Crime and Crack

by Tyler Cowen on May 23, 2005

Economists should be forward rather than backward-looking, so I will consider Steve’s new paper — "Measuring the Impact of Crack Cocaine," co-authored with Roland Fryer, Paul Heaton, and Kevin M. Murphy.

The contributions of this paper are clear:

1. Crack cocaine enters the U.S. market in a big way in the mid-1980s.

2. It is possible to construct a meaningful time series for crack usage.  This is the paper’s most significant contribution.

3. Murder rates rise dramatically when crack comes.  If I recall Steve’s talk correctly, some young black homicide rates nearly tripled, of course starting from an already-high rate.  This is a startling statistic.

4. Today crack continues to be consumed at high levels but mostly by longer-term addicts.  This kind of crack consumption causes much less crime.  Such a causal story, however, is closer to speculation than a proven result; it cannot be read off the numbers.

A central question is why crack has caused so many more murders than have other drugs.  I can think of a few possible (and possibly false) arguments:

1. Heroin and pot make you sleepy.  Crack gets you riled up.

2. Crack was a new drug when it hit the market.  Gangs were competing to hook new buyers.  This is a far more violent activity than serving established drug clientele.

3. In dollar terms crack was a "bigger" drug than ever before.  The gross and the profit margins were bigger.  The resulting turf wars over profits led to murders.  It is not worth killing people over a few marijuana sales.  (Yet still I find this puzzling.  Falling prices have taken profits out of the market; the gangs must either have had an extraordinarily high discount rate or they behaved irrationally in killing each other.  In the latter case we have no economic explanation at all for the hike in crime.)

4. Perhaps you buy other drugs from your friends, but you buy crack from dealers.  (Most people get Ecstasy from their friends, and this market is not very violent.)  The new question is then why this might be.  Could crack somehow require less personal certification from trusted acquaintances?

These market features might have implications for drug policy.  Most of the discussion focuses on either how to stamp out drugs or why we should legalize drugs.  We should consider a new approach based on differential features of drug markets.  Let us find out why crack differed from other drugs so much.  Then watch for when these critical features resurface.  If "another crack" comes along, what should we do to ameliorate the crime problem?  Encourage people to buy it from their friends?  Encourage exclusive territories for the new drug?

Steve’s paper has given us considerable food for thought.  But I would like to see him take the next step of pinning down the relevant policy questions.  Why has the crack market become safer rather than more dangerous?  Which features of drugs lead to associated crime waves?  Can we make drug markets — yes "black" ones — less dangerous?



Ginger Yellow 05.23.05 at 11:36 am

1, 2, 3 and 4 are all true, as I understand it, but I think the key factor is that crack is a very intense but very short high. Combined with its extreme addictiveness, this is a recipe for disaster. A heroin addict will get his fix and not come back to his dealer for a day or more. A crack addict could be back in half an hour. More transactions, more risk, more chance of confrontation, more chance of going into debt, on top of a crack addict’s inherent violence. It’s not going to be pretty.


Ginger Yellow 05.23.05 at 11:38 am

As for this: “Why has the crack market become safer rather than more dangerous?” Long term users, while chaotic in their personal lives, are more likely to have a handle on their addiction than newer users. They’ll plan ahead (as far as the next fix, anyway) making the debt situation less likely. Even if it’s only “I need to steal two DVD players to fund next week’s supply” it’s still planning.


Michael H. 05.23.05 at 12:14 pm

Hi Tyler
Interesting post.
Number 4 is intriguing. Is it possible that the crack cocaine producers are the Starbucks Coffee of the drug world: they didn’t want to franchise? Just a thought.


Barry 05.23.05 at 1:02 pm

IIRC, somebody blamed part of the violence on the idea that, in the late 80’s/early 90’s, police in major cities were successful at destroying large crack organizations as they (through lots of violence) consolidated. This kept things in the high-violence consolidation stage, for an unusally long time. Supposedly, by the late 90’s most of the large cities had consolidated crack organizations, and needed less violence (i.e., their reputation was doing the work).


dipnut 05.23.05 at 1:17 pm

…the gangs must either have had an extraordinarily high discount rate or they behaved irrationally in killing each other. In the latter case we have no economic explanation at all for the hike in crime.

