I was discussing Kevin Drum’s post on falling crime rates in America and the old Freakonomics argument came up – that the only possible explanation for this phenomenon is the after-effects of Roe v. Wade. The fewer unwanted pregnancies, so the argument goes, the fewer potential criminals.
Interestingly, alongside falling crime rates, we also see lower teen pregnancy rates, lower rates of teen consumption of alcohol and drugs, lower violence in school, and a host of other “oh I guess the world isn’t ending” figures when we actually look at the data available.
However, whereas the abortion rate has leveled out, crime rates continue to go down. This may poke a hole in the Freakonomics argument. Then again, maybe that argument wasn’t all that great to begin with – though it was certainly controversial and contrarian.
So two things here:
First of all, there is probably no one reason that crime rates have fallen. This is a very complex issue, and I don’t think anybody can realistically say they’ve controlled for every other factor and voila! here is the answer to why crime is falling.
Second – what about video games? This thought just struck me today – but looking at these charts, crime has been steadily decreasing since 1990 – about the time when video games were going mainstream and video game consoles were becoming affordable for most Americans. Video game sales have increased steadily during this time period as well. You could draw a line in the very opposite direction as the lines in these charts, for instance, charting video game sales between 1990 and 2008:
If anything, the fact that crime has been dropping for the past twenty years (along with things like teen pregnancy, etc. as I noted above) while more and more young people consume more and more video games should put a lie to the notion that video games actually increase crime and violence. I did a little googling and found this paper by Adam Thierer [pdf] which doesn’t exactly support the idea that more video game consumption has directly contributed to less crime, but certainly suggests that it’s a possibility.
One really fascinating bit from the paper is this excerpt from Killing Monsters: Why Children Need Fantasy, Super Heroes, and Make-Believe by Gerald Jones:
One of the functions of stories and games is to help children rehearse for what they’ll be in later life. Anthropologists and psychologists who study play, however, have shown that there are many other functions as well—one of which is to enable children to pretend to be just what they know they’ll never be. Exploring, in a safe and controlled context, what is impossible or too dangerous or forbidden to them is a crucial tool in accepting the limits of reality. Playing with rage is a valuable way to reduce its power. Being evil and destructive in imagination is a vital compensation for the wildness we all have to surrender on our way to being good people.
I think reading children’s literature is extremely important in the development of empathy in young people. You read about kids who are outsiders, who get picked on, who have to use their wits to get out of bad situations, and you learn a little bit more about what it means to be a human being. You learn what other people in your class might be going through – though with fewer wizards and dragons involved. Maybe playing video games or getting into these role-playing situations where you can be the villain, the monster, the criminal, or even the hero, the special-ops troop, and so forth is an important way to develop another kind of empathy – an empathy with the person we could be or would like to be, or at least to explore that part of ourselves that we will never become – maybe so that we never become it.
Or maybe video games are just very entertaining and relatively cheap – and fairly addictive. Maybe they’re just a good alternative to many worse things we could be doing with our time.
In any case, I think the idea that video games have led to a decrease in crime rates is at least as compelling as the abortion argument. Indeed, I think it’s quite a lot more compelling. But I think the question remains too complicated to boil down to any one neat and tidy explanation.
MikeJ
Global warming has increased since 1990 too. Perhaps it’s too hot to commit crimes.
LarsThorwald
I know why crime is down.
(looks around nervously)
The…the Bat Man.
DougJ
I think a lot of it is the demise of crack. Crime is down especially in urban areas where crack was previously a big problem and it’s not down so much in rural areas, where meth is now a problem.
Jon H
Cable TV and video games (and the internet).
There’s more to veg out on than there was in the 3-bland-network days, so people are more likely to stay home than be out on the streets trying to find something to pass the time.
scav
Disappearance of Bees? Decline in number of Pirates? Or have Criminal Masterminds been going extinct because of the loss of the rain forest?
