Small Lies, Big Lies, and Economists

The timing could not be more inappropriate. The United States is finally moving to join all the civilized nations of the world by ceasing to execute people. It still has quite a way to go. So far, most states that still impose the death penalty have recently suspended executions until the Supreme Court rules on whether the ways people are currently 'put down' constitutes a cruel and unusual punishment; that is, whether lethal injection violates the constitution.

Enter a bunch of economists with a model that, they say, shows that the death penalty deters crimes and hence saves lives. Because most people do not read the math involved and have a hard time following the intricacies of these models, these kinds of "findings" are often taken seriously. Indeed, even a major progressive legal scholar, Cass Sunstein, told the press that "the evidence on whether [capital punishment] has a significant deterrent effect seems sufficiently plausible that the moral issue becomes a difficult one."

Actually the data on which this model is based are extremely thin. First, there are simply not enough executions for most statistical methods to work properly; in 2003, while there were more than 16,000 homicides nationwide, there were only 65 executions. Adding to this statistical shortfall, too many other factors change over the same period as the number of executions changes to permit a sound conclusion. These include the overall crime rates, policing methods, gun laws, levels of economic growth and of upward mobility, drug rehabilitation and education policy, among others. Things are complicated even further when one recognizes that these factors differ not only over time but also among various states that execute more people than others. The economists gloss over such challenges by using dummy variables, artificial weights, and other slights of hand.

But all that the economists ultimately demonstrate is some (deeply questionable) correlation between capital punishment and murder. That is, they show that as the number of executions increased, homicide rates decreased. It doesn't take a doctorate in economics to know that correlation does not prove causality. As Cassandra Stubbs of the ACLU wrote in a recent letter to the editor of the Wall Street Journal, one might argue that "children with larger shoe sizes perform better in reading tests." And when one notes, as Professor Ronald Allen of the Northwestern University School of Law has, that during the time-period in question the homicide rate fell more or less nationwide--that is, not merely in states with capital punishment--one wonders whether these economists are not doing some shoe-gazing of their own.

Finally, there are some measures that civilized societies just don't employ, whether they are efficient or not. And if these measures are used, the societies at issue should be profoundly shamed rather than comforted with pseudo-scientific mumbo jumbo. Good societies do not hang people, do not torture people, and in select key areas, they allow their values to take precedent over all other considerations. After all, if our legal system followed the cost-benefit analysis of economics, we would shoot all young, first time offenders--because we know they are very likely to offend again. Not to mention that the costs of incarceration are higher than many college tuition rates; some $30,000 per year, per inmate. A bullet and cremation can be readily had for a few bucks.

The death penalty, especially given how often innocent people have been executed, belongs on this--granted short--list of no-nos. It is this taboo which I urge the United States to finally embrace, and also which is the only reason I oppose hanging a few economists in order to deter the others from building such deadly models.


Amitai Etzioni is University Professor at the George Washington University and author of The Moral Dimension: Toward a New Economics.