After the Census Bureau released population estimates showing that core counties were (at least in some metro areas) growing faster than exurban counties, the media was full of headlines about this alleged trend. An extreme example came from the Washington Post: "An end to America's exurbia?" (1)
But in fact, these estimates prove almost nothing, for two reasons. First, Census mid-decade estimates are only estimates- which means that their reliability is a bit questionable. To show one extreme example, 2009 Census estimates "showed" that the city of Atlanta's population had risen to 540,000 (2) - about a 30 percent jump from its 2000 population. But the 2010 Census showed that the city's population had grown only slightly, to 420,000 (up from 416,000 in 2000). (3)
Second, this Census estimate is just for one year, and thus may be either (a) a statistical fluke or (b) a temporary result of the recession (which, by making houses worth less, discouraged people from selling them and thus impeded residential mobility generally). It may be the case that later estimates will yield differing results, or that migration to outer suburbs will increase when housing values pick up and people can sell their existing houses for more.
On the other hand, it may be the case that rising energy prices and falling housing prices create a long-term trend- but we won't know for sure until an actual Census is taken eight years from now.