely," is that something that government can double down on to reduce crime, or simply a happy, random accident that cannot be intentionally replicated? Your guess is as good as mine, so what's yours?

H/T: Sentencing Law & Policy

24 comments:

Arce said...

Sociological criminologists in the 1970s and '80s predicted a drop in crime when the echo-boom, the children of baby boomers, began to age out of the prime crime committing years (late teens and early 20s. Many echo-boomers are now in their 30s and some in their 40s.

So, perhaps part of the explanation is that the number of those likely to commit crime is down b/c of population aging, incarceration of some, and additional understanding of the factors that lead to crime.

Anonymous said...

"Reported crime" has declined, or is it "the reporting of crime" that has declined? What happens to those who report? As they say, "keep your mouth shut or else."

doran said...

Grits,

What is going on with the street prices of illegal drugs, particularly heroin, cocaine and meth? The prevailing dogma years ago was that those addicted to such were the main perpetrators of theft of consumer goods, stealing to raise cash to feed their addictions. If the street prices of these drugs are down, then that might have some influence on the theft stats.

On the other hand, if the street prices are not down, then....never mind.

Anonymous said...

Perhaps the increased incarceration rates also caused a deterrent effect beyond those violent offenders who are actually incarcerated. In other words, maybe society in Texas has finally gotten the message that Texas is tough on crime.

Don said...

anon: 8:21 Incarceration and threat of punishment as a deterrent only if you see the criminal as a totally rational being. That's usually not the case.

Don said...

Should have read "only makes sense".

mike connelly said...

You can add to your list of possible factors the new work correlating: attitudes about institutions and about consumer confidence, the increase in private security forces and police, immigration (first and second generation immigrants reduce crime in the areas they settle), and the reality that much of the violence of the period eliminated violent people. What's always missing from all these analyses, including the ones you mention, is the fact that probation numbers paralleled imprisonment numbers in their rise in the period, yet no one has ever attributed the changes in crime rates to the impact of probation and community sanctions. Also, those analyses ignore the point Zimring made a while back that Canada demonstrated similar crime patterns without ANY of the policy changes the US made, prison, law enforcement, rain dances, anything. Finally, here's this. Those models "proving" a 25%-33% impact of prison DID NOT include the factors that we're talking about here. Unless/until they are run with similar probation numbers and the factors detailed, their reliability is on the order of an urban legend, "conventional wisdom" that is all of the first term and none of the second. But thanks, as usual, for the thoroughness and thoughtfulness of the post. Still a rarity in these days of commentary.

Brad said...

I tend to agree with Grits' favorite theory: technology and the cheapness of mass market goods. Young people who may be more predisposed to deviance, or to committing violent crimes like assault, battery, rape and other forms of sexual assault now have alternative means to channel their potential for crime and deviance. I don't think it is a stretch at all to contend that they can achieve cathartic release through either violent video games, movies, TV shows, or the plethora of niche pornography available on the internet, rather than actually acting on those impulses. Plus, I think it was the Heritage Foundation that published a report saying that the vast majority of households considered under the poverty level today have things like refrigerators, microwaves, TV's, computers, DVD players, etc. While the report was trying to say the poor aren't really poor (put that aside), I'd posit that the cheapness of these goods today and the likely fact that even the lowest of the low-level jobs can allow someone to afford these goods, makes the risk of arrest and imprisonment far too great when even the most low-skilled workers and crappiest of jobs can enable people to afford these goods.

Anonymous said...

the data submitted to the feds by politically minded law enforcement ageny heads are being modified. face it - all law enforcement heads are politicians, who in effort to get reelected need to show that crime is down. plus, the police today even load up all types of extra charges on defendants, yet statistical analysis reflect that crime is down. these reports smell.

Anonymous said...

Keep in mind it is "reported crime." There may be high levels of apathy that result in rising unreported crime rate.
It could also be the simple fact that jail bed space, court appointed attorney's and other high costs just can't be supported by the tax payer.
And crime may be going high tech and law enforcement and prosecutors haven't caught up with criminal technology.

Paul-UK said...

I found this report on the Huffington Post UK which the Royal Statistical Society and the American Statistical Association into the average takings in the UK are about 20,000 pounds, in the US the average takings during a bank robbery is $4330.

http://www.huffingtonpost.co.uk/2012/06/12/robbing-banks-crime-doesnt-pay-economists_n_1589493.html?utm_hp_ref=uk

Anonymous said...

I'm behind the video game/internet reasoning.

One trend I feel like I'm seeing is also the trend of drug violence happening outside of public view. I think the drug market has realized that customers feeling safe buying drugs is better for business so the actual crime frequently happens outside the cities or away from the sales.

Gritsforbreakfast said...

To those (7:50, 10:52, 1:34) saying more crime is going unreported, please re-read the next to last paragraph in this post. If what you say is true, crime victimization surveys would not confirm the downward crime trend, but they do. There's no evidence to back up what you're saying, and fairly strong evidence it's BS.

Gritsforbreakfast said...

