By Rob Gurwitt

Slow news cycles being what they are, the Census Bureau’s release last week of its state and county population estimates for the 1990s got plenty of play. With the presidential contest having settled down and no sudden foreign crisis to cover, even my local newspaper, which is devoted to covering a small and obscure patch of northern New England, gave a good chunk of space to a story about how minorities are transforming the face of the region.

There’s no doubt that, taken as a whole, the census estimates for 1990-1999 were quite interesting. The calculation that states in the South, for instance, saw startling growth in their Hispanic populations — which more than doubled in Georgia, Tennessee, North Carolina and Arkansas — suggests just how quickly Hispanics are spreading into the country’s nooks and crannies. The same is true, though not quite as dramatically, for Asians. It was nice to be reminded that, as a country, we are growing ever more demographically engaging and complex.

But what you might not have pulled out of last week’s news coverage is the conclusion that that’s about all you can say about the census estimates. This is because national and even statewide demographic trends are actually exercises in pointillism — pictures that, when looked at up close, turn out to be made up of thousands of small, local developments. Just what’s driving the Hispanic expansion, for instance, isn’t entirely clear. To make sense of it all you really need to be able to look at the county-by-county figures.

And that’s where last week’s census news is less than solid, a fact that the Census Bureau itself signaled by carefully disclosing its methodology and choosing to do nothing more dramatic than issue a press release noting it had put the estimates up on its Web site. “Whenever we put anything up on the Internet, we notify the press as a courtesy,” says Signe Wetrogan, assistant chief for population estimates and projections in the bureau’s population division, “but it’s not as if we went out with a full-blown press conference.”

The reason for the bureau’s modesty about the county estimates for race and ethnicity is that they are based essentially on a mechanical model that extrapolates from the estimates of each county’s total population growth and from the more detailed statewide numbers, which are based on data collected from the IRS and from the Social Security Administration. To make a long statistical story short, what that means is that if a county in 1990 already had some statistically significant minority population, the bureau probably caught some measure of what’s happened to it over the past decade. But if a new poultry plant opened up five years ago in some Southern hamlet that was all white and African-American in 1990, the fact that there are now a couple of hundred Vietnamese and Laotian families living there didn’t show up in last week’s numbers.

All of which is to say, last week’s revelations were really just a tease. Those of us who want to know how the country’s fabric really got rewoven over the past decade will just have to wait until early next year, when the Census Bureau will finally begin releasing detailed results from this year’s census.

Rob Gurwitt is a staff correspondent for Governing.