DougJ
And don’t forget the way the nation came together after 9/11. That had to bring crime down a lot.
slag
Personally, I think crime rates have fallen because you can no longer buy Sudafed off the shelf. You don’t know what mass quantities of that stuff was doing to people.
ETA Although I now see DougJ begs to differ. I smell a book offer.
scav
@DougJ: yes, it must be clearly inversely related to the number of visible flags on cars and firetrucks.
MikeJ
Atheism is also trending upwards while crime trends down.
Mark S.
Why do criminologists love to pick just one thing for why crime goes up or down? I didn’t know the Freakonomics idiots settled on abortion, but I’ve also read lead paint and our insane rate of incarceration as the only reason crime has gone down.
Zifnab
@DougJ:
/reaches for WoW again
/stops
/shudders
/slowly reaches for WoW icon again
Yes, thank god we’ve gotten rid of the crack.
Roger Moore
@scav:
In case you haven’t been following the news, pirates are back in a big way- largely because of Bush’s failed policies in Somalia. And I’m sure that RIAA and MPAA will tell you that piracy is rampant these days.
scav
@Roger Moore: oh no! The FSM is in danger! Aux barricades!
joe from Lowell
I think the horrible urban renewal and urban highway projects of the 1950s-1970s played a bigger role in the rise in crime than is generally acknowledged, and the more humane urban policies begun under Jack Kemp and, especially, under Bill Clinton played a significant role in the decline.
The Hope VI Program did this country a lot of good.
Zifnab
Well, the abortion argument really boils down to the youth poverty argument. If women have abortions to prevent raising children they can’t afford, the question becomes “Do we still have a large youth population that parents can’t afford?” If the answer is “Yes” then the abortion argument goes out the window. If “No” then abortion – or perhaps simply a decrease in child poverty through some host of other measures – may have played a roll.
@Mark S.: Because people want to treat crime like some kind of bacterial infection for which a single penicillin cure will cure us all of.
All that said, I would venture another possibility. Perhaps crime hasn’t actually gone down at all. Perhaps, instead, the crime has gone from blue collar armed robberies and muggings to much more white collar PayPal account hacks and email scams. Perhaps crime has evolved from street thuggery to a more sophisticated, less violent practice. And because it doesn’t involve a fist or a knife or a gun, it goes increasingly unreported.
Mike from DC
I read an article a couple of years ago that found a correlation between the phase-out of leaded gasoline and the drop in crime. The theory goes that the brains of children born after TEL was banned from gasoline developed differently, and were less likely to be violent. Of course, this has the caveat correlation != causation.
http://www.nytimes.com/2007/10/21/magazine/21wwln-idealab-t.html?_r=1&adxnnl=1&adxnnlx=1193069109-tIv/I01qmqYqqX/fw3A7Iw
http://www.washingtonpost.com/wp-dyn/content/article/2007/07/07/AR2007070701073.html
El Cid
I’ve seen arguments before that when unemployment and poverty is more generally and equally shared throughout a large community (i.e., a nation), there may be lower rates of crime than when poverty and unemployment are (a) much more geographically concentrated and (b) easily seen as a drastic contrast with the norm. Sort of a social coherence argument.
Cackalacka
The abortion argument may still pan out. An unwanted child can still produce unwanted children.
slag
And can I just say…Look how many problems legalizing abortion has solved! Thank you, Planned Parenthood!
djork
Demise of crack? Take a stroll with me to the grocery store sometime. In that 5 block walk, someone will either try to sell us crack or will try to bum money for crack. And I live in a decent neighborhood, though rough at the edges, obviously.
AdamK
It’s all da internet pr0n. Bad guys home alone gettin bizzy.
HobbesAI
Expect violent crime to begin trending up again. (The sunspot cycle is currently at its minimum.)
DougJ
@djork:
I think the use of it really is down overall, though.
scav
Availability of easy credit got any takers? Though I still prefer pirates. Or maybe the popularity of pomegranate juice.
