The Next Crime Wave is Upon Us, Right?

The following are headlines from this week’s major newspapers following the release of official 2006 crime statistics:

From the Washington Post:

Violent Crime, a Sticky Issue for White House, Shows Steeper Rise

From the Los Angeles Times:

Violent Crime Rises Again

From the Philadelphia Inquirer:

Rise in Violent Crimes is Higher than Expected

It is official. The next crime wave has hit. According to the Washington Post, “[v]iolent crime in the United States rose more than previously believed in 2006, continuing the most significant increase in more than a decade, according to an FBI report released yesterday.”

As I have previously noted, the media loves to put the most frightening spin possible on crime statistics, no matter how benign they might actually be. They did not disappoint with the latest set of statistics.

What actually happened to crime? The number of violent crimes increased 2 percent and the number of property crimes fell 2 percent. The more relevant numbers are the crime rates, which take into account changes in population. Since the U.S. population is growing by 1 percent per year, the number of violent crimes per person rose 1 percent and the number of property crimes per person fell by 3 percent.

This figure from the L.A. Times tells the story far better than the headlines.

It is also interesting to note this dinky little A.P. item the other day about the substantial drop in New York State crime in 2007:

New York State’s overall crime rate continued to drop in the first half of 2007, with the number of rapes decreasing 19 percent and robberies dropping 13 percent, according to a report issued yesterday. Gov. Eliot Spitzer announced that the number of reported crimes dropped 5 percent, to 208,265 in the first half of 2007 from 219,550 in the first half of 2006. Murder was down 4 percent, vehicle thefts dropped 15 percent and burglary decreased 7 percent, the report said.

Like they say, if it bleeds, it leads; if not, it doesn’t get the front page or a big scary headline.

That said, the incredible declines in crime we saw in the 1990s do seem to have run their course. This is not a surprise. The major forces that I believe drove that decline — increases in the prison population and legalized abortion — have also leveled out. Prison populations are creeping up, but not growing by the leaps and bounds of the last few decades. On the abortion side, virtually every active criminal these days was born after abortion became legal, so the theory doesn’t predict any further declines in crime.


Would you mind extrapolating your theory on the link between violent crime and legalized abortion?


I agree with your position generally, but, if the number of crimes increased by 2 percent, then, the population would have to double if the rate were only to increase by 1 percent. Or am I missing something?


To Louis:

If you want to read the theory, do two things:

1. Buy the book, and;
2. Click the "Abortion" tag above. It will take you to all of the Abortion postings on the site


Louis: Read Freakonomics (the book, not the blog).


Go Barnes & Noble, find "Freakonomics", buy it, and read it!


Here is a pretty decent crime report amp website. At least in DC, its much better and more user-friendly than the government websites on crime.


As my Crim Law Professor put it yesterday, "If you really want to see what it is like in the world, open your front door, look left, look right, and go back inside. It is often not what the media portrays it to be".


It's iPods!


Well, actually if you want to read about the abortion and crime rates thing in detail, you should just read the academic paper he wrote about it :)


Have you read freedomnomics yet? The rebuttal where he says that abortion actually increases crime.

I'm not sure he's 100% correct, but it's good brain fodder.

pat ballew

MIke (number 7) You reminded me of a t-shirt one of my students wore the other day... it said..
" I've been outside once, and the graphics weren't that great."


Alan, if crime is up 2%, it's at 102% of the prior value. If the population's up 1%, it's at 101% of its prior value (and at a constant crime rate, crime would also be at 101%).

So crime is really only up by 1/101 = 1% (or a little less) over the previous rate.


Whites are more likely to have an abortion then African Americans. Assume that Levitt is correct and abortion reduces the crime rate. Does that mean that the face of crime should be increasingly black?

Also I think Levitt is wrong on the abortion crime link. I would look at how society dealt with unplanned pregnancy in the 50's. If you are the oldest child in your family, and you were born in the fifties, the odds are that your mother was with child on her wedding day. But most families dealt with the issue, often in the context of extended families.

By the 60's doing "the right thing" for a pregnant women changed. Welfare and other social programs eased the burden of out of wedlock parenting. Males were more likely to be absent without as much stigma. Being sexually active with numerous partners was less taboo.

So for me, the change was in how society, and families dealt, with unexpected pregnancies. The collapse of the family lead to an increase in crime. Abortions were a partial answer to the collapse of the family and did moderate crime rates. But the true root cause, was the increasing collapse of the traditional family.

This is important if you want fix a problem, fix the root cause.



Gary, Thanks . . .(I actually had to do a numerical example for myself!!)


"I would look at how society dealt with unplanned pregnancy in the 50's."

Actually, going on anecdotal evidence - if a girl was pregnant during those times, the choices were:

(a) quick marriage - provided she was old enough and the boy suitable to the family. Children in this situation may or may not have been at risk.

