Writing Well Is the Best Revenge


For the first years of its existence, radio seemed like an economic loser without government subsidy. No one could figure out how to make listeners pay, and, consequently, radio hosts and entertainers usually worked for free. In 1922, Herbert Hoover, then Secretary of Commerce, asserted, "Nor do I believe there is any practical method of payment from the listeners." But that same year, AT&T discovered that it could make money by selling ad time on radio. It sounds obvious now, but it wasn’t then. After that, radio thrived and is still thriving.

John Lott tells that story and many others — and tells them well — in his latest book, Freedomnomics. Although the title is clunky — it’s obviously a version of the more famous title Freakonomics — that’s all that’s clunky. Lott’s book is an analytic tour de force as he walks the reader through the economics of issues as varied as the origin and development of American public schools, why government grows, crime, gun control, political influence, gasoline prices, the price of wine at restaurants, and voting machines. In each case, he explains concepts that most people don’t understand but want to understand — concepts that help us figure out our world. And, in the process, he makes the case — summed up by the book’s subtitle — for "Why the Free Market Works and Other Half-Baked Theories Don’t."

Many of the half-baked theories Lott refers to are those produced by his rival and lead author of Freakanomics, University of Chicago economist Steven Levitt. In case you haven’t heard, Levitt and Lott are engaged in a feud right now. Lott alleges that Levitt defamed him in Freakanomics and sued to enjoin Levitt and his publisher, HarperCollins, from selling additional copies of the book until they had deleted the alleged defamatory statement. This part of the suit has already been thrown out, but what remains to be tried is Lott’s claim that Levitt defamed him in a private e-mail to an economics professor named John McCall. Lott, in a July interview, told me that Levitt had sued Lott’s publisher for using the title Freedomnomics and that the publisher agreed to put on the cover, "A Rebuttal to Freakanomics and More." I’m not an expert on defamation, but as the late economist Murray Rothbard argued, no one has a right to a reputation — that is, to be thought of a certain way by people. I’m inclined to agree with Rothbard. Lott and Levitt should settle this on the field and not in the courts.

And Lott does settle it on the field. While Freakanomics is slick, Freedomnomics is much more solid analytically. Freakanomics reads as if written by a journalist who knew some, but not enough, economics. Freedomnomics was written by a first-rate economist who knows a ton of economics and who, by the way, also writes well. Writing well — and knowing your subject matter — is the best revenge.

Take one of the big issues that divide the two authors: why violent crime fell in the 1990s. One controversial explanation, which Levitt and journalist Stephen Dubner push in Freakonomics, is that liberalized abortion in the early 1970s caused there to be fewer unwanted babies. Their idea was that a disproportionately high percentage of these babies, had they been born, would have committed crimes. But Lott systematically takes apart their argument. He points out that in their data analysis, Levitt and his academic co-author, John Donohue, classified as "banned" states many states in which abortion was not banned. Rather, abortion was legal when the life or health of the mother was threatened. In some states, notes Lott, doctors interpreted this provision quite liberally. In some states that Levitt and Donohue classified as "banned," abortion rates were even higher than in some states classified as "legal." Writes Lott, "Donahue [sic] and Levitt, whose main results assumed that no legal abortions occurred in any of the u2018banned’ states before 1973, thus began their study with flawed statistics."

Furthermore, Lott argues that increased abortion leads to increased crime. How so? The availability of abortion reduces the "cost" of sex because the downside — an unwanted pregnancy — becomes easier to eliminate. With this comes an increase in pregnancies, but what if the woman decides not to have an abortion? The man might feel tricked into unwanted fatherhood and refuse to marry her. The "shotgun marriage" that I often heard about in the early 1960s seems to have pretty much disappeared. The result, then, is a child who is raised by an unmarried mother. As is well known, children raised out of wedlock score high on pretty much every social problem, including crime. So there are two offsetting effects of abortion on crime: the fewer unwanted children effect and the more children out of wedlock effect. Which one dominates? Lott cites his own work and work by Boston Federal Reserve economists Christopher Foote and Christopher Goetz, both of which find that the latter effect dominates: more abortion leads to more crime. The Economist magazine called attention to Foote’s and Goetz’s work, which caught an error in Donohue’s and Levitt’s work. In an article titled "Oops-onomics," The Economist pointed out that in running their test, Donohue and Levitt made a crucial computer coding error.

Moreover, notes Lott, if the Roe v. Wade liberalization of abortion really was responsible for the later reduction in crime, one should see a lower rate of crime among those born after Roe v. Wade than among those born before. But Lott presents a graph showing that at each age, the exact opposite is the case. That is, the murder rate among those born in the four years after Roe v. Wade was substantially above the murder rate among those born before. Interestingly, all this evidence was available and well-known long before Levitt and Dubner wrote Freakanomics.

