Hard to believe there is a "Hunger Lobby." I did business with the man who ran a major homeless shelter in LA. He caught the train in every day from Orange County, was met by a car, and took at least three Cruise vacations a year.
Of Fuzzy Math and 'Food Security'
By TOM ZELLER New York Times
WE are not concerned with the very poor," wrote E. M. Forster in "Howard's End," his 1910 novel of culture and class. "They are unthinkable, and only to be approached by the statistician or the poet."
Where the statistician ends, however, and the poet - or the politician or the advocate or the salesman - begins has long dogged the science of probability and averages. And to the extent that Americans are concerned with the very poor, it is not surprising that cold numbers and impressionistic reasoning should mingle uneasily.
Take the annual survey of homelessness and hunger published by the United States Conference of Mayors, a nonpartisan organization encompassing 1,183 cities with populations of 30,000 or more. The 2003 installment arrived in mid-December, and as it has for each of the last 16 years, the survey reported an increase in "demand for emergency food" - this time by 17 percent.
The figure was quickly noted in the press.
"Homelessness, Hunger Worsen,'' was the Boston Globe headline. The Washington Post had it as "Survey Indicates More Go Hungry, Homeless.'' And an editorial in The New York Times cited the survey in taking note of "continuing increases in hunger and homelessness in the nation's major cities.''
According to a chart included with the results of the survey, which was conducted in 25 major cities, demand for food assistance increased by a sizeable percentage every year since 1988. (The best year, 1995, had a 9 percent increase; the worst, 1991, was up 26 percent.) But the math behind that trend had, for the last several years, caught the eye of the conservative Heritage Foundation, which published a pre-emptive rebuttal a few days before the survey was released.
"Policy makers should be wary of the claims of 'increasing hunger' that are likely to be made in the upcoming 2003 U.S. Conference of Mayors report," the foundation's paper warned.
So what's the real story? Social ills like hunger and need for food assistance are notoriously difficult to measure. In a flagging economy, or even in boom times, those on the front lines - soup kitchen operators and food bank administrators, who must rely on donations and government subsidies - are always in the shadow of a shortfall. So it is not surprising that the mayors' survey, which relies primarily on the collected impressions (and varied recordkeeping) of these agencies, would consistently report an increase in demand.
"We can argue about the numbers and the methodology," said Steve Brady, the president of the Sodexho Foundation, which supports hunger-related initiatives and was a co-sponsor of the mayors' survey. "But it is certainly a clear indicator that the country as a whole is experiencing this problem."
In publishing a table titled "Hunger and Homelessness in America's Cities: A 16-Year Comparison of Data," however, the mayors' survey suggests an annual percentage increase that proves difficult to track. When asked for the numbers behind the percentages, Eugene Lowe, the assistant executive director for community development and housing for the mayors' conference, said there weren't any.
The mayors' conference, Mr. Lowe explained, requests the percentage change in demand from each city, and from that publishes an annual average percentage change for all the cities in the study. "The cities do the research themselves," Mr. Lowe said.
The Heritage Foundation took that methodology to task, comparing the rate of increase implied by the mayors' data with measures taken by the Census Bureau and the Department of Agriculture. Those agencies have found a slight rise in what the government calls "food insecurity" (see chart at right) in the last few years, but it remains lower than in 1995. The mayors' report has demand for emergency food rising an average of 16 percent a year in the same period.
"No set of data is perfect," said Melissa G. Pardue, an author of the Heritage Foundation analysis. She pointed to Census Bureau figures on the topic, which are generally considered to be too low. "But the undercount is expected to be roughly the same from year to year," she said, which means the trend should be a fairly accurate snapshot of changes in demand. That snapshot is at odds with the mayors' findings.
It is also not likely, according to Mark Nord, a sociologist with the Agriculture Department, that trends spotted in the 25 cities surveyed would so drastically deviate from national trends. "There are a lot of weaknesses in the methodology of the mayors' report," he said. "In terms of identifying trends, it's not a very good source for that."
OF course, neither side in the debate wants to divert attention from the millions of Americans who, at one time or another, need help getting enough to eat. And Mr. Lowe of the mayors' conference said it was a mistake to turn their numbers into something they are not.
"There are always people who will try to use these percentages to reflect the actual number of people" seeking emergency food aid, he said. "But you cannot do this with our numbers." Instead, Mr. Lowe said, the trend simply shows that the pressure on the informal and formal networks intended to address this problem is always great.
The Heritage Foundation concedes this point. "It's certainly important that we're talking about this," Ms. Pardue said. "But using exaggerated numbers is not helpful to the debate and to the needs of those people who, in the midst of a recession, may actually be hungry."
Or as a quotation attributed to the late business professor Aaron Levenstein has it: "Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital."
Copyright 2004 The New York Times Company |
