• Choose a category
• All
• Classic-Fiction

By Root 488 0

system employing one of the National Science Foundation’s supercomputing centers to assign personal mileage targets that meet each of the 200 million motorists’ needs while maintaining the previous annual average of 14,300. In method 2 we tell drivers not to stress out over it and to drive as much or as little as they please with no regard to how far they drove the prior year. If Uncle Billy Bob, who used to walk to work at the liquor store, decides instead to log 100,000 miles as a shotgun wholesaler in West Texas, that’s fine. And if Cousin Jane in Manhattan, who logged most of her mileage circling the block on street-cleaning days in search of a parking space, gets married and moves to New Jersey, we won’t worry about that either. Which method would come closer to the target of 14,300 miles per driver? Method 1 is impossible to test, though our limited experience with gasoline rationing indicates that it probably wouldn’t work very well. Method 2, on the other hand, was actually instituted—that is, the following year, drivers drove as much or as little as they pleased without attempting to hit any quota. How did they do? According to the National Highway Traffic Safety Administration, that year American drivers drove 2.88 trillion miles, or 14,400 miles per person, only 100 miles above target. What’s more, those 200 million drivers also suffered, within less than 200, the same number of fatalities in both years (42,815 versus 42,643).

We associate randomness with disorder. Yet although the lives of 200 million drivers vary unforeseeably, in the aggregate their behavior could hardly have proved more orderly. Analogous regularities can be found if we examine how people vote, buy stocks, marry, are told to get lost, misaddress letters, or sit in traffic on their way to a meeting they didn’t want to go to in the first place—or if we measure the length of their legs, the size of their feet, the width of their buttocks, or the breadth of their beer bellies. As nineteenth-century scientists dug into newly available social data, wherever they looked, the chaos of life seemed to produce quantifiable and predictable patterns. But it was not just the regularities that astonished them. It was also the nature of the variation. Social data, they discovered, often follow the normal distribution.

That the variation in human characteristics and behavior is distributed like the error in an archer’s aim led some nineteenth-century scientists to study the targets toward which the arrows of human existence are aimed. More important, they sought to understand the social and physical causes that sometimes move the target. And so the field of mathematical statistics, developed to aid scientists in data analysis, flourished in a far different realm: the study of the nature of society.

STATISTICIANS have been analyzing life’s data at least since the eleventh century, when William the Conqueror commissioned what was, in effect, the first national census. William began his rule in 1035, at age seven, when he succeeded his father as duke of Normandy. As his moniker implies, Duke William II liked to conquer, and in 1066 he invaded England. By Christmas Day he was able to give himself the present of being crowned king. His swift victory left him with a little problem: whom exactly had he conquered, and more important, how much could he tax his new subjects? To answer those questions, he sent inspectors into every part of England to note the size, ownership, and resources of each parcel of land.6 To make sure they got it right, he sent a second set of inspectors to duplicate the effort of the first set. Since taxation was based not on population but on land and its usage, the inspectors made a valiant effort to count every ox, cow, and pig but didn’t gather much data about the people who shoveled their droppings. Even if population data had been relevant, in medieval times a statistical survey regarding the most vital statistics about humans—their life spans and diseases—would have been considered inconsistent with the traditional Christian concept