Advertisement

SKIP ADVERTISEMENT

Census Bureau Reports It Overlooked 5,300,000 People in 1970

Census Bureau Reports It Overlooked 5,300,000 People in 1970
Credit...The New York Times Archives
See the article in its original context from
April 26, 1973, Page 93Buy Reprints
TimesMachine is an exclusive benefit for home delivery and digital subscribers.
About the Archive
This is a digitized version of an article from The Times’s print archive, before the start of online publication in 1996. To preserve these articles as they originally appeared, The Times does not alter, edit or update them.
Occasionally the digitization process introduces transcription errors or other problems; we are continuing to work to improve these archived versions.

WASHINGTON, April 25—The Census Bureau said today after a reanalysis that it had overlooked an estimated 5.3 million Americans in the decennial Census of 1970.

Bureau statisticians pointed out that the estimated error —approximately 2.5 per cent of the previously reported total of 203,235,000 people counted — was significantly smaller than errors in the two previous national head counts. But it acknowledged the increased importance of accurate figures now that they are the basis for distribution of millions of Federal dollars through revenuesharing programs.

Despite the acknowledgement of an undercount today, Census officials say the new figures will not be used to change the population figures upon which revenue shares are allocated or the apportionment of population in designing districts for elected offices.

The reanalysis was done primarily as a check on the techniques used in an effort to refine them for a more accurate count in 1980.

Because the bureau is allowed by law only to report people counted, the final 1970 total population figure will not be revised on the basis of today's estimate of undercounting.

Even though census data are the basic tools used in hundreds of governmental decisions —ranging from allocations of revenue sharing funds to apportionment of election districts —the bureau says it is not possible on the basis of the reanalysis to read just the population figures.

Adjustments Ruled Out

Reliable figures in internal migration and other factors that have changed since the 1970 Census are such, they say, that an adjustment of specific city, county or state figures is not possible without having an entirely new Census.

Although an analysis of the undercount by the bureau disclosed that two–thirds of the number missed were whites, those blacks missed in the counting amounted to a much higher rate (7.7 per cent of the total) because of the smaller size of the total black population.

The unusual announcement of the analysis, unrequested, of its own work by a Government agency was made today by the acting director, Robert L. Hagan.

“The 1970 Census was probably the best ever taken,” Mr. Hagan said, “but, like its predecessors it was imperfect.”

Because of better counting techniques, such as integrating postal data, introduced since the 1960 Census, he added, it is estimated. — based on the 1960 experience—that about 2.3 million people who would have been missed were included in the original 1970 Census and helped reduce the final margin of error.

The 5.3 million estimate of the number of persons missed in the count is not a fixed figure but what the bureau calls “the best estimate” within a range of error that extends from 4.8 to 5.8 million people.

The undercounting rates for the 1960 and 1.950 censuses — the only others subjected to detailed analysis — were 3.3 per cent and 2.7 per cent respectively.

Key elements in the reanalysis of the 1970 Census included the following:

¶The undercount for whites was 3.45 million (1.9 per cent) and 1.88 million for blacks (7.7 per cent).

¶The only large segment of the black population in which the undercounting in 1970 was worse than in 1960 was in the coverage of black children under 10 years of age. The omission rate there was 8.6 per cent, compared with 5.3 per cent in 1960.

¶The bureau was unable, because of no reliable data, to make any estimate on the number of Spanish‐speaking Americans missed in the 1970 Census.

Because of the composition of the undercounted population — largely black and presumably urban — statisticians conceded that some areas were “more undercounted than others.” But, they add, their information indicates that the most undercounted area could not be more than 3 or 4 per cent of the total reported figure.

The estimate of undercounting was arrived at by Census statisticians by calculating the “expected” total population as of April 1, 1970, from other sources.

Such sources included checking birth records, Medicare and Social Security applications, immigration and death statistics. By a series of such analysis, the bureau was able to come up with total population figures that could be checked against those produced by the actual census.

Problem Factors

The actual count figures are always questionable, these officials point out, because of certain factors that have always hidden people from Census takers.

Among the factors that made the 1970 Census vulnerable to an undercount, they say, are the following:

¶lncreased resistance to Census takers because of changing life styles and more alienation and distrust of authority.

¶The existence of a number of organized attempts to protest the Census as an invasion of privacy.

¶The reluctance of some Census takers to work, especially at night, in some urban areas.

Although Census officials declined to discuss details of such things as “resistance to Census takers,” it was clear they were talking about families who wished to avoid the inclusion in the count of draft‐age sons and of welfare recipients reporting male residents.

They were at a complete loss to understand the wide margin of error in reporting the number of black children under 10 years of age.

“As far as we can determine,” one official said, “there is no reason whatsover not to report such children and the only theory we have now is that something in the way the questionnaires were made out prevented an accurate count.”

The reanalysis of the census did not give an estimate of the undercount of Spanish‐speaking Americans because of unreliable basic data. Birth records kept by the states usually indicate a white or black birth, but in many cases do not identify parents as Spanish‐speaking. Birth records are a basic tool in the reanalysis of census data.

Advertisement

SKIP ADVERTISEMENT