Baltimore's preliminary census count fell far short of projections. The mayor was annoyed and went door to door seeking residents missed in the head count. The city sued to keep Census Bureau offices here open longer than planned.
Despite resemblances to 1990, this was 1980. Fresh evidence of Baltimore's plummeting population battered city pride. It has become a decennial ritual.
For all the horror stories and gnashing of teeth engendered by the 1990 census, the 1980 head count was even unkinder to Baltimore, save in one respect: People did participate.
Two weeks after Census Day in 1980, nearly 79 percent of Baltimore households had filled out and mailed in their forms. At about the same stage this year, only 61 percent had.
The story was true across the country, and it may be the story of the 1990 census: What if you held a census and nobody came?
"If people don't mail back the forms, it's really almost impossible to have an accurate census," said Eugene P. Ericksen, a Temple University census expert. "We learned from research in the 1980 census that omissions occurred in areas where the mail-back rate is low."
The Census Bureau expected a drop in participation. Americans now are increasingly rushed, besieged by telemarketers and junk mail, jealous of their privacy and wary of government, surveys show.
"People who tend to respond are people who are married and in families, have a middle-class income, are relatively well-educated and middle-aged or older. These are all people we used to have more of, proportionally. We're now a far more diverse population," said Martha Farnsworth Riche, national editor of American Demographics magazine.
And so the bureau spent more time and money than ever on census promotion, even commissioning ad agencies to craft campaigns targeting blacks, Hispanics and Asian-Americans.
The response was mass apathy.
Despite its residents' apparent lack of interest, Baltimore's preliminary population came closer to estimates this year than in Ten years ago, the city's preliminary count was 737,557 -- fully 81,000 under the most accepted local estimate. The count was eventually revised upward by nearly 50,000, but Baltimore still slid from seventh to 10th place in city population, behind Phoenix.
By contrast, this year's preliminary figure of 720,100 is "only" about 24,000 shy of the state's best estimate.
After Census Bureau checks, the city's population will probably rise modestly. Baltimore has now dropped to 13th place, behind Indianapolis.
More than ranking is at stake. Baltimore's declining population costs the city political representation and federal aid. This hurts doubly because Baltimore invested, by Mayor Kurt L. Schmoke's own account, $1 million to promote a "complete count" in the 1990 census.
Mayor Schmoke, like his predecessor William Donald Schaefer, has protested the results, if more quietly than 10 years ago. To be sure, he has horror stories to tell. Despite extensive city help, for instance, the census missed fully 10 percent of public housing units in Baltimore.
And the mayor has joined a debate over how the census should be taken -- by traditional head-counting methods, statistical sampling or a combination of both.
The drop in census participation islikely to intensify the debate for tworeasons:
* The fewer people the Census Bureau counts by mail, the more it has to spend going door to door to find them. Census expenses have already spiraled well beyond the rate of inflation. At $2.6 billion, the 1990 census will cost over 10 times more than the 1970 count.
* People who don't mail back their census forms are often those who are hardest to count by other methods: blacks and other minorities, the homeless, undocumented workers, anyone with reason to distrust the government.
In 1980, by the Census Bureau's own estimates, blacks accounted for 12 percent of the population but more than half of the net "undercount" of 3.2 million U.S. residents.
Big cities led by New York have sued to force the government to use statistical methods to adjust the 1990 figures to eliminate the undercount.
Mayor Schmoke told a congressional hearing last month that the census misses two broad categories of people: "the neglected and the uncountable."
The neglected, Mr. Schmoke said, are people the census overlooks because of mistakes -- errors perhaps inevitable in a mammoth undertaking that one census watcher compared to mounting the invasion of Normandy with an army of six-week temporary workers.
"The uncountables," the mayor said, "are the invisible Americans. We keep them out of our consciousness most of the time. So it comes as little surprise that they're not to be found when we set about looking for them once every 10 years."
How can future censuses be taken to cap costs, increase accuracy and count the "uncountables"?
For starters, Mr. Schmoke and others favor statistical adjustment of the count.
"We have reached the point where scientific projections are more accurate than counting," the mayor testified.