Just Freaky


Freakonomics (William Morrow, 2005) by Steven D. Levitt and Stephen J. Dubner has been reviewed several times on (1, 2, 3), but there are some important points to add.

Levitt and his coauthor Stephen Dubner have written an interesting and informative book and the title is all too accurate. Readers will be exposed to precious little economics, only the statistical unraveling of a few freaky puzzles. This is not to suggest, as the book does in a number of places, that the material or subject matter is all trivial. The problems of crime and African Americana are among the most vexing ones facing America and other nations around the world.

Austrian economists will quickly notice the methodological flaws of this book. For Levitt and the economics profession in general, science is measurement. Ordinary readers may accept this notion of economic science, especially when it comes to the production of physical, tangible, and seemingly homogeneous goods, but hopefully many will recognize that “measurement” is fraught with problems, especially when it comes to the new world of information technology, the production of culture, and the ever-changing basis of what underlies consumers’ satisfaction — the changing tastes and preferences of people.

The book itself hints at the limitations of quantitative measurement by noting that very small causes can have enormous effects, and the authors even provide a good example of this with the case of a private citizen who non-violently brought the KKK to its knees. It also uses several examples where the data collected to measure various phenomena have been erroneous.

For example, public school teachers cheated on their students’ standardized tests to gain economic rewards and avoid punishments (Levitt caught them by using a computer algorithm — not economics) and he also discovered that the City of Atlanta Police Department underreported crime in order to win the City’s bid for the 1996 Summer Olympics.

However, when it comes to Levitt’s novel suggestion that the dramatic drop in crime during the late 1990s was caused by Row v. Wade, the potential of bad data is dismissed. Based on all the reports I have heard about people not wanting to report crimes to the police and for police to under-report crime, I find this dismissal hard to accept.

For example, many of the hardest-hit areas of New Orleans were considered “cop-free” before Hurricane Katrina hit. Large areas of urban America are controlled by drug gangs and private property owners are responsible for their own security. Levitt can respond that certain types of crime — specifically murder — are almost always reported even in the deepest corners of the urban welfare jungle because bodies have to be disposed of somehow. Indeed, the number of murders per capita has been declining at a noticeable rate in recent years.

The wrinkle in this argument is that the murder rate has taken a statistical nosedive not because of the dramatic decline in the number of attempted murders, but because of a more dramatic decline in the number of successful attempted murders. The big reason for this is that emergency medical services have been improved and are able to respond quicker (e.g., 911) and they can respond with more effective life-sustaining technology, including drugs, life-saving devices and procedures and because ambulances now have direct communication and patient data transmission capabilities with emergency room service providers. In other words, entrepreneurs are reducing the success rate of murderers by introducing new life-saving products. Combine all this with the more widespread knowledge of CPR and knowledge of how to reduce profuse blood loss and the decline in the murder rate is a much less encouraging statistic.

The other reasons behind the reduction in crime are also not encouraging. Levitt finds that more police and more people in prisons are also reasons for lower crime rates. The increase of police and prisoners behind bars make it obvious that we continue to bear a heavy cost of crime. The cost of crime has not been reduced by the state, it has only been spread out a little more evenly over taxpayers and property owners. Levitt also ignores the positive role that gun ownership has in reducing crime.

The true economic content of Freakonomics is largely limited to mere economic catch-phrases like “incentives matter.” This was most glaring to this reviewer in the chapter that discussed the war on drugs and the problem of crack cocaine. What explains the emergence of crack? It turns out — according to the book — that some people just happened to invent crack, it was cheap and it became popular. In other words, just a vague generalization about incentives, entrepreneurship and technological change — no real economic explanation — crack just happened.

Crack cocaine is described as the worse thing to afflict the African American community since slavery, but there is no explanation of what caused it, what made it spread, and what made it take hold in America — the key questions to be answered. Instead the book tells us that street drug dealers are poor and face a high probability of violent death — things that are regularly reported on in the New York Times and other sources. I have shown that the reason for higher potency and more dangerous drugs is stricter prohibition enforcement.

Other trivial facts that we learn in the book include: not everyone pays for bagels on the honor system, people lie about themselves on dating websites, and that some people are phony about their open-mindedness about race-related issues. We also learn that Sumo wrestlers who have already qualified for championship rewards do not perform “as expected” later in tournaments when faced with opponents who have not qualified and face elimination from the tournament.

Another well-known point raised in the book is that professionals will often cheat their clients if the proper incentives and opportunities are present. Most people understand that the incentives of licensed professionals do not necessarily line up with their clients (e.g., real estate agents and doctors). The key question is "why this is the case." It might just be a small misalignment of incentives in the market economy, but it could also be a persistent one that has a big overall economic impact. What is the answer? Levitt does not provide the obvious answer that professional licensing provides a monopoly privilege and a barrier to entry that prevents real competition from reducing price, increasing the quality and diversity of service, and better monitoring of corrupt professional practices.

Ironically, Levitt’s true strength lies not in his much-acclaimed computer algorithms and regression results but in his admitted disinterest and marginal abilities in the mechanical approach of modern mainstream economics (math and stat). These so-called weaknesses are no doubt the key to his creativity and innovative thinking, however misguided, as most graduate programs rely almost entirely on math and statistics to beat all the creativity out of young scholars in order to produce more automatons of economic orthodoxy.