Lawsuit calls for accurate counting in 1990 census. Major cities seek adjustments to avoid minority undercount

Almost as soon as the 1980 census was over, controversy began over how the 1990 census would be tabulated. But reforms are still on hold as the decennial count of Americans approaches. At issue is whether Census rules undercount minorities. Critics of Census Bureau plans want the final population totals to include an adjustment for people missed in the survey.

Efforts to adjust the census were turned down by the Reagan administration and passed over by Congress. An adjusment measure died in the last Congress.

The matter has now gone to court. Earlier this month, New York City filed a lawsuit seeking to force the bureau to make an adjustment to avoid a minority undercount. (Blacks look for political home with clout, Page 5.)

The lawsuit (filed on behalf of Los Angeles, Chicago, Miami, The National League of Cities, and the United States Conference of Mayors) confirmed a prediction by a census committee staffer that the thorny issue will not fade away.

Adjustment would entail a survey taken after the 1990 census that would take samples of the population and match them against the original count. Those who are missed represent the undercount that would be added to the original survey. Undercounting is a special concern because census data determines US House and state legislative district boundaries. It is estimated that New York City alone was undercounted by about 500,000 in 1980.

Democratic congressmen from urban areas have the most to gain from adjustment. They argue the undercount robs minorities of equal representation.

Adjustments have been done in every census since 1950 but never have the results been added to the original count.

The representatives supporting adjustment are largely from states with large populations. They say the nearly 3.2 million Americans missed, living in the nation's barrios, ghettos, and urban centers, are additionally shortchanged when federal program funds are allocated. The money is distributed in proportion to population data.

``With $50 billion in federal funding to states directly linked to census data, with resources tougher to get, elected officials are very sensitive about the accuracy of the count,'' says Census Bureau spokesman Jim Gorman.

Blacks and Hispanics were undercounted by as much as 6 percent in the 1980 census, which was conducted primarily by mail.

Barbara Bailer, a former statistician at the bureau who quit because of the political wrangling over adjustment, says that when testing the post-census survey, interviewers discovered many reasons for the undercount.

In Mississippi many blacks lived in hard-to-reach backwoods areas where interviewers were ``afraid to go into.'' In Los Angeles they found Hispanics living in converted garages and vans, places where the ``mail didn't come to,'' she says.

In city housing projects, families were ``doubling up'' because of homelessness and feared being counted because they were violating federal housing regulations. People thought the information would be shared with other federal agencies like Housing and Urban Development and the Immigration and Naturalization Service, Ms. Bailer says.

Lastly they found an attitude of ``Why should we be counted? It's not going to help our lives.''

The bureau rebuffs adjustment formulas as ``error prone,'' pointing out that opinions vary on what equation best estimates the number of persons missed. Bailer labels this a ``weak argument'' because a formula is already used to adjust for some obstacles in the original count.

During congressional hearings, Rep. Robert Garcia (D) of New York accused the bureau of knuckling under to the Reagan administration and opposing adjustment. Robert Ortner, undersecretary of commerce, took strong exception, saying an elaborate mathematical system that would allow numbers to be ``made up'' would not solve the problem. The Census Bureau is a branch of the Department of Commerce.

Terry Ann Lowenthal, staff director for the House population subcommittee, says, ``It's to the advantage of certain powers to maintain the status quo. The question becomes why would anyone not support a measure to correct the situation.'' Rep. Constance Morella (R) of Maryland says a better question is why the $2.5 billion earmarked for the census can't produce accurate numbers.

The assistant director of the bureau, Peter Bounpane, says it's easy to ``make us out as the bad guys,'' but, ``it's not that simple.''

He illustrates with a hypothetical example. ``Say you measure the undercount for the state of Florida and it's 2 percent for blacks. The simplest way to adjust is to [use] a 2 percent undercount for every subarea. But it's liable to be higher in Miami and lower in Tallahassee. The assumption [is] not provable.''

Another monkey wrench is time. The bureau is mandated to begin the census in April and report to the President by December. That time frame makes it impossible to complete a post-census survey.

The bureau, however, has made other efforts to avoid an undercount. They include:

Hiring minority-owned advertising companies to launch promotional campaigns.

Encouraging church organizations to get the word out.

Starting a school project to encourage children to bring news about the census home.