Freakonomics Summary and Analysis of Chapter 4


In Chapter 4, called "Where Have All the Criminals Gone?", Levitt expands on the crime and abortion correlation previously discussed in the book's introduction. He starts with a case study on Romania. When Nicolae Ceausescu became the communist dictator of Romania, he made abortion illegal. The aim was to boost Romania's population in order to strengthen the nation. Before that point, abortion rates were high in Romania, with four abortions for every live birth.

Within one year of the abortion ban, the Romanian birth rate had doubled. These children born after the abortion ban would lead especially miserable lives, less successful in school and in the workforce on average than children born before them. They were also likely to become criminals. The abortion ban continued until Ceausescu lost his grip on Romania, when in 1989 he was captured and killed by protestors largely consisting of the youth of Romania—the ones who would not have been born, had he not instated the abortion ban.

The Romanian abortion story is the reverse of the American crime story. When crime began to drastically fall in the U.S. in the early 1990s, experts sought an explanation for the fall. Levitt reviews the eight most-cited explanations for this drop. He reveals that only three of these have been shown to have contributed to the drop in crime, and that the number-one cause for the drop in crime is not even mentioned in newspapers at all.

The strong economy, which was the number-six most cited cause, had not been proven to be correlated with a crime drop at all. The number-two most cited cause, increased reliance on prisons, actually did have a significant impact. Imprisonment acts as a deterrent to crime, despite being an extremely costly procedure, and accounted for roughly one-third of the drop in crime. Capital punishment (i.e. the death penalty), however, did not significantly reduce crime, since executions so rarely happened in the United States anyway.

An increased number of police in cities did have a significant effect on lowering the crime rate. However, the number-one most cited cause for the drop—innovative policing strategies, such as smart policing and cracking down on small crimes to prevent larger crimes—has not been proven to have a strong effect, primarily because these strategies only went into play in big cities like New York well after crime had begun to fall.

Levitt then leads an in-depth discussion of one of the other cited causes: increased gun laws. But policies on both sides of the political spectrum had not been shown to reduce crime. Gun control laws like the Brady Act were unsuccessful because of a thriving black market for guns that existed anyway, and right-to-carry laws—the right-wing policy that advocates for putting more guns in the hands of the right people—do not bring down crime either. One explanation that did have some effect on the crime drop was the bursting of the "crack bubble," when profits for dealing crack began to fall so the crack-dealing tournament lost its allure and gang violence abated.

Levitt moves on to explanations that center on demographic change. The first—the aging of the population—was too slow to explain the sudden crime drop. The demographic change that did have an effect was the legalization of abortion, as discussed in the introduction. By the year 1900, abortion was illegal across the U.S. By the 1960s and 70s, abortion slowly became legal in extreme circumstances, and then on January 22, 1973, abortion was suddenly legalized across the entire country in Roe v. Wade. The decision acknowledged that when a woman does not want to have a baby, she usually has a good reason, often acknowledging that the state of her own life is not conducive to raising a child.

In the wake of Roe v. Wade, 1.6 million American women had abortions by the year 1980. Because it had also gotten less expensive, the type of woman likely to take advantage of legalized abortion was unmarried, poor, or in her teens. Had these children been born, they would have been 50 percent more likely to live in poverty and thus extremely likely to have a criminal future.

Levitt acknowledges that it may be more comforting to believe that the newspapers were right and the crime drop was due to brilliant policing or clever gun control. He says this is because we prefer to link causality with things we can touch and feel, rather than with a distant or difficult phenomenon. But legalized abortion undoubtedly had a dramatic effect on the crime rate: states with higher abortion rates had higher drops in crime, and states that legalized abortion earlier saw crime start to drop before other states.


The primary takeaway from this chapter is that large effects sometimes have distant, unexpected causes. Aside from abortion itself being a controversial subject, the main reason why it is difficult for people to accept Roe v. Wade as the number-one factor contributing to the crime drop is that, since it happened nearly 20 years before, it seemed so far removed from the present. It is easier to believe that some direct action we have control over—like changing policing strategies or instating gun control laws—produce the immediate consequences we seek, but often discovering the real cause involves taking Steven Levitt's approach and digging deeper into the data.

Using the story of Romania as a reverse case of what happened in the United States helps Levitt to accentuate the latter account. Since Romania experienced the exact opposite effect—banning abortion led to a drastic increase in crime approximately a generation later—it is clear that abortion truly is linked to crime. It also paints a chilling picture of the kind of crime levels that may have been similarly seen in the U.S. had the Roe v. Wade decision gone the other way and the crime rate continued rising.

But it is important to remember that correlation does not always prove causation. Even if two things show a relationship, one may not necessarily be causing the other—further analysis of the data through isolation of the variables in question is necessary in order to prove causation. Levitt does this in his analysis of all of the additional explanations cited to explain the 90s crime drop; while innovation policing strategies appeared to be correlated with the crime drop, in reality it was the rise in the number of police officers—which went along with the change in policing strategies—that had the more significant effect.

This chapter also introduces the concept of black markets. A black market—sometimes called an underground economy, or a shadow economy—is a market characterized by behavior that is, in some way, noncompliant with the law. This means that black markets are either selling illegal goods—drugs, for instance—or selling legal goods in an illegal way, like guns being sold without background checks or permits. It is often difficult to enforce laws regulating the sale of certain goods because of the prevalence of black markets, as Levitt notes in his analysis of the effectiveness of gun control laws.

Finally, the content of this chapter highlights the importance of knowing the difference between positive versus normative analysis. A positive analysis is objective and fact-based. Positive analyses explain a phenomenon that is observable through data, such as the fact that legalized abortion and low crime rates are correlated. Normative analysis, however, is subjective and value-based. A normative analysis of this phenomenon might ask the question of whether or not abortion should be legalized. This chapter only consists of positive analysis, not normative analysis—even though Levitt's data has shown a relationship between these two factors, he does not attempt to pass a value judgment on the idea of legalized abortion itself.