Friday, May 21, 2010

Come here, my pretties . . .

Twice in the past week I've seen the same issue pop up - the alleged disparity in the criminal justice system based on whether someone is good looking or not.

First, fellow Texas blogger Grits for Breakfast addressed the issue on Wednesday, quoting a CBS story that cited a Cornell University study claiming that: "unattractive defendants are 22 percent more likely to be convicted than good-looking ones. And the unattractive also get slapped with harsher sentences - an average of 22 months longer in prison."

A lot of 22s there, friends, and that's a real specific number for something so imprecise.

I saw it again, reading a book review of The Beauty Bias (written by an ugly person, apparently) in the Christian Science Monitor.

Grits rightly questions the use of undergrads in the Cornell study but says, "one imagines it might be possible to devise a study that tested this finding empirically."

I doubt it, myself, but he's a smarter man than I am, so I read on. He suggests this:

"Take a subset of cases that went all the way to a jury. Rank defendant mug shots by attractiveness (perhaps using focus groups). Then chart the outcomes, grouping similar types of cases, for the homely and handsome alike, following up with juror interviews to round out the dataset with oral accounts."

Sounds like a decent idea, to me, in principle. Certainly it would be fascinating to know how much attractiveness plays a role in jury trials.

But here's the thing. Just because we (a) want to know something, and (b) live in a scientific world, that doesn't make knowing that thing a scientific possibility.

And here's why not, in this particular instance.

First, you have the problem of nailing down what is "attractive." Without knowing what that means, an answer is impossible. I suppose, on the left and right of the bell curve it's easy enough (Lyle Lovett v. Brad Pitt, for example) but it gets very hard from there on in. Grits makes a decent enough suggestion ("Rank defendant mug shots by attractiveness (perhaps using focus groups")) but at the end of the day you are left with a puddle of subjective opinions that forms a shape on the floor, and calling it immutable doesn't make it so. I mean, anyone ever pick a mate using a focus group? Remember that in a focus group you can have twenty people thinking a dude is hot, and one who doesn't. But in a jury it only takes one person to spoil the party. Just one.

You next have the larger problem of comparing cases, and this I would suggest (and have mentioned before in posts) is impossible. Within one "case" the following things are never identical:

-- the victim's role (provocateur, total innocent, etc)
-- the impact on the victim (sensitive soul, tough guy, etc)
-- the degree of harm (broken bone, snapped spine, amputated leg, etc)
-- the degree of intent to cause the harm (reckless, knowing, intentional, etc)
-- the defendant's criminal history (dude is always breaking people's spines, or did it just once, etc.)
-- the degree to which the defendant takes responsibility (pleads guilty but goes to jury for punishment, pleads not guilty, pleads not guilty and testifies to blame victim, etc)
-- likelihood of rehabilitation for defendant (smart guy with family support, not that, etc.)
-- the whiteness of the defendant's teeth (just kidding! Some don't even have teeth.)

See, with all these jelly-like variables being used to assess the impact of a jelly-like variable called "attractiveness" you are simply wasting time.

And please, feel free to waste time. Yours, of course, not mine.

Look, I wouldn't dispute the claim that in general attractive people often have an advantage over ugly mugs. But I don't think that means it's true in the criminal justice system -- the jurors I have always spoken to have taken their roles very, very seriously, looking at the evidence and not injecting their opinions as to irrelevancies like the size of someone's boobs or how nice their hair is. They don't do that for the women, either.

I suppose my point is that there are some absolute, unarguable, definitive problems with our justice system that could be studied and addressed. The quality of defense lawyers appointed to represent the indigent. The dearth of treatment and education programs for offenders. Low pay for dedicated (and very handsome) prosecutors. ADAs forced to look out the window at jail cells all day. The dearth of on-site, free massages for county employees.

You know, that kind of stuff.


  1. Well this isn’t really a place to get into a discussion of technological details, but mug shots can be objectively categorized with a set of mathematically rigorous descriptors. These methods were initially developed at MIT and are referenced in general by the term “eigenfaces” because they are based on eigenvectors of pixel intensity matrices and principal component analysis. Details are outlined at for the benefit of any who may be curious.

    The problem is that “attractiveness” is subjective and cultural. For example in Tonga, where the monarch is massive, and probably in other places where food is scarce, obesity is commonly considered attractive. There have been several psychological studies using focus groups where the purpose is to establish correlations of subjective perceived attractiveness with objective mathematical eigenvalue distribution patterns (just google “eigenfaces” and “attractiveness” simultaneously to get links).

    In any case, the subjective nature of “attractiveness” could be bypassed by using objective eigenface metrics as a proxy when assessing correlations of jury conviction and sentencing outcomes.

  2. Mirror, mirror on the wall, who's the fairest juror of all?

    "Beware my Queen," the mirror replied, "You're ugly as sin and they think that you lied!"

    Those of us at the Bar who will not ever be mistaken for Pierce Brosnan can't help but wonder if we too get the short end of the deal from jurors.

  3. Naturally, those of us who benefit from the beauty bias tend to be less critical of it. :) But I still think it's potentially quantifiable. IMO the selection of comparable cases is a much greater barrier than rating attractiveness. And there are some categories of cases - DWI and drug cases come to mind - where the fact circumstances are similar enough and the cases numerous enough that it might be possible to control for most variables. It's not how I plan to spend my summer vacation, but if academics are going to study the topic anyway, there's got to be a better way than interviewing undergraduates.

    Also, I regret I can't immediately lay my hands on it, but I read some research on the SSRN last year disputing the meme that "in a jury it only takes one person to spoil the party." A ballot by ballot study of jury behavior found that's technically true but in practice rare. Though I think the research was on capital juries (which may not be 100% transferrable results), the gist was that if on the first ballot only one or two jurors say not guilty, the other jurors usually end up later convincing them. They found it was not unheard of but extremely rare for a lone holdout on the first ballot to remain a holdout. In most cases jurors in the minority don't withstand the peer pressure unless there's a critical mass of 4-5 for 'not guilty' on the first ballot, as I recall, and then often it ends up flipping the other way.


Comments posted to this blog are NOT the opinion of the Travis County D.A.'s office, under any circumstances. They are only the personal, non-representative opinion of D.A. Confidential if posted under his name.
I welcome all comments, as long as they are expressed with politeness and respect. I will delete all comments that I deem to be personal attacks, or that are posted merely to antagonize or insult.