So what did cause the 1990s’ reduction in crime? Lott attributes the reduction, in part, to the death penalty. He points out that the Supreme Court suspended the death penalty nationwide between 1968 and 1976 and that during that time murder rates skyrocketed. The 38 states that reinstituted the death penalty after 1976 had a 38-percent larger drop in murder rates by 1998 than states that did not reinstitute it. Interestingly, executions increased significantly during the 1990s, which was when violent crime fell. Also important, writes Lott, were increases in arrest rates and conviction rates and the increased passage during the 1990s of laws allowing people to carry concealed handguns. Lott also points out that his view of concealed carry is controversial and gives a table listing studies on each side of the issue. A look at the footnotes here is worthwhile because he goes through the back and forth between his work and that of Yale economist Ian Ayres and the aforementioned John Donohue.

Lott has also done some of the path-breaking academic work on campaign finance reform and has a nice section on that in his book. He argues that campaign finance reform has made it much more difficult to have an insurgent campaign. Antiwar candidate Eugene McCarthy, he argues, would never have been able to accomplish his strong showing against President Lyndon Johnson in New Hampshire if the current limits on individual contributions to candidates had existed in the 1960s. McCarthy relied heavily on just six big donors and raised almost as much money, adjusted for inflation, as George W. Bush did from 170,000 donors by the time of the first 2000 primaries.

Lott also addresses education. Have you ever wondered why so-called public schools — that is, schools funded by taxes and run by government — have a geographic monopoly? Lott tells why and it’s likely to upset you. It had nothing to do with school-aged youths not getting enough education in the private sector, as the late education economist E.G. West has pointed out. Rather, in the early 19th century, state governments in New York and elsewhere were concerned that Catholic families were sending their kids to — are you sitting down? — Catholic schools. Oh no! We can’t have that, thought the government. In the legislative debates at the time, writes Lott, various politicians expressed the view that Catholic values would lead the students to a life of crime. So, to instill its beloved Protestant values, the government started subsidizing Protestant schools as a way of drawing customers away from the non-government-subsidized Catholic schools. It worked, kind of. Protestant schools were now competing with each other for these customers. How do you compete? By giving the customer what he wants. Thus, they focused on reading, writing, and math and played down Protestant religious training.

What could the government do? If competition gives you a result you don’t like, it’s obvious what to do: kill competition. So the government started to give subsidies only to the approved Protestant school nearest a student’s home, thus creating a subsidized geographic monopoly. When the subsidies got large enough, which they did in New York in 1867, the government simply took over the subsidized schools.

Robert Higgs has written that government grew because it acquired powers during crises in the 20th century that it did not completely relinquish when the crises end, thus creating a ratchet effect. Higgs points to the two world wars and the Great Depression as the three main government-growing crises of the century. But Lott points to a different explanation that he supports with evidence: giving women the vote. He notes that women — especially single women or married women who expect to be divorced — have a different voting pattern than men. They tend to favor government attempts to make people more secure with various subsidies and regulations. (Although, as I point out in "Big Government — Big Risk," government actually makes people less secure, but this is not well understood.) So, women’s suffrage, writes Lott, actually led to the growth of government. Because women received the vote at the state level in various years, the data allow him to track what happens to government spending and taxes as states allowed women to vote. He finds a consistent pattern: within 11 years of giving women the vote, states’ real per capita spending had doubled.

Lott also takes on the issue of gasoline prices. A standard lesson we economics professors teach students is that price controls designed to eliminate "price gouging" cause shortages and line-ups and reduce gasoline sellers’ incentive to save inventory for times of increased scarcity. Lott lays this out nicely, but he does more. He also shows why the difference between the full-serve price and the self-serve price for gasoline is smaller for premium gasoline than for regular gasoline. For both regular and premium gas, he notes, service stations charge extra for service. But buyers of premium gasoline tend to buy more gas at a time than buyers of regular gasoline. So the lower increment in price per gallon of premium gasoline times the higher number of gallons purchased makes the overall charges for service roughly equal.

When it comes to judging Lott’s work, I should admit my bias. John Lott is a fellow graduate of the Ph.D. program at UCLA in its glory years from the early 1970s to the mid-1980s; I graduated in 1976, he in 1984. And I’ve have found that virtually everything he writes about government policy makes sense. I don’t know enough about the data to judge his work on concealed carry laws, which has been the work most challenged by other academics. It’s simply that his argument makes sense. The only failing I can find in the book is his reasoning about public school teachers. He argues that public school teachers have an incentive to teach that government solves problems, despite the evidence, because they "have an abiding personal interest in perpetuating the growth of government." I agree with him about the empirical fact he is trying to explain: virtually all of my and my daughter’s public school teachers, to the extent the issue came up, favored more government. But I’m not sure this was in their interest. The larger government becomes, the less will it be available to pay the absurdly high pensions granted to teachers. If I were a public school teacher just looking out for my own interest, I would favor smaller government in order to maximize the probability that the government could afford to pay my pension. Of course, for someone to see this, he or she would need a detailed understanding of the budget numbers. My own view is that there is self-selection: people who want to teach in a government setting tend to be people who like government.

Political Theatre

LRC Blog

LRC Podcasts