Ten states had lower populations in 2000 than the Census Bureau expected.
A reality check on Census Bureau population projections.
Most demographers regard the first release of data from Census 2000 as a nonevent. Still, it's a fascinating look at how the government projects its numbers.
The new count didn't offer much beyond confirming trends that analysts have observed for years: The population is growing fastest in the West and the South. Indeed, the numbers show that Western states had the largest percentage growth: In 2000, they had nearly 20 percent more residents than they did in 1990. The South, no laggard, recorded a 17.3 percent growth rate. Those two regions eclipsed the Northeast's 5.5 percent growth as well as the Midwest, with 8 percent growth since 1990.
Score one for the forecasters? Maybe not. In 1996, the bureau released projections of the population in all of the states through 2050, including its forecast for the number of people in each state. And although the bureau's demographers were remarkably accurate, they didn't hit their mark every time. While most of the 50 states have more residents than the bureau had projected in 1996, 10 states actually have a lower population in 2000 than the bureau expected. Overestimates occurred in Alaska, Hawaii, Idaho, Montana, New Mexico, North Dakota and South Dakota, West Virginia, Vermont, and Wyoming.
Why? There's no one explanation for the overestimates. Paul Campbell from the agency's population projections branch says several factors contribute to the miscalculations. One possibility: flawed migration projections. Since birth and death statistics are less prone to error, migration miscues are the likely culprit. Gregory Spencer, chief of the population projections branch, says the bureau's reliance on Internal Revenue Service data from previous years to predict future migration may be the problem. When the Census Bureau conducted its last projections, 1992 and 1993 IRS returns were the most up-to-date data source. At that time, California was in a recession and many people were leaving the state. But the California economy recovered and the mass exodus from the state subsided, so states that were predicted to gain refugees from California never did so. It's no coincidence, notes Spencer, that many of the states for which population was overestimated, such as Idaho, New Mexico, Montana, and Wyoming, were expected to become hot spots for former Californians. Thus, overestimates for those states at California's expense.
Another possible reason for the overestimation: When the bureau projects population in off-census years, it first projects the population of the United States on a national level, and then goes back and projects each state's population. There are always discrepancies between the two numbers, and the bureau adjusts state figures to fit its more accurate national projection. Less populated states would likely have been beneficiaries of this somewhat flawed mathematical adjustment.
As Census 2000 drills down geographically, more precise information on migratory trends will come to light. The forthcoming analyses will provide an intriguing scorecard of the bureau's ability to forecast. Stay tuned.