Roger Moore
@Zifnab:
I doubt it. That implies that there’s some kind of move of criminals from one kind of crime to another, but that doesn’t seem likely. Computer crime requires a completely different skill set, and even a different mindset, from violent crime, so people don’t just drift from one to the other.
Wile E. Quixote
Do you want to know why crime has decreased over the last twenty years? It’s because of the Reagan tax cuts and defense build up of the 1980s. Also the fact that crime is down proves that global warming is a myth and that the earth is actually entering a cooling period. If it were warmer criminals would be out on the streets committing crimes but since it’s so cold they’re choosing to stay at home instead.
OK, can I have a job at Reason magazine now?
Zifnab
@Roger Moore: Well, this would be more of a generational shift. I’m not saying a guy who moonlights as a club bouncer is going to drop his baseball bat and take up credit card fraud. However, I am suggesting that perhaps credit card fraudsters are growing in number while guys with baseball bats hiding in alleyways are decreasing.
Tim in SF
Prove it.
What I have read tells me empathy is a biological trait — something we share with other social primates. It is hard wired.
Here, give THIS a listen:
http://www.radiolab.org/2007/aug/13/
WereBear
I think the video game angle is not so far fetched.
I base this on having lots of rural relatives, now aged, who grew up when there was nothing to do. The kind of bone deep boredom where you would do something life threateningly stupid just to feel something.
I see the effects it had on them; none of them good, by the way. I mean, I grew up rural, with only my brothers for company and playmates, and we used our imaginations a great deal. But we also had books and television; very different from when a lot of people did the same three tasks from dawn to dusk and then went to bed.
This has an urban angle, too; there are more amusements in a city, obviously, but a lot of them weren’t good for you, either.
djork
@DougJ:
In this house….er, neighborhood they ain’t!
But you’re right.
Martin
Well, let’s look at some of those crimes. Vehicular theft is going down rapidly – less than half of what it was 20 years ago, but technology has made vehicular theft much more difficult just to initiate and WAY more risky due to Lo-Jack and other services.
Similar but less dramatic improvements to technology for home security now as well, as well as for portable electronics. Laptops and phones now have features that track the device when stolen.
For violent crimes, the proliferation of cameras makes these crimes less and less likely to be anonymous, and the expansion of electronic media makes any crime data easier to widely distribute. Truth is, the world is getting smaller at a very rapid rate, and that’s a bad scenario for criminals.
I think there’s a whole host of incremental and fairly unremarkable reasons that taken together explain why the crime rate is dropping. I think the video game explanation is kind of silly. We had affordable video games in the first wave of consoles back in the 70s/80s. More likely is that the 90s brought cheaper computers and wider access to the internet giving criminals a new place to operate – the ability to steal from people with essentially zero risk of violence.
Pangloss
Cars are more difficult to steal than they have been in the past, burglary only gets you someone’s used junk that’s not even worth a trip to be ripped off at the pawn shop, the police are better trained at their jobs, technology has made it easier to prevent or deter crime, and the sentences handed out for convictions are brutal.
Linda Featheringill
@Mike from DC:
Lead:
This could be a factor. The US/states/local health departments have been working to lower exposure of young children to lead for some time now. Maybe the total amount of the stuff out there has declined to the point where all this effort is having a positive effect.
We might also consider the benefits of the Head Start Program, started in 1965 [according to Wiki] and expanded since then.
And then there is the WIC Program [Women, Infants, & Children] which has improved the nutrition of pregnant women and young children.
There are probably more programs that I just don’t think of at this time.
But these have a common goal, producing healthier children that are better able to deal with life.
WereBear
And we simply can’t discount the lessened impact of racial prejudice.
Not that we don’t have a ways to go; but there really are more opportunities for young people than they used to be.
At least, until recently.
Tim in SF
This whole post is PREPOSTEROUS. There is NO CAUSAL RELATIONSHIP between violent video games and societal violence or empathy development in children or any other thing, the decline of which which coincides with 1990. NONE!
Video games cure one thing and one thing only: boredom. And people do not commit violent crime because they are BORED.