Glenn Reynolds, pondering a review of Freakonomics, suggested that crack dealers are motivated more by status than money. Certainly, the economic rewards are meager, especially in view of the high risk of murder in the business. Also, crack is associated with a status-obsessed culture (replete with symbols such as flashy cars, bling-blings, “bitches”, etc., as well as the roles and expectations of a gang hierarchy) for which there is little or no parallel with, say, heroin or marijuana.


Jos Marq 05.23.05 at 4:29 pm

they behaved irrationally in killing each other. In the latter case we have no economic explanation — Tyler Cowen

crack is associated with a status-obsessed culture — Dipnut

Any business activity — whether you’re Willy Loman or the corner drug dealer — is associated with status in a culture where your profession and how much money you make IS who you are. The millionaires who play the market are not, strictly speaking, doing it for the money. They’re doing it for status, just like their poorer counterparts in less legitimate enterprise.

Because rap videos are how many white Americans encounter Black culture, it’s not surprising that this very narrow genre of expression has become a quick reference guide for understanding “that mentality.”

Well, that mentality is not specific to Black people and it’s certainly not restricted to the black market of illicit narcotics. Everyone in America believes in “bling,” and if the current administrations initiatives to prevent teen pregnancy are any indication, lots of people, rich and poor, black and white, believe women are possessions, call them bitches or not.

Every industry, every social group has hierarchies and resorts to force to settle disputes. If you’re IBM and you’re unhappy with what a competitor is doing, you take them to court. If you’re a drug dealer, you can’t very well do that. It’s not hard to understand that competition will lead to “litigation by firearm” nor is it likely that the cheap and plentiful supply of firearms in America is unrelated to the “epidemic” of black-on-black crime in American inner cities in relation to the introduction of crack.

Those of us who have had access to good education, to higher education, who read web sites like this one, can’t possibly understand why kids would kill one another for such meager earnings as can be had from the crack trade. But if your education is so lacking that you don’t believe you have more choices (not the same thing as knowing, but, really, believing it), you’ll take the risk.

After all, the rich, white President of the U.S. is going around the country putting on a very expensive show to convince Americans that they should bet it all in the hopes of becoming very well off when, in fact, they’re most likely to become very poor by buying into his policies.

“Who wants to be a millionaire?” It’s not just the gangbangers.


ambienceman 05.24.05 at 8:22 am

This correlates so well with the release of Chappelle’s Season 2.


Gabriel Mihalache 05.24.05 at 8:46 am

Your first argument is in doubt. From Sigmund Freud’s “On Cocaine”:

A few minutes after taking cocaine, one experiences a certain exhiliration and feeling of lightness. One feels a certain furriness on the lips and palate, followed by a feeling of warmth in the same areas; if one now drinks cold water, it feels warm on the lips and cold in the throat. One other occasions the predominant feeling is a rather pleasant coolness in the mouth and throat.During this first trial I experienced a short period of toxic effects, which did not recur in subsequent experiments. Breathing became slower and deeper and I felt tired and sleepy; I yawned frequently and felt somewhat dull. After a few minutes the actual cocaine euphoria began, introduced by repeated cooling eructation. Immediately after taking the cocaine I noticed a slight slackening of the pulse and later a moderate increase.

It might be that the high and actual symptoms are of call and pleasure while the withdrawal is violent.


Andrew Edwards 05.24.05 at 8:46 am

…the gangs must either have had an extraordinarily high discount rate or they behaved irrationally in killing each other. In the latter case we have no economic explanation at all for the hike in crime.

Is it that hard to believe that someone in such an incredibly dangerous line of work would have a very high discount rate? If I didn’t really expect to live past 30, I think (at 26 years old) that I’d severely discount even mid-term, let alone long-term future returns.

