Follow the Stat   

A much-criticized Time magazine estimate of the number of people who enter the country illegally takes on a life of its own.

No one can possibly know how many people illegally entered the United States last year. Yet few statistics are as fiercely contested among immigration activists, thanks in part to a Time magazine cover story that ran last September.

In a 9,000-word investigation on the porous U.S.-Mexican border, Donald L. Barlett and James B. Steele reported that a whopping 3 million people would enter the country illegally in 2004. The estimate prompted accusations from immigrant advocacy groups and former government officials that the reporters embellished the figure--by at least 2 million--lending ammunition to those calling for the placement of troops on the border.

Perhaps unaware of the dispute, news outlets accepted Time's estimate at face value. Within weeks, the 3 million figure appeared in newspapers including the Miami Herald, the Philadelphia Inquirer and New York's Daily News, in editorials, columns and stories decrying the country's immigration dilemma, as well as on Lou Dobbs' and Bill O'Reilly's nightly CNN and Fox News programs.

Even CBS' Bob Schieffer, moderator of the final presidential debate, repeated a variation of Time's estimate. "I'm told that 8,000 people cross our borders illegally every day," Schieffer told the candidates and an estimated 51 million viewers.

Carin Pratt, the executive producer of CBS' "Face the Nation," said she simply divided Time's estimate by 365 to derive Schieffer's statistical gem. Had she been aware of the dispute surrounding the figure, Pratt says she would have looked elsewhere. "But how was I to know?" she asks.

Journalists often lack the time or inclination to double-check statistics, especially numbers reported in trusted publications. As a result, inaccurate figures can take on a life of their own, gaining momentum through what academics call "the anchoring effect."

"It's the tendency we all have to be glommed on to a number that's presented to us," says John Allen Paulos, a Temple University math professor who designed a statistics course for the Columbia University Graduate School of Journalism. That tendency allows tenuous claims to become "the basis of discussion, and only months or years later people say, 'Oh, that wasn't right'--if people even care about it by then."

Many of those who study migration trends, such as Jeffrey Passel of the Urban Institute, say about 750,000 people illegally cross the border each year. After subtracting migrants who die, return home or become legal, the U.S.' undocumented population grows by some 350,000 to 500,000 annually, according to data from the U.S. Census and the former Immigration and Naturalization Service.

Time's estimate, experts say, relied upon dubious assumptions. Interviewing agents and residents on the Arizona-Mexico border, Barlett and Steele heard a troubling statistic: For every migrant caught, three more get through. With the Border Patrol recording more than 1 million apprehensions last year, the reporters concluded that 3 million migrants eluded capture.

One problem with this formula is that apprehensions track total arrests, not distinct individuals. That is, if Border Patrol agents catch the same migrant 10 times, the government records 10 apprehensions. For the 2004 fiscal year, the Border Patrol reported more than 1.15 million apprehensions, though agents actually caught fewer than 750,000 people.

The validity of the 3-to-1 getaway claim is harder to gauge. The same ratio has been cited over the years even as the Border Patrol tripled its force. T.J. Bonner, president of the National Border Patrol Council, a union representing some 10,000 agents, says of such immigration figures: "By no means do we claim this is an exact science."

Readers of Time, however, would never know that. Neither would journalists who repeat the estimate.

Some say that dwelling on the accuracy of Time's estimate misses the larger point. "It's a big difference numerically, but is it a meaningful difference?" asks Steven Camarota, director of research for the Center for Immigration Studies, a group that promotes lower immigration levels. Indeed, the number of people who jump the border is unlikely to sway those who already desire stronger immigration enforcement.

But in the larger debate over immigration reform, statistics can shape public opinion. Advocacy groups commonly further their causes with alarming figures--like the millions of people already living illegally in the country and the thousands who have died crossing the border.

Barlett told AJR he is puzzled that the 3 million estimate has garnered more attention than the story's larger point--how easily people can slip into the country from Mexico and the problems they cause for ranchers, hospitals and law enforcement officials. At any rate, he says he did not trust Census data and that the 3-to-1 ratio is, if anything, too low. Asked where reporters could verify Time's estimate, Barlett says they should not rely on the magazine for statistics in the first place.

"Don't take our word," he says. "Go out and do the work."