Washington. -- To the untrained eye, it was just another of the numbing numbers by which journalism calls attention to this or that crisis: ''Every year, the World Health Organization estimates, 220,000 people die from pesticide poisoning.'' To the trained eye of Richard McGuire, New York's Commissioner of Agriculture and Markets, that assertion in an upstate New York newspaper's editorial looked implausible.
It was. Follow Mr. McGuire as he follows the slithering number to a lesson about the strange life led by some statistics and the terrible data on which government often makes decisions.
A call from Mr. McGuire's office to the upstate editor revealed that he had received the editorial from a California syndicate. A call there revealed that the 220,000 number was from information supporting Sen. Patrick Leahy's bill to prohibit U.S. companies from exporting pesticides whose use is banned in America. Mr. Leahy was concerned about America importing foods containing residues of chemicals banned here.
Senator Leahy's office directed Mr. McGuire to the World Health Organization, which directed him to a WHO report. Mr. McGuire wrote to the author in Switzerland, who wrote back to say the figure of 220,000 deaths came from another WHO publication.
The author had warned readers that ''reliable data on pesticide poisonings are not available and the figures given are derived from various estimates.'' Unfortunately, he said, quoted figures often acquire misplaced momentum because they are shorn of their tentativeness.
Here is what the WHO publication the author relied on actually said: ''Of the more than 220,000 intentional or unintentional deaths from acute [pesticide] poisoning, suicides account for approximately 91 percent, occupational exposure for 6 percent and other causes, including food contamination, for 3 percent.'' Of the 3 percent (itself a guess), we are left to guess what portion involved food contamination.
WHO's basic message was that there were actually 20,000 deaths from unintentional pesticide poisoning in a world population of 5 billion. The numbers floated downstream, from the WHO to the senator's office to the editorial writer's office where this was written:
''Every year, the World Health Organization estimates, 220,000 people die from pesticide poisoning; 25 million fall victim to injury or illness. There are no reliable numbers on how many of these casualties result from exposure to unlicensed chemicals imported from this country. . . . But there is no question that the American manufacturers who continue to traffic in these poisons are a significant part of the problem.''
That is, American traffickers in poisons are unquestionably a significant part of the problem, if there is a significant problem. (U.S. Food and Drug Administration tests on imported foods reveal no significant problem with chemical residues on food imports.)
The use of nutty numbers to advance political agendas may result from cynicism or from confusions born of carelessness. The result can be foolish public policies, feeding on and fed by the journalism of apocalypse.
Twenty years ago The Public Interest published ''The Vitality of Mythical Numbers,'' by Max Singer, then president of the Hudson Institute. He dissected a then commonly cited number, that New York City's ''100,000-plus'' heroin addicts were stealing upward of $5 billion worth of property a year.
The assumptions behind the numbers were: 100,000 addicts were each spending an average of $30 a day on their habits, or $1.1 billion a year (100,000 x 365 x $30). Stolen property is fenced for about one-quarter of its value, so addicts must steal upward of $5 billion worth.
Mr. Singer was skeptical.
Most stealing by addicts then was by shoplifting and burglary. All retail sales in the city then totaled $15 billion (including cars, carpets, diamonds and other goods not susceptible to shoplifting). All losses from all forms of theft and embezzlement were about 2 percent. Even if shoplifters accounted for half of that (they don't; employees steal much more), and if all shoplifters were addicts (they aren't), the addicts' shoplifting total would be $150 million.
Burglary? Even if a fifth of the city's 2.5 million households were burglarized each year (they weren't), and accepting the police estimate that the average loss from a burglary was property worth $200, the burglary total ($200 x 500,000) was $100 million.
So even with inflating assumptions, the burglary and shoplifting sum was a quarter of a billion dollars worth of property. That is not chopped liver but it is one-twentieth of $5 billion.
Probably the ''100,000-plus'' number of addicts was inflated. A pertinent question about such numbers is: Whose interests are served by a numerical exaggeration? The answer often is: Those whose funding or political importance varies directly with the perceived severity of a particular problem.
Here, then, is a helpful number: two. When an advocacy group cites hair-raising numbers about the problem for which they are advocating solutions, or a bureaucracy cites such numbers about the problem its programs address (homelessness, drug abuse, teen-age prostitution, whatever), divide the numbers by two.
Similarly, when the Office of Management and Budget issues deficit projections, multiply by three.
George F. Will is a syndicated columnist.