Outdoor air pollution suffers a similar “common pool” problem; if too many individuals or companies emit too much pollution, the combined result can produce illness and even death. Everything from cars to power plants emit byproducts that could be classified as harmful, but no one would argue that we should eliminate cars and power plants because their pollution costs outweigh all their benefits. Similar to fishermen out at sea, individual car makers or factory owners are unlikely to take into account the cost that their pollution imposes upon others. Altruism only goes so far. This creates a legitimate space for government intervention—governments can regulate pollution levels by limiting, taxing, or otherwise restricting pollution emissions.
The problem here is that the government’s inevitable tendency is to increase the scope of its authority. Allowing the government—whether federal, state, or local—to regulate pollution may be necessary, but we can only watch with dismay as the government uses this authority to steadily expand its coercive powers. In doing so, it inevitably begins mandating solutions to tangential “problems” that are best left to the market to solve.
This is most evident today in the restaurant industry, where local governments are becoming increasingly intrusive in ensuring clean air by banning smoking. A restaurant owner will face competitive pressures to decide whether his customers want to be allowed to smoke, just as he must figure out what food to serve, how big the portions will be, and what kind of décor to have. Restaurants that don’t satisfy their customers on these issues will quickly go out of business.
That’s why smoking in restaurants should not be considered a common pool problem subject to government intervention. Restaurants that allow smoking don’t base their policy merely on their love of smokers; they are responding to competitive pressures and customer preferences. And of course, the choice doesn’t have to be between perfectly pristine air and air so cloudy that you can’t see more than a few feet; many restaurants simply create smoking and non-smoking sections. But this solution, too, is outlawed by government smoking bans.
If restaurants can go out of business when they don’t get small things right, like whether to include an extra side dish with a meal or what kind of background music is played, why is there any less reason to believe their decision on smoking will likewise reflect customer demand? If anything, the fact that people feel so strongly about smoking—both for and against—implies that restaurants will very carefully tailor their smoking policies to their customers’ wishes.
Restaurant smoking is more like a private fishery than a common pool. And the best thing the government can do to ensure people get the air quality they want is to let the market decide.
To conclude, the free market is remarkably inventive. Given enough time, all kinds of seemingly insurmountable problems can be overcome if someone stands to earn a profit by finding a solution. In contrast, government intervention, although often well-intentioned, faces inherent economic and political problems that make it costly and naturally inefficient. As Milton Friedman famously pointed out, “Nobody spends somebody else’s money as carefully as he spends his own.”72 Looking at examples from Airbus to NASA to state licensing systems, it’s hard to disagree.
4
Crime and Punishment
If something becomes more costly, people will do less of it. This is the fundamental principle of economics—a simple notion that also explains a lot of human behavior in realms seemingly far removed from trade, industry, and finance.
Take sports, for example. When college basketball’s Atlantic Coast Conference increased the number of referees per game from two to three in 1978, the number of fouls dropped by 34 percent. Why? Basketball players fouled less often because they were more likely to get caught. In fact, the actual decline in fouling was probably even larger, since fouls that may have gone unnoticed by two referees were more likely to be caught when there were three.1
We find the same kind of incentives at work in baseball. The American League has more hit batsmen than the National League, but this difference only appeared after 1973, when the American League removed its pitchers from the batting lineup in favor of designated hitters. Since American League pitchers no longer worried that they themselves would be hit in retaliation if they threw at an opposing batter, they began throwing more beanballs.2
The reverse of this fundamental principle also holds true: when you make something less costly, people will do more of it. And if you offer a meaningful reward for some type of behavior, you can bet more people will do it, even if they shouldn’t. This is borne out by a study of a specific type of fraud by air traffic controllers.3 To receive disability benefits due to job-related stress, air traffic controllers must present a well-documented stressful incident—a collision or close call—that has caused a deterioration in their performance. Unsurprisingly, when it became easier to file for disability, flights suddenly started experiencing more “close calls.” And these were not cases that the air traffic controllers could simply make up; they were reported by a sophisticated performance evaluation called the “Operation Error Severity Index.”4
Analyzing incentives in this way is a particularly good method for studying crime. It helps reveal what factors increase criminality, and what methods work best to reduce crime rates. Studying incentives also leads us to some surprising conclusions about the effects of abortion, affirmative action, and the death penalty on crime, while showing how certain kinds of high penalties and other policies can inadvertently increase crime rates.
Why did Crime Fall During the 1990s?
“[It] remains one of the great mysteries of our time.”
