The Realistic Bigotry of Reality in the Workplace

(While discussing Murray’s Quiz, I made the following comment and Will thought it was good enough for a post, so here’s the post version, slightly edited.)

As silly – and limited – of a metric as it is (for various reasons), IQ is interesting because of its distribution. And (at least in my experience), IQ distribution maps to real intelligence distribution, even if the measure itself when applied to an individual doesn’t necessarily place you accurately on the bell.

Here’s the curve:

(In other words, if we could have a real measure of intelligence that accurately mapped the population, you’d still see a distribution that looks a lot like IQ distribution, just different individuals would be in different places on the curve)

When looking at task-oriented skill maps, it’s pretty likely that the distribution is also a bell… about 60% of people are within one standard deviation of “average” for any particular skill set. 20% are above that, and 20% are below.  So the distribution of competencies for skill complexity will look something like the curve above.

Now, there isn’t a 1-1 correspondence there (if you’re right at the 50% percentile mark for general intelligence, you’re not necessarily at the 50% percentile for the ability to grok complexity, but it’s likely that you’re very close).

In practice, this applies to just about everything from digging ditches to being an academic; in any particular population set for a profession, you’re going to see a microcosm of that sort of distribution, barring intentional structural modification.

So if you got all the ditch diggers together, and you measured their performance as individuals, you’d see that most ditch diggers were average, and some were really good at it (good body kinesthetics, smart about where to pile the dirt, etc), most were average, and some are really bad.

The problem is that those curves overlap an awful lot. Someone who is going to be a good physicist might also be a good line worker at a machine shop or a good ditch digger. Someone who is going to be a good machine shop worker might also be a good physicist or a good ditch digger, and someone who is a good ditch digger might make a good line shop worker or a good physicist. All of that is true.

But the likelihood of someone being good at all three of those things is vastly smaller than the likelihood of them being good at two or one of them… and the person who is likely to be good at only one of them is vastly more likely to be good at only ditch digging.  Now, I work at Caltech and I can state definitively that the stereotypes of Caltech people that you see on shows like The Big Bang Theory are in fact stereotypes, but like all stereotypes this one comes from somewhere.  There are people who are awesome at physics who would make terrible ditch diggers.  Not just the obvious example, for obvious reasons, but there are people who are very complex thinkers who don’t have good proprioception or a good physical constitution, or what have you.  Some tasks don’t even become actual paying jobs until you get to the pinnacle of performance: you can’t make a living playing basketball unless you’re better than 99.999% of the people who play basketball.

I mean, I can sit at the screen and talk all I want about how Devin Ebanks is never going to be a starter in the NBA… Devin Ebanks would murder me in a game of one on one.  He could spot me 10 points and the ball and we could play to 11 and I’d bet he beats me 11-10.  But I digress.

The problem is, as the base job – whatever it is – gets more complex, the apex of the curve among the working population itself moves farther to the right of the greater population.  So any collection of physicists will have a bell curve distribution for how good they are at being physicists, but the middle-of-the road physicist is still going to be somewhere significantly to the right on the IQ curve… and they’re very likely to be somewhere significantly to the right on the “ability to grok complexity” curve, too.  So someone who has the ability to master a lot of complexity can do very complex jobs, but is also likely to be able to master very simple ones (although they’re likely to dislike that).

But the converse is simply not true.

Back to IQ: as limited of a measure as it is, there is a definite correlation between IQ and the ability to master complexity. So roughly half of the population ain’t very good at dealing with complexity, and of the half that is good at it, ~80% of those are just average.

This is all well and good provided that your available jobs are well distributed across complexity, too. If your available job skill sets more or less look like a bell curve, then it works out fine: everybody gets a job that they’re potentially good at.

But that’s not what we have here in the good old U.S. of A.

To be clear: I’m totally fine with outsourcing basic manufacturing for all the good outcomes, from a globalization perspective. Having other countries have a robust working economy is good for everybody – for international stability reasons, if nothing else.  There’s obvious justice implications, etc., etc.

But the consequence of that is that a good number of lower complexity jobs just aren’t around to be had, here in the U.S. And those lower complexity jobs that are available to be had are vastly outnumbered by the individuals looking to get them, because most people just aren’t that good at even medium complexity jobs.

So we have a situation where the jobs that we have available require skill sets and cognitive abilities that most of the population doesn’t have… and most of our unemployed people have skill sets and cognitive abilities that map to “jobs that aren’t here any more and aren’t going to return any time soon”.

This is not just a potential problem in the U.S.

Most people – worldwide! – are incapable of performing what the OECD calls “level 3 science” (<40%) and “level 3 mathematics” (<50%) at 50% proficiency. Level 4 – 6 are increasingly smaller numbers of the population. Reading? That’s slightly better, but not by much.

Level 3 math and science… actually, not very awesome. We’re not talking about the ability to do complex differentials on paper.  If you are a regarded as proficient at Level 2 in mathematics, you can balance your checkbook with 50% reliability.  That’s an “F”, in American schools.

Now, neither math nor science nor literacy is going to map directly to the ability to handle complexity, generally, but put the three together as a proxy measurement and it’s not all that bad.

In most jobs that are available in the modern economy, you’d have to have at least the complexity mapping skills of someone who does better than 50% at these tasks, regularly. Like, 85% or so. And yet, if reading, math, and science are a reasonable proxy, about… half… of the population has the potential (the potential!  not the capability!) to accomplish this at above a 50% success rate.

This is why people are constantly bitching about their co-workers, every industry, everywhere in the U.S. Because the failure rate of your average worker to perform even the basic skills of their job is so freaking high. And it’s not because they’re lazy, it’s because they don’t possess the capacity to do complex jobs with any sort of reliable success.

And this isn’t because our educational system sucks, because everybody knows that education in Finland is way better than the U.S. and most people in Finland suck too.

We can’t fix this problem with better education. People, on average, are just really bad at handling meta information and multiple layers of abstraction and context-switching and multi-threaded tasking.  And this is the “information economy”.

Moderately complex jobs are the basic job, now.  Computers, robots, earthmoving equipment… technology has enabled us to ignore lower-complexity jobs…or vastly, vastly increase the productivity of a single worker to the point where we don’t need many workers to perform that lower-complexity job.

We’re slowly turning all non-entry-level, career path jobs in the country into “in order to perform this job at better than ‘passing’ levels of proficiency, you need to be at least as good at your job as Devin Ebanks is at playing basketball”.  But Devin Ebanks is in the 99.999%.