Falling for the Gap

The New York Jets have just won the Super Bowl. It must be true: There's a story on the front page of The New York Times, and there are color pictures in Sports Illustrated. And indeed it is true. Or, rather, it was true, a few decades ago. Only a truly inexperienced sports writer would suggest that the New York Jets are the current champions of the NFL.

Did you hear that the Dow Jones industrial average has topped 1,000? That, too, is old news, as even the most junior financial writer must know.

How about this one: There is a broad and widening gap on the Internet between white and minority Americans. This familiar claim, often asserted as a fact by policy makers and digerati alike, is also based on old information. Reinforced by White House press releases and presidential candidates' speeches, the idea is so ubiquitous that even the usually well-informed have come to believe that white Americans are online and minorities are not.

Not so. It may have been true in 1996 or 1997, when the Internet was only a few years old as a popular medium and personal computers cost thousands of dollars. But today, with dirt-cheap Internet access and computers approaching the costs of television sets, assertions of a "digital divide" or "racial ravine" are as correct as identifying Joe Namath as football's current MVP or pinning last week's Dow at 1,000.

Misled by stereotypes, misinformed about survey techniques, and misdirected by interest groups, the media have treated the "digital divide" as a crisis requiring government intervention. As a result, billions of dollars might be spent to address needs that no longer exist.

To understand how this happened, start with stereotypes. East Coast journalists typically equate "minority" with "African American," portraying the country as divided between black and white. This view omits the fastest growing minority group, Hispanic Americans, who in just a few years will be the largest minority group in the country.

Confusing "minority" with "African American" also leads journalists and analysts to forget that it is not among whites but among Asian Americans that Internet and computer use are approaching levels of penetration comparable to those of the telephone, television, and indoor plumbing. So even using the old survey data, it was always inaccurate to claim that minority Americans were not online in large numbers.

But the issue of dated information is crucial, especially because a year or two in "Internet time" is the equivalent of a decade for older media. The findings of the most frequently cited "digital divide" study, released last summer by the U.S. Department of Commerce, were presented and widely reported as new information. The study was actually an analysis of surveys in 1998 and earlier. When it was released, more-current information was already available from market research firms, but only a handful of news organizations reported the newer data.

The Commerce Department study made page one headlines with its conclusion that the United States faced a "racial ravine" dividing online white Americans from information-poor minorities. "For many groups, the digital divide has widened as the information `haves' outpace the `have nots' in gaining access to electronic resources," it said. "Between 1997 and 1998, the divide between those at the highest and lowest education levels increased 25 percent, and the divide between those at the highest and lowest income levels grew 29 percent."

That sounds impressive, but if you look more closely you may spot a crucial methodological flaw. Among reporters for the major daily newspapers, only John Schwartz of The Washington Post noted the problem. "Last year's studydid not collect information about out-of-home access," wrote Schwartz. "It is not possible, therefore, to say whether the digital divide is growing based on access from all places." In other words, the Commerce Department's claim of a "widened" gap was not supported by the data it cited, because the surveys asked different questions from year to year.

"We never stated that we have any information about widening with regard to anywhere access," says Larry Irving, who directed the government study before he resigned as assistant secretary of commerce. "But certainly we can prove the in-home access gap is widening."

Yet according to every survey taken in the last few years, Americans get their online access at work and at school in far larger numbers than at home. According to The Internet News Audience Goes Ordinary, a 1999 report from the Pew Research Center for the People and the Press, 62 percent of employed Americans go online through their jobs, and 75 percent of students go online from their schools. The Commerce Department study reported only on use of personally owned computers, thus excluding the millions of users (including this writer) who are online every day but do not own a computer. This is like assuming you don't need a driver's license unless you buy a car.

Regardless of whether the questions in the federal survey were correctly phrased, they were asked in 1998. Surveys conducted this year have found not only that minorities are not falling behind but that they are catching up.

"If you missed Christmas [1998], you missed a big surge," says Ekaterina Walsh, author of The Digital Melting Pot, a report based on 1999 data collected by Forrester Research of Cambridge, Massachusetts. "Quite a lot of people got cheap PCs. We were surprised ourselves, because we were projecting lower numbers for online penetration and commerce [than the study found]. Even a month made a big difference." Walsh adds that the federal report may undercount or ignore WebTV, which in 1998 was one of the lowest-priced devices enabling consumers to go online.

"I think we did miss a certain amount of information with regard to lower-priced PCs since December," concedes Irving, the former Commerce Department official. But he stands firm on the question of whether the department's study was misleading because it tracked only computer use at home. "No one has been tracking out-of-home access, as far as we know," he says.