Linda Featheringill
@Martin:
Or maybe all of those nice social programs for poor children are producing smarter criminals who can now commit more technical but less violent crimes? :-)
Ash Can
I have no idea whether this video game theory is correct, but, as the parent of a 10-year-old boy who’s enamored with online games that involve guns, bombs, bazookas, explosions, and other assorted mayhem (age-appropriate and non-graphic, of course, but still), I sincerely want it to be correct.
DougJ
@Tim in SF:
That only makes sense if you think we descended from monkeys. Hence, it is a liberal theory.
andy
@Mike from DC: Actually, one of my favorite smells when I was kid (back in the 60’s) was leaded gasoline. Mmm- that and El Marko’s when they came out.
Still, switching to unleaded did take away the blue haze and that greasy two-stroke smell that was everywhere there was a lot of traffic.
Cain
@joe from Lowell:
How so? I’ve never heard of this before.. links would be appreciated.
cain
Adam Lang
Freakonomics is wrong. Water is wet. Things mostly tend to fall down, rather than up.
I could keep going, if you like.
frosty
@Ash Can:
Me too. My kid’s on the xBox most of his free time and I can at least say I know where he is. There’s a few sketchy guys around that I’m glad he’s not outside hanging around with.
Meganomalous
In the Freakonomics study, Steven Levitt first thought abortion might be connected to crime rates when he looked at Romania’s enacting an abortion ban and how their crime rate went up afterwards. Obviously abortion can’t be the ONLY factor, but it’s probably the main one. The reason people are afraid to come out and say it is because it’s not in vogue to support abortion these days. Even if you’re pro-choice you have to temper your opinion with something along the lines of, “well, I wish no one HAD to have an abortion…”
It is interesting that the rise of video games correlates (somewhat) with reduction in crime rates, but you have to take into account two things: the video game revolution began a little less than a decade prior to the crime drop, and crime stats for ADULT age groups declined, not just juveniles. In fact the central argument was that unwanted children who WOULD have reached adulthood and began committing adult crimes were never born after the Roe vs. Wade decision. Juvenile crime stats did not figure into the argument. If anything, video games would have a stronger effect on adult crime rates since 2000 or so, when all those college age kids who grew up on Nintendo were playing more adult-themed games on Playstation or computers.
Like it or not, abortion is the number one explanation for the consistent decrease in crime we’ve seen over the last 20 years. The last unwanted child born before Roe vs. Wade is 37 years old, so expect crime to keep decreasing for at least a few more years as that age group mellows out.
Mark S.
How about demographics? The vast majority of crimes are committed by young males. The population has been getting progressively older, to where the median age is now about 37.
I doubt that’s the only factor, but I think it’s an important one.
jl
I think a big part of the drop is due to the aging of the boomers. The country ain’t no spring chicken, populationwise, like it used to be. Older, and wiser, and perhaps most importantly, feebler people tend to commit less of this kind of low class crime.
On other hand… white collar crime, that may be a different story.
Does ED Kain have a link or reference for regional stats on crime rate by rate of abortion?
Carol
@Tim in SF: But video games and the internet get kids off the street and presumably away from a lot of temptation. Stay at home, stay out of trouble. Content is irrelevant, the fact that kids stay home is relevant.
A lot of kids won’t do much unless there’s a crowd to back them up and give them cover. It takes initiative to do stuff alone. In the old days, kids used to hang out outside all of the time, and if someone had a bad idea, others could get swept up in that bad idea immediately. But if you have to gather people in advance, its a little harder to do.
The lead thing is not only eliminating lead from gasoline, but lead has been removed from paint-no more paint chips. When our apartment building was remodeled from its old 1920’s vintage (for the worse in appearance), the lead paint that had been there for decades was removed and replaced with latex and enamel paint.
wmd
@MikeJ:
Global warming is related to the number of pirates. What kind of anti-Pastafarian are you?
Tim F.