BTW, Mike Connelly, thanks as always for your excellent comments. Great point about those models failing to account for changes in probation. I know that adult probation has a lot more research-based tactics at its disposal than 20-30 years ago that a lot of Texas departments have embraced, and you're right I've never seen anyone attempt to quantify the effects of those changes.

Anonymous said...

What percentage of victims are too afraid to report the person who assaulted or raped them? How about those who live in a neighborhood where violent crime is tolerated and where it takes a brave person to "run their mouth." We don't want to believe in victim intimidation, but it's there.

Tim said...

Thanks for this excellent blog, which I just came across while reading Drucker's 'A Plague of Prisons'.

I wonder whether another reason for the downturn in crime, both in the official figures and in victim report studies, may not be that prisons have become, for some inmates, the 'business schools' of crime, where criminals learn to network more efficiently, negotiate with less need for open violence, and to switch activities to less visible - and often more lucrative - forms of illegal money-making? These kinds of crime will be less likely to show up in either the official or the self-report studies, as victimization is more diffuse, less individualized, and less dramatic. By this reading, imprisonment does not necessarily reduce crime, but is instrumental in changing its nature.

Gritsforbreakfast said...

Thanks Tim.

And to 7:38 (and 7:50, 10:52, and 1:34 - all likely the same person), I've already answered your question in the next to last paragraph of this post, which you simply ignore. So let's restate it here. Please respond directly:

"Finally, there are usually a few, typically anonymous zealots/trolls who show up whenever FBI crime data are reported to claim that they far understate actual crime, much of which, the argument goes, remains forever unreported. (Often this absurd construction attributes unreported crime to the success of the "stop snitching" meme.) But Grits believes those arguments are debunked by essentially similar trends demonstrated in national crime victimization surveys, which also show crime declining. I'm willing to believe numbers are juked in some jurisdictions, but that's something that must be demonstrated, not merely alleged."

If victimization surveys showed high crime rates and reported crime dropped, we might surmise (as you do, sans evidence) that more crime is going unreported. But overall crime victimization has gone down in tandem with reported crime and all the evidence says your (always anonymous, repetitive, cite-free) theory is just wrong. There's little wonder, certainly, why you don't attach your name to it.

A Critic said...

There are surely many factors - what about CSI and other TV shows? Most people have the impression that those shows are accurate - perhaps the idiots likely to commit crimes are sometimes dissuaded because they believe they'll be caught via forensic technology?

Anonymous said...

How can anyone say that tough on crime doesn't work? Maybe tougher sentences we have been handing out do work. It just takes a little time to see the results trickle down.

Anonymous said...

Arce mentions sociologists and a theory why the crime rate might have fallen. In the 1990s, there was an expectation that a crop of super-predators would arise that would cause an upward spike in crime, that these were the offspring of parents/grandparents who did not discipline these offspring, so that they became functionally feral. The point may say more about the difficulty in coming up with a hypothesis that does account for any change in the crime rate. For certain, though, trends that don't fit at least make for a living discussion in any criminology course...

Anonymous said...

Grits: "But Grits believes those arguments are debunked by essentially similar trends demonstrated in national victimization surveys."

Those arguments can be demonstrated to be bogus by their lopsided nature.

Should crime go up, Anonymouse would never claim that citizens' reporting frequency just rose. He'd claim it was crime itself that was up. But should it go down, citizens are clamming up alofa sudden.

My idea: is arson going down becuz the gov less frequently uses pseudoscience to classify a fire as "arson"? Auto theft, because of gadgets to prevent theft?

~L~

Anonymous said...

I've wondered if there might be a sort of snowball effect at work - let's assume that human behavior is mostly learned, and thus that some amount of criminal (or especially violent) behavior is learned. As criminal behavior declines, fewer (young) people will see crimes being committed around them and "learn" to act that way. So whatever the original reasons for declines in crime, be they demographic or technological or chemical or what-have-you, such declines might have a tendency to become self-perpetuating or self-multiplying beyond the magnitude we might expect from the original cause. Just a theory!

Anonymous said...

Why isn't anything I posted showing up this week?

I like the ageing population theory, the video-game effect, the CSI effect, and A Critic's snowball effect.

SSRI drugs would increase crime. Lead was removed from gasoline but children have more vaccines.
Maybe the public is just too sick to commit violent acts!

My suggestion was that technology has advanced enough to prevent car theft, and arson is about the same but we used to cry "arson!" after an accidental fire.

~L~

M Johnson said...

I am involved in intergenerational crime/delinquency research and some preliminary findings based on national data examining delinquency among adolescents in the late 1970s (age 11-17 at beginning of the study), AND their offspring of the same age in the early 21st century show substantial drops in traditional forms of crime from older to younger generations. There is some indirect evidence that technologically-oriented normative AND criminal behaviors are playing a major role in these shifts. However, preliminary information shows few meaningful generational differences in "causes" of delinquency.

If you accurately include technology-aided crime in statistics, the rates would be much higher overall, though violent crime would still be low.