Hunting for Numbers

A haunting description from David Simon of analytics and “accountability” gone bad in the Baltimore police department :

“How do you reward cops? Two ways: promotion and cash. That’s what rewards a cop. If you want to pay overtime pay for having police fill the jails with loitering arrests or simple drug possession or failure to yield, if you want to spend your municipal treasure rewarding that, well the cop who’s going to court 7 or 8 days a month — and court is always overtime pay — you’re going to damn near double your salary every month. On the other hand, the guy who actually goes to his post and investigates who’s burglarizing the homes, at the end of the month maybe he’s made one arrest. It may be the right arrest and one that makes his post safer, but he’s going to court one day and he’s out in two hours. So you fail to reward the cop who actually does police work. But worse, it’s time to make new sergeants or lieutenants, and so you look at the computer and say: Who’s doing the most work? And they say, man, this guy had 80 arrests last month, and this other guy’s only got one. Who do you think gets made sergeant? And then who trains the next generation of cops in how not to do police work? I’ve just described for you the culture of the Baltimore police department…”

Sure, the worst case scenario is probably right here, with this kind of thing happening to law enforcement. But that doesn’t mean this shouldn’t be instructive to everyone else. Measurable goals are great, if and only if you really want the logical extreme of that goal. Think long and hard  about that the next time you talk about how you want your team, department, or company to be more “results oriented”.