(FORTUNE Magazine) – BRAINS IN THE OFFICE

One was nodding along with Bill Gates's views about IQ, as recently (November 25) reported in this magazine, and yet one yearned to add a few codicils. Bill says that his managers need to look for intelligence above all when they hire people, that his company's real competitive advantage is its supersmart people, and that "in terms of IQ, you've got to be very elitist in picking the people who deserve to write software." As the author of a 1994 paperback called A Question of Intelligence: The IQ Debate in America, a work that has absent-mindedly not been plugged in this space for going on a year now, one could not dissent from any of these statements. Obviously, a leading-edge company like Microsoft has a special need for brainy people. There are, however, a couple of missing thoughts in Bill's formulations. These are that (1) all companies would benefit from hiring smarter people, and (2) IQ matters in all jobs, including sweeping up the place after the programmers go home.

These statements get a lot of people hot and bothered, but they are empirically unassailable. A few years back, a panel of the National Research Council (an arm of the American Academy of Individual Sciences) took a long look at the U.S. Labor Department's General Aptitude Test Battery--in effect an IQ test. A major question before the panel: Do GATB scores predict job performance? Crisp answer: You bet. In jobs all across the skills spectrum, higher test scores are associated with shorter training times, greater productivity, and lower turnover rates.

Having recently gotten an advance peek at a forthcoming fascinating issue of the journal Intelligence, which elaborates these themes in greater detail, one senses that Microsoft will not long be alone in promoting worker intelligence as a business strategy. Edited by Linda S. Gottfredson of the College of Education at the University of Delaware, the issue incorporates an avalanche of data about the workplace correlates of general intelligence, or g--essentially a measure of reasoning and problem-solving ability. g, which is lowercased even when starting off sentences, correlates at about 0.4 with an average job. (Positive correlations range between 0.0 and 1.0.) For Gates's programmers, the correlation would probably be around 0.8. For the sweeper-upper, it might be down around 0.2. Still, the correlation would always be positive, i.e., more g means better sweeping, and g gives you a far better prediction of job performance than does the applicant's level of education or past experience. The only human trait that sometimes competes with g as a performance predictor is "conscientiousness," numerous tests of which are now available to your human resources department.

The statement that you are better off with smarter workers does not mean that you want to hire Ph.D.s to sweep the factory floor (even though, one gathers from various morose articles in the Chronicle of Higher Education, some of them could be available these days). What it means is that you want workers at the higher rather than lower ends of the intelligence ranges for their jobs. Among the nuggets in the Intelligence report are some data on current effective IQ ranges for different occupations. The IQ range for working stiffs at a warehouse loading dock would be around 80 to 100. For a car salesman or a cop patrolling city streets, the range would be around 92 to 110. For a registered nurse, maybe 98 to 118. For an advertising manager, 108 to 126. (The report offers no data on this person's golf game.) For Microsoft bosses, very to extremely elitist.

ACTION UNLIMITED

Sid Diamond, an ebullient chap who runs the sports books for the seven Circus Circus casinos in Nevada, was filling us in the other day on an amazing new technology. A sports book, for folks not into Sid's world, is a place where you can bet on the outcomes of sporting events--baseball, football, basketball, you name it. You can bet dozens of events at a time and loll around watching them all on huge screens.

But as you already intuit, this is still not thrilling enough for some guys. Their problem is now being addressed at Excalibur, a Circus Circus casino with a medieval theme, where the kids can watch knights jousting. The casino's solution is to offer parents bets not just on game outcomes but on play action during the game.

You walk into the book's main room, go over to the cashier's desk, and buy a certain amount (but not less than $50) of betting credit. The cashier takes your money and gives you a PIN number. You type in the number on a computer whose display has live action of the event you're betting on--a football game, say--but also offers various instant propositions. Will they pass or run on third down? Will they punt or go for it on fourth down? Will the punt returner call for a fair catch? The odds on each bet are set by a database reflecting past experience in thousands of similar situations (and also reflecting the house's need for an edge of 5% or so). Every time you make a bet, the balance in your account (also shown on the screen) changes, for better or worse. When it's all over, the computer prints out a summary of your action, and if your balance is positive, you go to the cashier and get paid.

And if you get wiped out, there's always jousting.