(b) adopted out - for younger girls or less suitable fathers. These children gre up with better opportunities. FAS was less likely to be a factor.

(c) kept - for the bottom of the social ladder. They had nothing to lose; the father probably wasn't worth marrying, or long gone. Also, at risk of FAS. These were the children at risk of growing up to be criminals. Frequently, these are the ones entering the foster circuit at 1 or 2 years old, after the mother discovered she couldn't handle them.

The first two groups probably have abortions nowadays at the same or a higher rate than the last. When abortion became a choice, however, it's the last group that affects the crime stats the most.

Of course, there is a much larger trend in second group is to keep the chid nowadays.



To Mr. Levitt,
What are your predictions for crime rates in the next few years? Do you think that the small increases in the past few years (for violent crime, anyway) will continue, or are these just random deviations around the baseline?
Do the authors read these comments?


DanC, youre post is hard to follow. Why do you say whites are 'more likely' to have abortions than blacks? Some evidence please? And do you mean as a raw number? That wouldn't be surprising given the vastly higher number of whites than blacks in the population. Do you mean as a percentage of the population? If so, by how much?

And what's with the playing of the race card with the "does that mean the face of crime should be increasingly black'? Do you think that trying to make someone feel like they are giving a 'bad' answer should change the truth? Worse, your question makes no sense in light of the supposed fact you gave.

You are assuming from the word go that x percentage of people will become criminals no matter what, and ditto for blacks.

So if both whites and blacks are good for say a ratio of 1/10 criminals/total babies then aborting the white ones who are most likely to be criminals would lower the criminal population no matter what.

I don't think this makes any sense. There's no cap or ceiling the criminal population based on race. It would make more sense to say that there is some percentage of hopelessly poor men that will become criminals and the more hopelessly poor men there are, the more of them will turn to crime. Since they were being aborted more often across the board, there would be less of them to turn to crime.

None of this has anything to do with 'family values' as you wish it would, it has to do with economic reality, and yes a stable family provides better economic possibilities to children. But this is a practical point, not a moral one. So please get off your high horse.



I was wrong. I was under the false impression that African American women were less likely to have an abortion. According to the CDC in year 2000: "The abortion rate for black women (30 per 1,000 women) was 3.1 times the rate for white women (10 per 1,000 women), whereas the abortion rate for women of other races (22 per 1,000 women) was 2.2 times the rate for white women."

Based on conversations I have had with people who work with agencies that assist low income individuals, they told that that their experience is that African American clients are far less likely to have an abortion. Unless the CDC is for some reason missing data on abortions, I will go with the CDC data.

Professor Levitt claims that the legalization of abortion explains, in part, a decrease in crime rates. The story goes something like this. Roe Vs Wade made abortion an option for poor, young and unmarried women. Having a mother who is young and uneducated, poor, and unmarried (or alone) puts you at greater risk for a life of crime. Abortion removes unwanted babies from the population, these babies were at greater risk for criminal activity, fewer unwanted babies means less crime.

One way to test Professor Levitt's theory is to see if it holds true if you control for race. That is not playing the race card. If African American women are having abortions at 3.1 times the rate of white women, then why is violent crime so high in the African American community. In the African American community, does abortion increase with income or education? I don't know. Perhaps absent abortion crime would be even higher in the African American community. Again, I don't know.

So perhaps DamienD is correct and I should assume that African Americans are more likely to be poor and to engage in criminal activity because they are poor. (Or please explain the racial composition of the American prison population.) Perhaps whites are better at aborting potential criminals and that alone can explain the total drop in crime rates and explain, in part, the racial composition in prisons. Or is that playing the race card.

Or perhaps there is a hard to reach minority within the African American community of children having children producing children where a life of crime is accepted. The root cause of the problems in this community is the collapse of the family.

A few other questions. Using the Levitt model. Why wouldn't widespread adoptions have the same impact on crime rates as abortions? Why wouldn't shotgun weddings, not much worse then the arranged marriages, create some family stability that in turn prevents at least some criminal activity? Shotgun weddings were common during a period of lower crime rates. Did Grandparents decrease the chances of their grandchildren becoming criminals by forcing marriages? I am happy to ride the largest horse I can find if that means that I think that an increase in abortions mask the social problems that run in tandem with increased abortions.



What is really correlated with high school student success ?


My comment to Steven Levitt.

Aren't abortions increasingly difficult to obtain in some parts of the country? Certainly in rural areas that's true. I think I read a news story that there is only one abortion clinic in all of Mississippi. In addition, how have the Hyde Ammendment (which stopped federal money paying for abortions) and this lack of access affected crimes rates? It would be interesting to look at this county by county or state by state.