Posts Tagged ‘profiling’

Profiling Doesn’t Work? More Profiling!

Saturday, April 23rd, 2011

When we were just starting out in the law, we frankly had no problem with the concept of profiling.  Not racial profiling — that’s just a logical absurdity along the lines of “most people who commit crime X are of race Y, therefore it’s reasonable to suspect people of race Y of committing crime X.”  We’re talking about profiling as the concept that a significant number of people who commit crime X exhibit the combination of traits A, B and C, which is a combination rarely encountered otherwise, and therefore if one were to look for people exhibiting traits A, B and C, then one might have a better chance of catching someone guilty of crime X.

Intuitively, this sounds reasonable.  If we were to know, for example, that certain serial arsonists are motivated by a sexual mania, that these arsonists tend to remain near the scene to masturbate or so they can masturbate to the memory later, that they tend to have spotty work and relationship histories, and that they tend to have crappy cars — well then, there’s nothing wrong in letting the cops scan the crowd of spectators at a fire, question any who seem to be getting a kick out of it, and investigate those who are single, unemployed, and drive a beater.  (This is an actual profile, by the way.  We didn’t make this up.)

And emotionally, profiling sounds wonderful.  Catching a psychopath is often difficult, because they don’t play by the same rules as the rest of us.  Wouldn’t it be nice if there were some, er, rules that we could follow — a formula of some kind — that would make it easier to identify and catch them?

As we said, in our early years we thought this was a great concept.  Whenever we encountered some findings that certain traits had been identified with this type of serial killer, or that type of terrorist, we thought it was fantastic.  But we didn’t think too critically about it.  And for sure we never bothered to look for the underlying data, much less examine the methodology used to determine how strongly these traits correlated with perpetrators of that crime.

The problem is, nobody else was doing that, either.

Profiling only works if the profile is accurate.  That should go without saying.  But it has become plain over the years that the various profiles out there are not accurate.  They are not based on actual data, but instead only on anecdotes.  (And as we like to say, the plural of “anecdote” is not “data.”)  Nor are these profiles based on any significant sample size.  No profiling study ever did even a simple regression analysis to determine whether any particular trait happens to be a meaningful variable.

We figured this out soon enough, of course.  After our first couple of years with the DA’s office, we were already joking about the silliness of profiles.  It was almost a party game to figure out which psychopathic profile we and our friends happened to fit (secure in the knowledge that hardly any of us were really psychopaths).

And the rest of the world soon caught on.  The Onion did a piece entitled “Crime Reporter Finds Way of Linking Warehouse Fire to Depraved Sex Act.”  Malcolm Gladwell wrote an outstanding piece in 2007 called “Dangerous Minds: Criminal Profiling Made Easy,” in which he solidly debunked the whole profiling scam, showing how there’s no science or statistics behind it, and even the data it’s based on is mostly useless.

It’s now fairly common knowledge that criminal profiling is about as useful as a Tarot deck.  So of course the FBI has stopped using it, right?

-=-=-=-=-

Wrong.

As a matter of fact, they’re expanding!  Just as the feds have (disastrously) tried to use street-crime investigative techniques like wiretaps to go after white-collar offenders, they are now (equally idiotically) starting to use criminal profiling to go after people for white-collar offenses.

Matthew Goldstein wrote an excellent piece on this for Reuters this week, called “From Hannibal Lecter to Bernie Madoff: FBI profilers famous for tracking serial killers are turning their attention to white collar felons.”  This (and the Gladwell piece linked to above) should be required reading for any white-collar defense lawyer now practicing.  When the Galleon case first came down, we were one of a handful of people doing white-collar defense who also had plenty of wiretap experience; now, of course, more of us are learning it the hard way.  Hopefully, with this new profiling issue, more of us will be prepared to deal with it ahead of time.  (And perhaps even nip it in the bud.  Like Barney Fife, we’re a big fan of bud-nipping.)

The agents in the FBI’s Behavioral Analysis Unit are the ones who profile serial killers and the like.  “The hope is,” reports Goldstein, that they “can get into the minds of fraudsters and see what makes them tick.”

“This originally started out as an attempt to find a way to prevent and detect Ponzi schemes,” said Peter Grupe, the FBI’s assistant special agent in New York in charge of white collar investigations.  “But it developed into (more…)