Lies, damn lies, and statistics

I listened to the radio a lot over recent days because of a flu bug that discouraged me from doing anything but resting in bed! One of the pundits I heard was Michael Smerconish, a conservative radio talk show host and attorney. He was crowing about a recent article in the NY Times regarding the death penalty and its presumed deterrent effect. (http://www.nytimes.com/2007/11/18/us/18deter.html) Smerconish argues that if the (in his view) left-leaning and therefore anti-death penalty NY Times would publish data showing that the number of homicides fall as the number of executions increases, then the numbers must be true. He also argues that the study now makes it impossible for anyone to mount a rational argument against the death penalty because if 3 to 18 lives are saved (or possibly more if other studies are also correct), then we are bound to use it because of the overall saving of lives.

Saving for the moment arguments based strictly on morality, let’s look at the data presented. First, the Times article notes that “Canada has executed no one since 1962. Yet the murder rates in the United States and Canada have moved in close parallel since then, including before, during and after the four-year death penalty moratorium in the United States in the 1970s.” Given this fact, one could argue that the change in rates is based on some other variable. I’m sure there are also comparisons that could (should?) be done with other countries without a death penalty option. Secondly, we aren’t talking about a large data set; there have not been that many executions to track.

The statement Smerconish made that most troubled me, though, was his response to a caller who questioned how many innocent persons may have been wrongly executed. His reply was that there was no proven case of an innocent person being executed. This response would be laughable were it not so sad. He assumes a 100% accuracy rate in application of executions, and there is nothing associated with humanity that functions at 100% accuracy. Fallible lawyers, judges and juries are bound to have made errors, whether we can prove it, or not. Studies have shown that the criminal justice system is skewed based on race, ethnicity and class, so poor minority prisoners who were executed would not have had the resources nor been in the spotlight that would attract persons desirous of proving wrongful execution.

The most persuasive argument to me, however, is that from scripture, where Jesus never advocates the death of anyone he encounters, no matter their offense, and actively works against the stoning of a woman accused of adultery. The Sermon on the Mount is clearly anti-violence. Tony Campolo’s words ring again in my ears: What if Jesus meant what he said?