America’s toughest sherriff (R) Joe Arpaio brought down crime everywhere in America by graciously concentrating it in his district, where violent crime has gone through the roof. It’s the only explanation.
Tim F.
Also, crime appears to have peaked around when I graduated high school. That cannot be a coincidence.
Cain
@Tim F.:
Are you making an argument that you had something to do with it?
cain
Tim in SF
@Carol
Those are nice theories. Let’s see your data.
JGabriel
Roger Moore:
Exactly!
That’s why rape and murder have gone down – because of the increasingly violent gangs of song and video pirates. It’s not that there’s less crime; it’s that the crimes and gotten WORSE!
.
Martin
@Cain: I don’t have any links but search around for studies on the effects of urban development in NYC in the 50s and 60s. Particularly the Bronx.
NY went through a big and quite ugly period of transportation development to better accommodate automobiles at the expense of mass transit. This had two direct effects:
1) The new roads had to cut through existing developed neighborhoods, breaking up longstanding communities and in areas that were mostly pedestrian, cutting people off form easy access to friends, family, jobs, stores.
2) The decline of mass transit was an added burden on this urban population that didn’t have and didn’t want cars, making it harder for them to get to jobs and school.
There were a slew of secondary effects as well – shifting spending from the areas where taxes were collected to favor suburban development, and so on.
It doesn’t seem like it would be that disruptive, but you have to understand how tightly connected urban, pedestrian areas can be. Cars were incredibly damaging.
Scott P.
Language is hard-wired too, yet children still have to be exposed to an environment where language is used for that potential to become actual. I don’t see why that couldn’t be the case for empathy as well.
Carol
@Tim in SF: I did a little googling and found this:
Crime rates and sedentary behavior among 4th grade Texas school children
Not as conclusive as I thought it would be, and the focus was on a concern about kids being sedentary, but there does seem to be a definitive link as far as parents’ perceptions are concerned.
Kids used to knaw on lead paint, lead paint that would flake and peel off of wooden windowsills and walls. There was a sweet, almost too sickeningly sweet flavor and scent to lead paint. Lead paint was outlawed in 1978, and even before that in urban renewal work, the old paint was scraped off and replaced with harmless latex paints. Add the families that left the slums for suburban housing that never had the stuff to begin with, and the exposure to lead paint declined precipitously. Paint no longer flakes like it used to, even when things get bone dry. So no more paint chips, no more lead exposure.
Nathanael
I really like the idea that street crime just doesn’t pay off decently any more, so the would-be criminals all go into identity theft and computer fraud. Seems likely. Both of those are well up.
jayackroyd
Look, I have avoided getting dragged into the criticize EDK business. It’s fine to have alternative voices, and so on and so forth, und so weiter…..
But this is just wrong:
Freakonomics doesn’t say this. It goes To Great Lengths not to say this. The authors test out any number of alternative hypotheses. They regard this as a Surprising Result, not the only possible explanation. It is decidedly non-intuitive. That’s why they write it up, for heaven’s sake.
If you are going to present other people’s arguments, you have a responsibility to do it in such a way that they would recognize them as their arguments.
jayackroyd
And, duh, Freakonomics uses multivariate analysis, not a series of charts illustrating the simple correlation coefficient. They controlled for shit in their equations.
Did you actually read the book, EDK?
Bnad
I’m surprised everyone seems to be missing the broad cultural trend here.
Episodes of Sesame Street from the 1970s are now too confrontational and violent for today’s kids, so they’re labeled “adults only.”
Enlightened parenting, esteem building, and pharmaceutical mood management are a two sided coin. On average, kids are happier and less rebellious. On the other, they’re more conformist, less original and less likely to call bullshit when it needs calling.
In other words, a future society of sheeple a la China.
lacp
Jeez, isn’t it obvious? All the criminals have gone Galt.
sherparick
Correlation is not necessarily causation, but the causation will always be a correlation.