If that’s true it makes sense that there would be abnormally intense competition for the early above-MC returns, since they’re the only returns anyone was even thinking about. Abnormally intense competition in illegal drug markets manifests as violence.

Moreover, this cycle would feed itself: intense competition would escalate violence, which would reduce life expectancy, which would intensify short-run competition.

I’m assuming here that we’re talking about a new technology that launches into the market, creates monopoly-level returns for early movers, and then turns into a more-or-less perfectly competitive commodity market fairly quickly (See, e.g., computer RAM, only faster). Which sounds like what you’re assuming.


Lockjaw the Ogre 05.24.05 at 9:01 am

An associate of mine, who took a LOT of drugs “back in the day” gave me one of the most insightful statements I’ve ever heard on crack.

He said pot made you mellow. Heroin was a real rush. Cocaine was a fast high. Speed was this, downers were that, acid was the other thing.

Crack, though, he said was “Satan.”

According to him, he had only used crack a few times, and it had been years, but he would be driving down the road and suddenly, he’d think, “Boy, I’d sure love some crack right now.” It just never let you go.


john b 05.24.05 at 10:55 am

Sigmund Freud clearly had some bad shit, man. Coke does not make one *feel* sleepy, notwithstanding its effect on the pulse rate…


Christine Hurt 05.24.05 at 11:29 am

I like “jos marq”‘s analogy to “litigation by firearm.” What about “consolidation by firearm”? If the introduction of crack to the drug market is similar to the introduction of any new technology or industry, then you can describe the crack market in the 80’s as a start-up market. In most start-up markets, you see fierce competition and rapid consolidation through acquisitions. However, in a drug market, players may not want to consolidate to be able to compete, so violence may ensue. Turf battles may take the place of hostile takeovers and firm failures. After the start-up phase of the industry, a more mature industry evolves with fewer players that compete on reputation and price.


anno-nymous 05.24.05 at 3:38 pm

I just thought I’d note that for Levitt’s undergraduate Economics of Crime class, we had a problem set that was essentially to figure all this stuff out as best we could. Come up with theoretical reasons for why crack use might lead to violence; find some sort of time series for crack use and prices; work out a simple model; find out what happened to crime and crack use over the 1980’s-1990’s. It was fun.


Steve Sailer 05.24.05 at 3:48 pm

Perhaps coincidentally, perhaps not, the crack wars took off in the late 1980s precisely in those metropolises that had legalized abortion, de jure or de facto, in 1970, three years ahead of Roe v. Wade: NYC, LA, and DC.

The crack wars, which were fought by exceptionally young males, then spread nationwide as the first generation born after Roe v. Wade entered their mid to late teenage years. Both the homicide and “serious violent crime” rates for ages 17 and under peaked in 1993 and stayed very high in 1994, according to FBI statistics collected under highly different methodologies.

And the wave of crack violence was worst among the group with the highest abortion rates in the 1970s: urban blacks. The non-white abortion rate, according to the Alan Guttmacher institute peaked in 1977, and was already at it’s long term level by 1976, so the timing fits much better for the theory that abortion increased violent crime than Levitt’s now-famous theory that it cut violent crime.

Correlation isn’t proof, but there seems to be about as much evidence for the proposition that legalizing abortion increased the murder and serious violent crime rates as for Dr. Levitt’s theory that abortion decreased it.

Why was there this geographical and demographic correlation between the usage of legal abortion and the crack violence wave? Dr. Levitt has tried to ignore this correlation for six years. He has tried to define the question surrounding abortion’s effect as only whether it played a role in the decrease in crime in the mid-1990s and avoided the whole question

Perhaps legalizing abortion had a corrosive effect on urban black culture.

