Officials Nationwide Heaping Discredit on 1990 Count : Census: An undercount is being charged. Congressional seats and billions of dollars in aid to cities hang in the balance.

Relying on independent data, their own estimates and gut instincts about urban populations, city officials in California and across the country are heaping discredit on the 1990 Census, charging that preliminary results of the census are too low--by anywhere from 2% to 12%.

Billions of dollars in aid to cities, as well as seats in the U.S. Congress, rest on the outcome of the census--an increasingly difficult task to count one by one a mobile and uncooperative population of 250 million people.

The final census will be reported to President Bush by the first of the year.

City officials said the figure for Los Angeles--3,420,000--could be as much as 10% off, and New York City officials say the bureau’s count of 7,033,000 could be 1 million in error. Other cities ranging from San Francisco to Chicago to Detroit to Houston echoed similar complaints.

The big losers could be major industrial cities of the Northeast and Midwest. Although cities in rapidly growing states such as California stand to gain much because of flight to the Sun Belt over the last decade, few cities even in those areas are satisfied with the government’s figures.

Officials in Los Angeles, New York and other cities contend that the preliminary figures suffer from the same flaw that beset the 1980 Census--a 5% or greater undercount of non-English-speaking immigrants, illegal aliens and homeless people.

City officials believe that as these populations have swelled over the last decade, the undercount has become more pronounced.

In response to these complaints, the Census Bureau stresses the preliminary nature of figures released earlier this week, and emphasized that canvassing is still going on. Bureau officials also said that local authorities will have a chance to challenge the figures over the next two weeks.

City officials don’t put much stock in the challenge because under the law they are not allowed to contest the number of people, just the number of households.

If the census has failed to accurately count the number of people living in the houses, there is no way to correct that mistake, the cities argue. And in many large cities with poor and immigrant families doubling up, it is easy to make mistakes.

In addition, local officials across the country said the bureau’s tally of housing units does not match their own records drawn from such sources as public utilities and building permits and the cities’ own estimates.

A high-ranking census official said Friday that communities should not be alarmed if the population figures seem low.

“People are reacting to the population count,” said Edwin B. Wagner Jr., assistant division chief for operations at the Census Bureau in Washington. “We provided that for information only,” and those figures are expected to climb as a result of the ongoing canvassing, he said.

“The real focus needs to be to review these counts of housing units” by cities that feel they are being undercounted, he said.

That is not always easy, some local authorities countered, especially in cities such as Los Angeles that have a high number of illegal immigrants and others who would prefer that the government not know they are here.

In Los Angeles, the census has missed “about 18,000 units,” said Jessica Heinz, deputy city attorney. Heinz also believes the census listed a lot of housing units as vacant that are actually occupied--including garages and facilities that house large numbers of illegals--but the census does not allow provisions for local authorities to challenge the number of units listed as vacant.

The count “is off (low) by as much as 10%,” said Heinz. That “undercounting” could cost the city millions in both federal and state funds.

“We figure a person is worth about 65 bucks a year in state money” alone, Heinz said.

Adding to the confusion is the state Department of Finance’s figures. For example, the department’s estimate of Los Angeles’ population is close to the census count. But department officials question the accuracy of their figures for some of the same reasons that the census figures have been questioned.

According to Linda Gage, demographer for the Department of Finance, the state’s estimate does not count homeless people and does not accurately measure the number of people in a given household.

“Neither one is accurate,” Gage said, referring to the state count and the preliminary census. “Ours is an estimate and subject to all the errors you can make when you’re making an estimate. And the census bureau number is preliminary and will certainly change.”

One city, Detroit, is in a uniquely difficult position.

Juliet Okotie-Eboh, chief of data analysis for Detroit’s Planning Department, said the census is about 140,000 too low, and that could translate into a devastating loss for the city. Detroit, by far Michigan’s largest city, benefits from state legislation that was tailored specifically for cities of a million or more inhabitants. The preliminary census puts Detroit’s population at 970,000.

That drop, from 1.2 million in 1980, will cost the city dearly if it cannot be boosted by 30,000, and Okotie-Eboh said she believes the city will succeed.

“We’re confident we’re going to get it,” she said.

Census officials expect the numbers to change, partly as a result of “post-census local review,” the process that allows local officials to challenge the number of housing units in the preliminary count. But that task is made more difficult by the fact that the bureau does not provide local officials with addresses to protect the confidentiality of the census. So instead, local governments are given the number of housing units for specific blocks within the area.

It is then up to local authorities to compare that number with their own records.

But some feel that is not a satisfactory approach and will still end up falling far short of the actual number of residents. That is of particular concern to New York City, which has lost a million residents during the last decade, according to the latest census.

Following the 1980 census, the “post-census local review,” as the housing verification is called, made very little difference, according to David Golden, assistant counsel for New York City.

“It added something like 76,000 people nationwide to the overall count,” Golden said. “And it reallocated only 56,000 people geographically.”

“That was the extent of its impact,” he said of the review process.

Local authorities across the country are hoping for far more, and they are compiling as much evidence as they can get in hopes of proving the bureau wrong.

Joseph Chow, assistant director for planning and development for Houston, said he could not understand how the federal count should have been as far off as his records indicate.

“I have found they have misreported some 10,000 units in one southwest area alone,” Chow said. He said he believes the bureau may have missed 200,000 units, including “588 units on one block which they identified as three units,” he added.

“They just missed it.”