An Innocent(?) Man

The Columbia Law Review is set to publish an article alleging that Texas executed an innocent man in 1989, Carlos DeLuna. He was convicted of the murder by stabbing of a women in a bar, but it's claimed that another Carlos (who looked similar) had committed the crime.

This isn't the first time that this has been alleged. There are claims that Cameron Todd Willingham was also wrongly executed for the arson murder of his children, in 2004. There, the allegation is that the expert evidence used to claim that the fire was deliberately started was mainly pseudoscience nonsense.

Jason Kottke linked to both of them. Both of them make for disturbing reading. I ended up reading them in reverse order, and it reinforced that there are actually two questions:

1. Has an innocent person been executed by capital punishment in recent times?

2. Will this fact likely be established beyond sufficient doubt such that public opinion believes that an innocent person was executed by capital punishment?

In terms of the first question, (particularly from reading the Willingham case) it seems likely to me that arson-based murder is a very likely candidate.

Think about it. For the vast majority of murders, there's not much doubt that the person was, in fact, murdered - the issue is mainly by whom. For arson, on the other hand, a large part of the determination is whether the fire was an accident or was deliberately lit. A guy that was nasty to his kids, and woke up one day to find his house on fire, looks a lot like an arsonist on paper. The main way to tell them apart is based on the scientific evidence about whether the fire was deliberately lit.

And from everything in the Willingham case, it's hard to avoid the conclusion that large amounts of this are junk science, folklore, and other kinds of nonsense. The field seems to be improving, so hopefully there's less chance of this kind of thing happening in the future.

Still, how much confidence does this give you about all the previous capital punishment for arson cases? They were very likely decided by the same kind of junk science. This makes it more likely that someone will have been executed wrongly.

But, and here's the rub, I imagine that arson cases are less likely to satisfy question #2. Precisely because it's hard to say that a fire decades ago was deliberately lit, it's also hard to say that it wasn't deliberately lit. Without this extra piece, the average person gets to the conclusion that maybe the guy might not have done it, but they aren't sure either way. Since people don't think emotionally in terms of probabilities (there's a 50% chance I should be outraged, and a 50% chance that justice was done, albeit by a shonky method), this doesn't get people fired up. Even in the Willingham case, I come to the conclusion that it certainly doesn't convince me beyond a reasonable doubt that the fire was deliberately lit (which is kind of the point). Which means he shouldn't have been convicted if I were on the jury. But there's a difference between failing to establish guilt (which is what courts determine) and establishing innocence (which, like it or not, is what death penalty opponents have to do).

Which is where the Carlos DeLuna case comes in. There, there is quite strong evidence of a specific alternative murderer, along with enormous evidence of a shockingly bungled case. So it's not just that he wasn't guilty beyond a reasonable doubt - death penalty opponents can also make an affirmative case that someone else was the murderer.

Without reading all this stuff in detail, I never quite know what to make of these cases by advocates that a particular person is innocent. I'm open to being convinced, but there seem to be a lot of false positives - death penalty opponents keep choosing pinup boys like Mumia Abu-Jamal where, despite the evidence that the investigation was not done well, it also seems quite likely that the guy did in fact kill the cop. Death penalty opponents would seem advised to be careful about crying wolf, because even potentially sympathetic audiences can start to feel like every new claim is just another beatup.

Personally, I'm a Bayesian - probabilities are never zero, and no system is failsafe. Done enough times, eventually someone will be wrongly executed. When this happens, it will be a travesty, but this doesn't turn into a categorial imperative against the death penalty, any more than the death of a single soldier provides a categorial imperative against war. Death penalty advocates are very reluctant to make the claim that a very small probability of a wrongful execution may still be worth it to secure justice for the rest.

The real mystery to me is why people focus only on death penalty cases as the source of outrage. When the state kills an innocent person after years of careful legal deliberation, all hell breaks loose. But when the state kills an innocent person immediately in a rash and shameful manner, people largely just yawn.