—Wesley Skogan, a criminologist at Northwestern University, on the long and steady drop in crime during the 1990s.5
Violent crime in the United States shot up like a rocket after 1960. From 1960 to 1991, reported violent crime increased by an incredible 372 percent. The disturbing trend was seen across the country, with robbery rates peaking in 1991 and rape and aggravated assault following in 1992. But then something unexpected happened—between 1991 and 2000, rates of violent crime and property crime fell sharply, dropping by 33 percent and 30 percent, respectively. Murder rates were more stable up to 1991, but then they also plunged by a steep 44 percent.6
Murder and Non-negligent Manslaughter in the United States from 1960 to 2005
Violent Crime Rates other than Murder in the United States from 1960 to 2005
Property Crime Rates in the United States from 1960 to 2005
This drop in crime was particularly surprising because it occurred after some academics had predicted that the advent of “super-predators”—a rising generation of conscienceless violent youth—would soon lead to an explosion of crime.7 Also unexpectedly, the phenomenon was not confined to the U.S.—a remarkably similar pattern was evident in Canada, where violent crime rates peaked in 1992 and property crime and murder rates topped out in 1991. While the declines in violent crime and murder rates during the 1990s were smaller in Canada than in the United States, the fall in Canadian property crimes was slightly larger.8
Homicide Rate in Canada from 1961 to 2005
Violent Crime Rate Other Than Murder in Canada from 1961 to 2005
What’s more, the drop in crime may be even larger than these statistics indicate due to one factor that is often overlooked: rising rates of crime reporting by victims. With the exception of the data on murder, the crime data cited above—supplied by the FBI—are based on crimes that victims reported to police departments. But, of course, not every victim reports when a crime occurs. Victims are most likely to report two kinds of crimes: the most serious crimes and the ones that they believe have the greatest chance of being solved. These trends reinforce each other: victims report the most serious crimes, which in turn receive the most attention from police. This makes serious crimes more likely to be solved and thus even more reporting is likely to occur. Two simple diagrams show how the rate of reporting of violent crimes and property crimes in t
he U.S. has increased alongside rising arrest rates, suggesting that the increased arrest rates have encouraged more reporting.9
How Did the Rate That Violent Crimes are Reported Vary with the Rate That Violent Crimes Result in Arrest?
How Did the Rate that Property Crimes are Reported Vary with the Rate that Property Crimes Result in Arrest?
Less serious crimes, receiving less police attention, are not as likely to be solved and are also less likely to be reported. During 2005, victims reported only about 20 percent of larcenies to police, compared to an estimated 63 percent of rapes and 68 percent of aggravated assaults. Since more than half of all larcenies involve items worth less than $100, this should come as no surprise.10 Many people are deterred from reporting petty crimes by the inconvenience of dealing with the police and the slim chance of recovering the items.
So why is the rising rate of crime reporting important? Recall the example of the basketball referees: increasing the number of officials discouraged fouls, but it also made any given foul more likely to be noticed. The same effect is at play with the crime rate, which fell sharply in the 1990s even while the rate of reporting violent crime noticeably increased.
This trend indicates that anti-crime efforts in the 1990s were even more effective than is commonly believed. After accounting for the rise in crime reporting, we find that violent crime during the 1990s fell by about 10 percent more than the FBI’s statistics show. Property crimes fell by about 30 percent more. Furthermore, the rate of reporting of violent crimes surged by 46 percent from 1999 to 2002. This implies that official data on recent violent crime rates—which show crime rates leveling off after the drop of the 1990s—are particularly misleading. In reality, fewer crimes are occurring, but a higher percentage of them are being reported. 11
So what explains the fantastic plunge in crime rates during the 1990s? A lot of the individual pieces to this puzzle have been identified. Analysts have advanced a variety of plausible explanations, but it is not always clear how these fit together. Some stress law enforcement aspects such as increased arrest and conviction rates, longer prison sentences, “broken windows” police strategies, and the death penalty. Others emphasize different factors, including right-to-carry laws for concealed handguns, a strong economy, the waning of the crack cocaine epidemic, or affirmative action polices within police departments. It is even argued that legalized abortion has helped to stem crime. Many of these explanations may simultaneously be true, but there is lively debate about which factors are more important than others, and whether some policies actually did more harm than good.
So how do we evaluate the competing explanations? Before we identify the successful policies that reduced crime, let’s look at a few factors that had the opposite result.
What Increased Crime? Part I
Legalized Abortion
Out-of-wedlock births in the United States have climbed to an all-time high, accounting for nearly four in 10 babies born last year, government health officials said yesterday.
—Washington Post, November 200612
Nearly everyone agrees that the breakdown of families—to take one indicator, one-third of all births in the country and two-thirds of black births are now out of wedlock—is feeding into a destructive cycle of poverty, educational and developmental deficits, and incarceration.
—New York Times, July 200613
Of all the explanations for the drop in crime rates during the 1990s, perhaps the most controversial is its attribution to Roe v. Wade, the Supreme Court’s 1973 decision to mandate legalized abortion. The large number of women who began having abortions shortly after the Roe v. Wade decision were most likely unmarried, in their teens, or poor, the argument goes, and their children would have been “unwanted.” This indicates a high probability that these children, if born, would have grown up to be criminals. But because they were aborted, these children, who would have been teenagers entering their “criminal prime” in the early 1990s, were not around to commit the crimes expected of them. According to Freakonomics, abortion thereby became “one of the greatest crime-lowering factors in American history.” 14 An attention-grabbing theory, to be sure. But, as we shall see, a thorough analysis of abortion and crime statistics leads to a contrary conclusion—that abortion, in fact, increases crime.