For some reasons, one the true weaknesses of human reason is our desire for a single, simple, cause for huge, complex social phenomena, such as the crime rate, or the Great Housing Bubble, Bust, and Recession of the Oughts. Our minds react with terror and anger at the chaos of events, so we set forth simple causes (abortion leads to lower crime rates). They are examples of the following error:
The cum hoc ergo propter hoc logical fallacy can be expressed as follows:
A occurs in correlation with B.
Therefore, A causes B.
In this type of logical fallacy, one makes a premature conclusion about causality after observing only a correlation between two or more factors. Generally, if one factor (A) is observed to only be correlated with another factor (B), it is sometimes taken for granted that A is causing B even when no evidence supports it. This is a logical fallacy because there are at least five possibilities:
1. A may be the cause of B.
2. B may be the cause of A.
3. some unknown third factor C may actually be the cause of both A and B.
4. there may be a combination of the above three relationships. For example, B may be the cause of A at the same time as A is the cause of B (contradicting that the only relationship between A and B is that A causes B). This describes a self-reinforcing system.
5. the “relationship” is a coincidence or so complex or indirect that it is more effectively called a coincidence (i.e. two events occurring at the same time that have no direct relationship to each other besides the fact that they are occurring at the same time).
A larger sample size helps to reduce the chance of a coincidence, unless there is a systematic error in the experiment.
In other words, there can be no conclusion made regarding the existence or the direction of a cause and effect relationship only from the fact that A and B are correlated. Determining whether there is an actual cause and effect relationship requires further investigation, even when the relationship between A and B is statistically significant, a large effect size is observed, or a large part of the variance is explained.
Hence it took many statistical and epidemical studies to show that smoking causes cancer.
sherparick
Correlation is not necessarily causation, but the causation will always be a correlation.
For some reasons, one the true weaknesses of human reason is our desire for a single, simple, cause for huge, complex social phenomena, such as the crime rate, or the Great Housing Bubble, Bust, and Recession of the Oughts. Our minds react with terror and anger at the chaos of events, so we set forth simple causes (abortion leads to lower crime rates). They are examples of the following error:
The cum hoc ergo propter hoc logical fallacy can be expressed as follows:
A occurs in correlation with B.
Therefore, A causes B.
In this type of logical fallacy, one makes a premature conclusion about causality after observing only a correlation between two or more factors. Generally, if one factor (A) is observed to only be correlated with another factor (B), it is sometimes taken for granted that A is causing B even when no evidence supports it. This is a logical fallacy because there are at least five possibilities:
1. A may be the cause of B.
2. B may be the cause of A.
3. some unknown third factor C may actually be the cause of both A and B.
4. there may be a combination of the above three relationships. For example, B may be the cause of A at the same time as A is the cause of B (contradicting that the only relationship between A and B is that A causes B). This describes a self-reinforcing system.
5. the “relationship” is a coincidence or so complex or indirect that it is more effectively called a coincidence (i.e. two events occurring at the same time that have no direct relationship to each other besides the fact that they are occurring at the same time).
A larger sample size helps to reduce the chance of a coincidence, unless there is a systematic error in the experiment.
In other words, there can be no conclusion made regarding the existence or the direction of a cause and effect relationship only from the fact that A and B are correlated. Determining whether there is an actual cause and effect relationship requires further investigation, even when the relationship between A and B is statistically significant, a large effect size is observed, or a large part of the variance is explained.
Hence it took many statistical and epidemical studies to show that smoking causes cancer.
uila
Have you considered the effects of the Chicken McNugget?
I heard a very reliable second hand story today about this guy who is apparently a very fussy eater who ate nothing but chicken mcnuggets his whole life, not necessarily from mcdonalds, but your basic fried processed frozen food chicken parts. So he and his wife have been unable to conceive, and the doctors are blaming his low sperm count on a lifetime of mcnuggets.
What I’m saying is, your pet theories about video games and abortion are garbage unless they control for McNuggets.
calling all toasters
The Simpsons first went on the air in December 1989.