To speculate: First, legalization tended to kill off marriage by removing the moral burden on the impregnating boyfriend to marry the girl because she could now get an abortion with the full approval of the law. Levitt claims that legalizing abortion increased the “wantedness” of the babies who were born, but it sure didn’t improve how much the fathers wanted them: the illegitimacy rate soared. As Levitt himself points out, being raised in a single parent home seems to double the chance of being a criminal.

Second, legalizing abortion, according to Levitt himself, drove down the birthrate by only 6% but increased the pregnancy rate by almost 30%. In other words, it vastly increased the rate of unwanted pregnancies. Let’s think about the impact on respect for life among young males. If you live in a culture where role models tell you that the proper response to an inconvenient unborn baby is an abortion, does that have an impact on your attitude toward the proper resonse to inconvenient business rivals?

Third, let’s look at the effects of legalizing abortion on urban blacks from a selectionist point of view. Both Levitt and I are much more selectionist in our thinking than is politically correct today — i.e., as you’ll see from his chapter on parenting techniques, he believes who the parents are is more important than what they do. Obviously, a big part of the sub rosa appeal of Levitt’s abortion-cut-crime theory is that blacks get more abortions than other groups, a point he made in his original 1999 paper with Donohue, although he dropped in later versions.

But which blacks were getting abortions? According to the Alan Guttmacher institute, almost as many unintended pregancies end up in a birth as in an abortion, so the effect on the wantedness of the babies actually born is much more uncertain than it sounds like in “Freakonomics.” It is quite likely in fact that legalized abortion culled proportionally more middle class and working class blacks, while underclass blacks tended to just have the babies. Thus, legalized abortion could have lowered the social standards of blacks. Indeed, if you compare the kind of music that blacks youths were listening to in 1970 (e.g., James Brown, Marvin Gaye) to the kind of music they were listening to in 1993 (gangsta rap, which celebrates the ethos of the crack dealer), it does appear that the social standards of blacks declined, which would be in line with this theory.

You can learn a lot that Levitt doesn’t want you to know about the relationship between abortion and crime at


Chris 05.24.05 at 5:08 pm

Only number 3 seems plausible to be.

A couple of points:

1) Are we talking about crime committed by users or crime committed by dealers? These are distinct types of crime. I believe an overwhelming proportion of drug crime is related to dealers, not users. If that’s the case, I don’t see why the nature of crack’s high is directly relevant.

2) Gangs compete for territory, but the idea that gangs run around “hooking” everybody on drugs is ridiculous. This is the talk of people who don’t actually use or buy drugs (people who don’t what they are talking about). Gangs sell drugs, and believe me they don’t have any problem finding buyers. They don’t need to actively try “hook” people.

I would guess more violence by dealers could occur with crack, relative to other drugs, if crack was substantially more profitable than other drugs. Perhaps the nature of crack’s high is indirectly relevant to this point… that is, perhaps crack is more profitable because buyers tend to buy more frequently and have relatively inelastic demand curves. I’m sure those plenty of other reasons why this crack could more profitable than other drugs.

On a side note, I almost always find it funny to hear/read people discuss drugs that have clearly rarely, if ever actually used drugs, purchased drugs, or been involved in drugs in any way.

I’m NOT saying only drug users should be involved in discussions of drugs (to make such a statement would be absurd). I am saying non-drug users should keep in mind they have zero experience with drugs and that most of their information probably comes from other people with a similar level of experience with drugs that also happen to have anti-drug biases.


Barbar 05.24.05 at 5:54 pm

Indeed, if you compare the kind of music that blacks youths were listening to in 1970 (e.g., James Brown, Marvin Gaye) to the kind of music they were listening to in 1993 (gangsta rap, which celebrates the ethos of the crack dealer), it does appear that the social standards of blacks declined, which would be in line with this theory.

