Selection Bias

If, several hundred years from now, our descendants want to consider giving representative government another go, they’ll need to figure out a better way of picking leaders.

I don’t mean things like “restricting the franchise to stakeholders” and “IQ tests,” though those are great ideas in themselves.  I mean they’ll have to overcome systemic selection bias — a kind of Peter Principle that promotes people not just to their level of incompetence, but based on a completely different set of skills.

Football coaches are a good example.  Chances are good that a brilliant coordinator will flame out as a head coach, simply because the skillsets are so different.  Head coaches have to “coach up” — their day-to-day jobs involve handling the owner, general manager, the media, and his subordinate coaches.  Their relationship to the day-to-day, X’s and O’s of the game that’s played on Sunday is usually pretty tenuous.  Coordinators, on the other hand, “coach down.”  They do the nuts-n-bolts stuff, handle the players and their issues, devise the specific schemes and match-ups.  There’s almost no overlap between those two areas of responsibility.

(The less said of college head coaches who jump to the pros, the better.  College kids aren’t pros, the boosters aren’t the owner, and you don’t get the ridiculous recruiting advantage bigtime college programs do.  Examples are numerous, but my favorite is Steve Spurrier — in his brief tenure with the Redskins, he really did seem to believe his “huck it downfield and let his five-star receivers blow past the opponent’s two-star DBs” would work in the pros.  But everyone in the pros is a former five-star recruit).

Either way, though, there’s simply no relationship between the two skillsets, and thus no way to judge.  A brilliant X’s and O’s guy, who gets the most out of limited athletes, might be great at schmoozing the owner and handling PR…. but then again, he might not.  The point is, there’s no way to tell if or how his X’s and O’s work will translate over to schmoozing and PR, and — given the demands of the business — there’s no way to give him a trial run.  Yet coordinators always get promoted to head coach, because… well, how else are ya gonna do it?

Politics works the same way.  The traditional cursus honorum — state legislature, national legislature, state governor, president — selects for a very different set of skills than those the President needs.  A dull-but-clubbable party man makes a great Senator, but a lousy President.  It takes some real skills to be a state governor, but high among them is the ability to massage entrenched local elites — you have to be wired in, but in a totally different way than a Senator does.

The system, in other words, is set up to produce dull-but-clubbable party men.  They were quite open about this in the 19th century, in case you think I’m making it up.  That’s why “nominating conventions” were real things back then — the wheeling and dealing was brutal, smoky back rooms weren’t just a metaphor, and sometimes it broke down spectacularly and you ended up running Franklin Pierce or someone like that.  This was because the 19th century actually believed in that “federalism” stuff, and the savvy operators avoided national politics for state governorships.

Trump is a huge anomaly who has exposed just how systemically flawed our process is.  We need to figure out how to overcome this selection bias effect… or, at least, our great-great-great-great-great-great grandchildren do.

Loading Likes...

4 thoughts on “Selection Bias

  1. Jay Carter

    Winston Churchill nailed it.
    “The best argument against democracy is a five-minute conversation with the average voter.”

  2. Frip

    “A dull-but-clubbable party man.” That’s me right there! Great post. Rotten is the chestnut cuz the chestnut gonna rot!

  3. Al from da Nort

    Your example of selecting football head coaches has an interesting parallel to the problem of finding competent wartime military commanders. Peacetime generals have historically not done well as a group in wartime and the faster they are brushed aside the better for all concerned. That’s because the skill-sets are likewise different. In peacetime you get promoted by checking the boxes, both formal and informal, while being pleasing to your superiors (i.e. a suck-up).

    In wartime, of course, it’s results that count, but you do have to know the systems involved to gain those results. Simply being unconventional is just courting danger, usually (but not always) causing needless casualties and wasting vital resources. Also, being competent in wartime at a lower hierarchal level is no guarantee either. But demonstrating incompetence in wartime is reason enough for removal for all the obvious reasons.

    The pattern in winning military systems seems to be that successful leaders just sort of emerge from the potentially competent people at the lower levels (coordinators in the football example) as the conflict progresses, much like successful head coaches. The trick is to recognize them early on and not let jealous peers sabotage them.

    1. Severian

      Worse yet, all wars are political, and politics exacerbates the problem of finding competent leaders.

      The US Civil War is a good example. Civil War buffs are always going on about how General X should’ve been replaced with General Y and the whole thing would’ve been over by Christmas, e.g. Braxton Bragg with James Longstreet, or George McClellan with anybody. But since armies were raised at the state level, state politics inevitably got in the way (pulling Bragg probably would’ve resulted in an all-Virginia high command, for instance, when every other state in the CSA already complaining about the “all Virginia, all the time” nature of the war). National politics, too — McClellan was, of course, Lincoln’s opponent in the 1864 presidential election. And then there’s the “politics” of dealing with the President himself – a lot more officers were loyal to McClellan than Lincoln, for example, and everyone knows what a prick(ly guy) Jeff Davis was.

      I’ve been told this was a big problem in Vietnam, too — someone like Maxwell Taylor might have been a better overall commander, but Westmoreland was a much better politician (I don’t know whether or not this is true, or justified). Add to that the 13-month tour of duty for platoon leaders, who rotate back to the world just as they’re becoming competent, and I’m amazed we accomplished anything there. As to how to fix it, well… I don’t think it can be done. Nobody knows how he’ll react to bullets until he’s shot at, and nobody knows if he can lead men or not until he tries. I’m sure the Army has many excellent tacticians who wilt at the sight of blood, since most of us do. Short of having a small bush war simmering at all times, such that we can rotate every West Point grad though, I don’t know how to address it. (And nobody wants that!)

Comments are closed.