Judged on the basis of U.S. Census data, swaths of Detroit would seem to be too poor to support much in the way of retail development -- a situation repeated in cities across the country. Social Compact, a 16-year-old nonprofit organization has been steadily working to correct such misperceptions, applying advanced data-mining and analysis techniques to spur urban development.

Using software from SAS and its DataFlux subsidiary, Social Compact researchers gather data from a broader range of sources than has often been used, including departments of motor vehicles; credit, mortgage and other financial histories; and utility records. All of the data is stripped of personal information that would identify specific people, but it is linked to individual neighborhoods. Citigroup is working in partnership with Social Compact, supplying analytics and depersonalized and aggregated data, as well as financial support.

Such partnerships are at the core of how Social Compact, which is based in Washington, D.C., is able to obtain and then analyze data that is much more complete and accurate than information from government and other sources. The organization's mission draws on the belief that "community development can no longer depend on public investment to succeed," and that it must "rely on public and private sector partnerships," says Social Compact CEO John Talmage.

Doing that, using data to look at neighborhoods in nine cities and the entirety of three others, Social Compact has found 850,000 people who were not counted in the most recent census in 2000, he says. By the end of next year, the organization's data-mining and analysis could boost that number to as many as 3 million people in just parts of 30 U.S. cities, "representing $80 billion in unrecognized household income."

With remarkable understatement, Talmage adds, "so the numbers are very large."

The Brookings Institution, involved in a project with the University of Michigan to come up with strategies for downtown development in Detroit, last year invited Social Compact to gather and analyze data and researchers found that the median income was closer to $59,000 than to the $40,000 reported in census figures. They found a higher home ownership rate than indicated by the census and that the median home value of neighborhoods studied also was substantially more, says Olga Savic, director of strategy and external affairs for the Detroit Economic Growth Corporation.

"The picture was better than what conventional sources of information might suggest, and it really helped to validate the work we're trying to do to attract retail to the area," she said. Savic is passionate about her love of Detroit. Of course, she's paid to be. But on the other hand, even though the renaissance underway in Detroit is often held up as an example, there is still an untoward, and unfair, stigma attached to the city because of unemployment, riots, and the penchant some residents used to have for setting things on fire.

"People can have a really good life here," she says, adding that she believes over the next three years of her organization's contract with Social Compact, "we'll be able to recast this image of Detroit as a place where people really want to be."

Washington, D.C., which has its share of misunderstandings that aren't associated with who is or isn't inhabiting the White House at any given time, is further along in its use of Social Compact data, which is directly linked to, for instance, a 550,000-square-foot retail shopping center in the Columbia Heights neighborhood. About 10 new residential projects are also in the works there, with planned developments involving a large full-service grocery store and other shops, says Michael Stevens, of City Building Consulting.