Furthermore, gangsta rap actually became mainstream in the late 1990’s, as your theory predicts. Suddenly it all fits together:

1) In the 1980’s, mainstream music is Tears for Fears, the Cure, and Madonna.
2) The legality of abortion pervades white America, leading people to place less of a value on human life.
3) This naturally leads to a decline in social standards in white America, the best evidence of which is the popularity of Ludacris.
4) When you combine this with the Monica Lewinsky blowjob and Bill Clinton’s consequent cover-up, it comes as no surprise that the late 1990’s were marked by skyrocketing crime rates, economic unrest, and the dot-com boom.

Thanks for your insight.


Steve Sailer 05.24.05 at 10:30 pm

As the crime rate came down in the mid 1990s, so did the abortion rate, and the illegitimacy rate started to plateau after shooting upwards for decades. It actually fell among African-Americans.

So, here’s a challenge to Dr. Levitt to make a courageous prediction about the future, rather than the fail to explain the past.

According to the pro-choice Alan Guttmacher institute, the abortion rate fell during the 1990s. It fell especially among non-Hispanic whites, from about 19 per 1000 women during their child-bearing years in 1991 to only about 11 by 1999. See p. 8 of this PDF:

Does Dr. Levitt believe that this decline in abortion is going to drive up the juvenile crime rate, especially among non-Hispanic white youths, beginning in a a couple of years? That’s what the logic of his theory says. Should we start building more juvenile detention facilities to accomodate the coming wave of unaborted young criminals? If he really believes it, let him make a prediction that we can evaluate in a number of years.


Strophyx 05.24.05 at 10:38 pm

IIRC, Freud and most cocaine users of that era injected it, rather than snorting or smoking. As a result it hits the brain more slowly and in a much more diluted form than either of the currently prefered routes of administration. It’s not at all surprising that the effects would be different in the two cases.


Tim Lambert 05.25.05 at 5:24 am

I don’t see how a fall in the abortion rate in the 90s leads to a prediction of increased crime under the Donohue-Levitt theory. Presumably the fall in the rate was due to a decrease in the demand for abortions rather that a decrease in the supply of abortions, so it doesn’t seem that their would be an increase in unwanted children.


Cole 05.25.05 at 8:25 pm

Steve Sailer, your tone is uncalled for, as are your countless posts.


Steve Sailer 05.26.05 at 3:01 am

“Steve Sailer, your tone is uncalled for, as are your countless posts.”

Another highly logical and empirically detailed defense of the facts and theory behind Dr. Levitt’s abortion-cut-crime theory!

Okay, you’ve convinced me. Dr. Levitt is right about everything and if you all take on faith everything he says, you’ll all follow him into bestselling authordom and media celebrityhood.

In contrast, Tim Lambert makes an interesting and constructive point about whether, under Levitt’s theory, the post 1991 decline in the abortion rate would have effects counter to the post 1969 increase.

Some pro-choice organizations lament the decline in abortions, arguing that is the result of the supply of abortionists (excuse me, reproductive service providers) being driven down by harassment and even murder. Plus, states have imposed some restrictions on abortion and funding. How much that hs limited the supply of abortions, I cannot say at present.

Or, it could be that the abortion rate has declined due to a return to moral traditionalism, as also seen in the fall in the crime rate, the leveling off of the illegitimacy rate, the popularity of fundamentalist Christianity, and so forth. At present, I don’t have a way to evaluate this quantitatively, but it seems not implausible and should be investigated further.

Or, it could reflect improvements in either contraceptive technology (e.g., long-lasting injections like Depo-Provera) or behavior (e.g., the return of the condom, due perhaps to fear of AIDS — by the way, Jonathan Klick found that legalizing abortion drove up the sexually transmitted disease rate by about 30%).

At this point, I can’t judge among these explanations (or other possible ones). In any case, since Levitt’s unwantedness model of how abortion would cut crime is so unrealistically simplistic — as he admits in “Freakonomics,” the birthrate only declined 6% after legalization while the pregnancy rate went up almost 30% — we would just be piling uncertainties on top of uncertainties.

So, since Levitt’s theory did such a poor job of predicting the past, I agree that it’s too much to ask it to predict the future.

Comments on this entry are closed.