Redshift
@Bnad: Er, the early Sesame Street DVD releases have the adults label because they include scenes that are now considered “unsafe,” like adults inviting a child into their house for cookies. I think it’s sad they have to assume today’s parents are that overprotective, but Sesame Street was never “confrontational and violent.”
Laertes
@Tim in SF:
That’s persuasive, of course, but I find that I’m still not 100% convinced. Perhaps if you were to repeat it verbatim a few times, and use all caps on every word instead of a quarter of them, I’d be convinced.
Belvoir
Video games are cheap? Last I looked, the hot ones, the ones that kids covet and talk about, are $50 for new releases. That’s cheap?
Where does a lower-income kid get that money for that game?
Also, I don’t understand the conflation here with childrens’ literature, video games, and crime. Children aren’t the ones out mugging people, young men in their 20’s do. I grew up in the Bronx in the 70’s, NYC crime epidemic at its worst, my dad a cop.
It’s a nice idea, that video games and the Internet are some sort of sedative for the underclass. But in reality, money and sex is what drives people. Young men in their twenties from a poor minority underclass won’t be playing video games 24 hours a day, sorry. And it’s a weird ellision in this optimistic reasoning. It’s sort of condescending, not saying who’s being talked about.
Premise: very debatable, and nerdy as hell. Where did you grow up, E.D. Kain? The city, or some leafy suburb where you imagine today that low-income future criminals all have X-Boxes and Wiis?
No consideration also for the corporate private prison system that throws young black men in jail for profit and actively lobbies for harsher penalties for offenses like pot-smoking? A thousand things.
I’d love to see a shred of evidence for what you’re asserting here.
kyle
increased rates of abortion will reduce crime. Maybe not per capita but overall more dead people = fewer people to commit crime. For this reason I think that abortion should be not just legal but the default. All these damn people are using up my oxygen.
Robert Waldmann
Get the lead out. My favorite freakonomicoid argument about crime is that lead causes people to have temper and self control problems (this is documented) and that getting rid of leaded gas (back when the current non-criminals were little kids and their elder siblings already had fried brains) was the key issue.
The advantage of the lead hypothesis is that leaded gas was phased out (then banned) in different countries at different times. Legalized abortion is highly correlated across rich countries. So is increased video gaming. The US banned leaded gas first, and crime peaked in the US first (and then decreased to a level much higher than the current level in other rich countries — but people burned much much more leaded gas in the USA back in the day).
About your argument contra the freakazoids, they have an 18 or so year lag between abortion and lower crime. Also, by their logic, contraception has the same effect as abortion. It’s not as if more kids are having kids instead of aborting.
Joseph Nobles
How about an inverse relationship between crime rates and obesity rates?
E.D. Kain
@Tim in SF: I don’t think you’re reading the post all that carefully buddy.
grendelkhan
I remember reading somewhere that the abortion hypothesis isn’t borne out by the differing years in which various states legalized abortion, but that the lead hypothesis is borne out by the differing years in which various states updated their building codes or started to require unleaded gasoline. I checked the original NBER paper, but didn’t see it in there. It may have been a different analysis.
Eric
It also correlates with a rise in obesity; could be that people are just too tired all the time and out of shape to go commit crimes.
Dr. J.
Looking at trend lines is insufficient to demonstrate causality. Among other things, you need a really good theory to explain why you think things are connected and then you need to rule out other possible explanations.
In social science, researchers often use regression analysis to rule out other possible explanations.
Regression is statistical wizardry and that is what Leavitt and Donohue did when looking at abortion and crime rates. Rick Nevin also used regression to look at the impact of declining lead exposure.
Interpreting the results means understanding the strengths and limits of regression analysis and it also means really looking at the measures used in the analysis.
From my perspective, it is important to not be dazzled by statistical magic while also remaining appreciative of how hard it is to determine causation in a complex world. We often find that we know a whole lot less than we think we do.
I use these studies in my text book because they stir up lots of discussion and have posted a short summary at:
http://researchdemystified.org/2010/09/explaining-the-declining-crime-rate/#more-519