The class in the room before mine is a political science class. As the door was open last week I caught the tail end of a lecture on the death penalty. It ended with the professor saying that he didn’t believe in the death penalty because, based on what he’s read, it’s not effective and it’s cheaper to keep someone alive because execution appeals are expensive. To back up his reasoning, he cited that Texas has the death penalty and also has really high murder rates.

I had two problems with his arguments. If he had just said that, based on the research he could find, it is not effective at deterring murders, he might have had a point. The literature is mixed and, from what I can gather, doesn’t clearly show that the death penalty serves as a deterrent. Some proponents of the death penalty don’t care about this and support it more for its retributive aspects, arguing that if you kill a bunch of people you simply deserve to die. But his example of Texas was cited as evidence of the fact that states with the death penalty have high murder rates, so clearly the death penalty isn’t working.

Using this professor’s logic, one could also say that since the unemployment rate increased despite the stimulus bill, the stimulus bill didn’t work. When Republicans say that, they’re chastised as being idiots. And they probably should be, since it’s an argument not supported by facts, and the stimulus bill did keep unemployment from rising — albeit at the cost of rising debt. Whether it added between 1.5 and 3 million jobs, as the CBO says, or a half a million, as economist Ray Fair argues, is debatable and depends on assumptions of each model. We’ll never know. But it did increase output and employment. You can’t look at the fact that unemployment rate rose and conclude that the stimulus bill didn’t work any more than you can look at a state that has a high crime rate and conclude that one specific aspect of that state (that it has the death penalty) has no impact on crime or murder, ignoring every other thing that might affect the crime rate in that state (its other laws, its economy, its proximity to Mexico, etc.).

This professor and I are cordial with each other, so I decided I’d discuss his statement with him and give him a similar analogy based on what he said. I told him that, by his logic, you could also argue that cities that have more police officers often have more crime, so clearly law enforcement measures don’t work and we should spend less money on police. He backtracked quickly from his first argument and then retreated to the “it’s more expensive to execute” argument. Personally, I agree with that position and, as an economist who a) doesn’t see much deterrent effect and b) doesn’t much care for eye-for-an-eye retribution, I’m against the death penalty, too. But his students didn’t hear that and now probably think that because Texas has a high crime rate, the death penalty doesn’t work.

As professors, we’re supposed to be fostering our students’ ability to use their critical thinking skills. In economics, we spend a lot of time on the difference between correlation and causation, and how difficult it can sometimes be to prove causation. We’re supposed to model that kind of analysis for them so they can use their own brain to decipher specious arguments. If you think the death penalty is too expensive, fine — I agree with you. If you think it doesn’t deter other murders, cite the research. But it’s irresponsible to cite one data point as evidence of your conclusion when, as all social scientists know, a study that proves causation requires much more than that.