From 2 million to 4 million illegal immigrants were living in the United States as of 1980, far fewer than had been estimated, and their numbers are growing more slowly than had been supposed, a research panel has concluded.

Estimates by official and private groups have ranged from 2.3 million to 20 million. The Immigration and Naturalization Service's estimate is 6 million to 7 million, and the Census Bureau offered a range of 3.5 million to 6 million for 1978.

The National Research Council, an arm of the National Academy of Sciences, reported its estimates last week as part of a blistering criticism of the "woefully inadequate" way the INS tracks "dismally limited" data on immigrants.

Because the INS has failed to produce reliable or timely data about immigration, including information about where immigrants come from and what they do, U.S. immigration policy "has been made in a data vacuum," the panel said.

The panel called the "File Room" of the INS office in New York a metaphor for the service's statistical operations.

There, immigration records of 2.1 million people are on paper files in a 300-foot room full of special metal cabinets. Because of the cabinets' design and placement, the use of one blocks access to others.

When several people are looking for records, the panel said, "mechanical gridlock" can result, sometimes for 20 minutes. With shopping carts full of records, clerks troop the aisles, and refiling records often takes three days.

The panel described the system as "Dickensian," noting that "the system for record-keeping is no better, indeed is clumsier, than that used by clerks of the Cratchit era."

The system is so poor, the panel said, that when President Jimmy Carter asked during the Iranian hostage crisis how many Iranian students were in the United States, the INS could not answer.

The report said the INS still cannot answer that question, but agency spokesman Duke Austin said last week that the answer -- 63,000 -- was supplied to the White House in April 1980, after five months of pulling paper files by hand and manually tallying the numbers.

"Data on all new immigrants is automated now," Austin said, adding that the INS spent $61 million last year on automation.

Austin dismissed the panel's estimate about illegal aliens as based on inadequate data.

"Nobody knows how many there are," he said. "If we could count them, we could catch them."

The panel's estimate involved no new effort to count immigrants, but consisted of a review of previous estimates tempered by a judgment on reliability of the methods used to produce them. All estimates "suffer from uncertainties," the panel said, noting that it used the least uncertain to develop its estimate.

The work was done by an 18-member panel of statisticians, demographers, economists, sociologists and other specialists under the leadership of statistician Burton H. Singer of Columbia University.

The panel did not deal with immigration policy questions but took the position that, without reliable data, sound policies cannot be crafted. Estimates Of Aliens Disputed Illegal-Immigration Figure Said Too High By Boyce Rensberger Washington Post Staff Writer

From 2 million to 4 million illegal immigrants were living in the United States as of 1980, far fewer than had been estimated, and their numbers are growing more slowly than had been supposed, a research panel has concluded.

Estimates by official and private groups have ranged from 2.3 million to 20 million. The Immigration and Naturalization Service's estimate is 6 million to 7 million, and the Census Bureau offered a range of 3.5 million to 6 million for 1978.

The National Research Council, an arm of the National Academy of Sciences, reported its estimates last week as part of a blistering criticism of the "woefully inadequate" way the INS tracks "dismally limited" data on immigrants.

Because the INS has failed to produce reliable or timely data about immigration, including information about where immigrants come from and what they do, U.S. immigration policy "has been made in a data vacuum," the panel said.

The panel called the "File Room" of the INS office in New York a metaphor for the service's statistical operations.

There, immigration records of 2.1 million people are on paper files in a 300-foot room full of special metal cabinets. Because of the cabinets' design and placement, the use of one blocks access to others.

When several people are looking for records, the panel said, "mechanical gridlock" can result, sometimes for 20 minutes. With shopping carts full of records, clerks troop the aisles, and refiling records often takes three days.

The panel described the system as "Dickensian," noting that "the system for record-keeping is no better, indeed is clumsier, than that used by clerks of the Cratchit era."

The system is so poor, the panel said, that when President Jimmy Carter asked during the Iranian hostage crisis how many Iranian students were in the United States, the INS could not answer.

The report said the INS still cannot answer that question, but agency spokesman Duke Austin said last week that the answer -- 63,000 -- was supplied to the White House in April 1980, after five months of pulling paper files by hand and manually tallying the numbers.

"Data on all new immigrants is automated now," Austin said, adding that the INS spent $61 million last year on automation.

Austin dismissed the panel's estimate about illegal aliens as based on inadequate data.

"Nobody knows how many there are," he said. "If we could count them, we could catch them."

The panel's estimate involved no new effort to count immigrants, but consisted of a review of previous estimates tempered by a judgment on reliability of the methods used to produce them. All estimates "suffer from uncertainties," the panel said, noting that it used the least uncertain to develop its estimate.

The work was done by an 18-member panel of statisticians, demographers, economists, sociologists and other specialists under the leadership of statistician Burton H. Singer of Columbia University.

The panel did not deal with immigration policy questions but took the position that, without reliable data, sound policies cannot be crafted.