Even before Roe v. Wade, supporters of abortion rights decried the crime and other social problems caused by “unwanted” children. Daniel Callahan summarized the argument in his 1970 book, Abortion: Law, Choice, and Morality: “To withhold the possibility of a safe and socially acceptable abortion for unmarried women is to start the chain of illegitimacy and despair that will continue to keep poverty, crime, and poor mental health high on the list of pressing social problems.”15
The argument was reiterated in 1972 by the Rockefeller Commission on Population and the American Future. Established by Richard Nixon, the Commission cited research purporting that the children of women denied an abortion “turned out to have been registered more often with psychiatric services, engaged in more antisocial and criminal behavior, and have been more dependent on public assistance.”16
The Commission appears to have been greatly influenced by a study published in 1966 by Hans Forssman and Inga Thuwe.17 The two studied the children of 188 women who were denied abortions from 1939 to 1941 at the only hospital in Gothenburg, Sweden. They compared these “unwanted” children to another group—the next children born after each of the unwanted children at the hospital. The study found that the unwanted children were much more likely to grow up in adverse conditions, such as having divorced parents or being raised in foster homes. They were also more likely to become delinquents and have trouble in school. Unfortunately, the authors never investigated whether the children’s unwantedness caused these problems, or were simply correlated with them. Perhaps a family’s poverty was the real cause of these dysfunctions, and women who sought abortions were more likely to be poor.
Nevertheless, the argument became axiomatic among supporters of legalized abortion. During the 1960s and 1970s, before Roe v. Wade, abortion rights advocates attributed all sorts of social ills, including crime and mental illness, to unwanted children.18 Furthermore, they found that “unwanted children are more likely to be abandoned, neglected and abused,”19 and they tend to be “poorly fed, poorly housed and poorly clothed.”20 Weeding these poor, crime-prone people out of the population through abortion was therefore presented as a beneficial deed that would make society safer.
More recently, two economists—John Donohue and Steven Levitt—became the first analysts since Forssman and Thuwe to attempt to present systematic evidence that abortion reduces crime.21 They argued that the drop in crime rates during the 1990s was primarily due to the increase in the availability of legal abortion in 1970—when abortion was deregulated in five states22—and especially in 1973—when Roe v. Wade deregulated abortion in the remaining states. The effect, they claimed, was staggeringly large—the pair attributed up to “one-half of the overall crime reduction” and up to 81 percent of the drop in murder rates from 1991 to 1997 to the rise in abortions in the early-to-mid 1970s.23 If accurate, they had surely found the Holy Grail for reducing crime.
The theoretical link between “unwanted” children and crime is simple and powerful. Most people who oppose this thesis argue from a moral perspective instead of trying to rebut the evidence. But whatever weight people put on moral arguments, the claim that abortion can prevent some murders and save lives causes at least some people to rethink their position. Unfortunately, the original arguments never acknowledged the possible pernicious effects of abortion on crime. And when we look at the data, we find that the argument doesn’t hold up empirically.
Let’s begin by looking at the overall status of abortion in the United States in the early 1970s. This is when Donohue and Levitt find that the “legalization” of abortion laid the foundation for the future drop in crime. It should be noted that, contrary to popular belief, there was not a b
lanket ban on legal abortion before the early 1970s. While closely regulated, the procedure was legal in various circumstances, such as when the life or health of the mother was at risk. In some states, doctors interpreted this clause quite liberally. As a result, legal abortion was much more widespread before Roe v. Wade than is commonly acknowledged. In fact, in 1970-1973, when abortion was “legal” in five states but “banned” in the rest, some of the “banned” states had similar or even higher rates of legal abortion than in the “legal” states. For example, Kansas had 277 abortions per 1,000 live births in 1971, outstripping “legal” states such as Alaska (160), Hawaii (261), and Washington (265). High rates of abortion could be found in other “banned” states and districts, such as Washington, D.C. (703), New Mexico (219), and Oregon (206).24 Donahue and Levitt, whose main results mistakenly assumed that no legal abortions occurred in any of the “banned” states before 1973, thus began their study with flawed statistics.25
A central problem with the “abortion reduces crime” thesis is that it conflates and blurs two different arguments. The first contention is that aborted children would have been more likely to cause crime specifically because they’re unwanted—since their parents did not wish to have them, they would likely have grown up in an unloving household, and would have therefore been more prone to crime. A separate, less savory explanation is that abortion reduces crime by culling out certain demographic groups that commit disproportionate numbers of crime, for example, young African American males. If abortion really reduces crime, then this “eugenics” effect could be highly significant, since African Americans have an abortion rate that has consistently been about three times that of whites and a murder rate about 6.5 times that of whites.26
Freedomnomics: Why the Free Market Works and Other Half-Baked Theories Don't Page 12