249,632,692? No Way!

December 28, 1990

The official 1990 count from the Census Bureau is in -- 249,632,692 (including Americans overseas). No one -- no one -- believes there are that few Americans. Even Commerce Department officials who describe the 1990 tally as "full, fair and efficient" know that several million Americans were not counted. Last fall, a Census Bureau estimate of the resident population (not including Americans overseas) was 253.4 million. The actual count of residents released Wednesday was 248,709,873. Experts in and out of government know the estimate is better than the count.

For gross apportionment purposes, the counting error is of no consequence. A perfect count would result in the same distribution of House of Representative seats among the states. Maryland would still have the same eight it had in the 1980s and will have in the 1990s based on the under-count. New York would still lose three seats, California would still gain seven, and so on.

But that does not mean imperfect counting is inconsequential. Census under-counting occurs primarily in urban areas, especially minority neighborhoods. This has two very harmful effects. One, redistricting for city and state legislatures and Congress is skewed against the people in those neighborhoods. Two, government aid programs based on specific groups of the population (the aged, the young, the ill, the unemployed, etc.) are not funded fully and fairly. In the last 10 years, Mayor Kurt Schmoke recently told a congressional committee, Baltimore was shortchanged $230 million because of the 1980 Census under-count.

The secretary of Commerce is required to study the need for an adjustment in the count, to be based on sophisticated demographic techniques. He has until July 15. He may at that point decide to go with the traditional enumeration, or he may agree to the adjustment. Most observers believe he will stick with tradition -- in part because the adjustment would provide more political clout to largely Democratic urban areas.

We used to think enumeration was preferable to statistical estimation. Now we don't. Nose counting is just not appropriate any more. The 1990 Census was the most expensive, expert and best planned ever -- and it didn't do the job. It missed more noses than the 1980 Census did. The nation is now too big, too diverse, too mobile and too suspicious of government officials with long lists of questions about intimate matters.

In our view, a simple count is still good enough for purposes of apportionment. But for deciding and allocating federal and state aid programs that are based on specific sub-groups of the population, the more precise statistical estimates should be used. Sen. Daniel P. Moynihan of New York and Rep. Thomas Sawyer of Ohio have shown interest in this approach in the past. We believe 1991 is the year to do something about it.

Baltimore Sun Articles
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.