Confederacion De La Raza Unida v. Brown, 345 F. Supp. 909 (N.D. Cal. 1972)

Robert L. Gnaizda, Jo Ann Chandler, Albert F. Moreno, Public Advocates, Inc., San Francisco, Cal., Mario Obledo, Miguel Mendez, Mexican-American Legal Defense & Educational Fund, San Francisco, Cal., for plaintiffs.

*910 U. S. Atty., James L. Browning, Jr., and Asst. U. S. Atty. Brian B. Denton, San Francisco, Cal., for defendants.

PECKHAM, District Judge.

Plaintiffs in this class action are individual Mexican-American residents of Santa Clara and San Benito Counties, California, as well as organizations representing the interests of Mexican-American persons in those counties. On behalf of all Spanish-origin persons, they seek a preliminary injunction restraining publication of the results of the 1970 census until, inter alia, a statement is added to all releases dealing with Spanish-origin persons to the effect that there may have been an under-count of such persons. Jurisdiction exists under 28 U.S.C. §§ 1331 and 1337.

Plaintiffs' claim is that Mexican-American and other Spanish-origin persons were miscounted by the 1970 census, either because they were not counted at all or because they were counted but not noted as being Mexican-American or of other Spanish-origin. The alleged undercount, plaintiffs contend, will cause them great and irreparable harm, since census data are used in disbursing federal funds to the poor and minorities, in allocating federal jobs to minorities, and in apportioning Congressional districts, as well as in formulating remedies in employment discrimination cases brought under Title VII of the Civil Rights Act of 1964.

The legal theories upon which plaintiffs base their claim for relief take varying forms; but they all reduce to the contention that plaintiffs have been denied various rights without due process of law by the conduct of the 1970 census. The rights which plaintiffs claim to have been deprived of are, inter alia, their right to vote fully and effectively in federal elections, their right to be counted in the decennial enumeration (U.S.Const. Art. I, § 2, cl. 3 and 13 U.S.C. § 141), their right to be free from discrimination on the basis of national origin in the administration of federal programs (42 U.S.C. § 2000d), and their right to benefits from certain federal programs which disburse their largesse on the basis of the number of poor and minority group persons in a given area (e. g., 42 U.S.C. §§ 1751-63; 2571-2628). Of course the recent Supreme Court decision in Goldberg v. Kelly, 397 U.S. 254, 90 S. Ct. 1011, 25 L. Ed. 2d 287 (1970) makes it clear that the due process clause protects individuals from deprivations of such important statutory rights as well as from deprivations of constitutional rights.

Apparently, the parties do not disagree over the requirements of due process in the context of the taking of a census. They concur with the court in City of East Chicago v. Stans, No. 70 H 156 (N.D.Ind., Oct. 21, 1970), as does this Court, that a census count satisfies the due process clause if its methods are "nondiscriminatory and reasonably adapted to the end sought to be achieved." Plaintiffs do not contend, and correctly so, that they have an absolute right to be counted.

In considering claims of the sort made in the instant case, the court in City of East Chicago, supra, declined to enjoin publication of the census results for that city. In reaching its conclusion that due process had not been violated in the conduct of the East Chicago census, that court made the following findings of fact:

5. The methods used in taking the census were designed to achieve the most accurate population count possible in view of the Census Bureau's resources, and the inherent difficulty of the task. They were devised with considerable care by recognized experts in the field of demography, and based on extensive research over the last ten years.

6. The methods used in taking the Census were not adopted arbitrarily, capriciously, or irrationally. They were well adapted to the end sought to be achieved. There was no discriminatory *911 intent on the part of the Bureau or its employees, and a special effort was made, in the face of apathy and at times even hostility, to reach members of minority groups which, in the Bureau's estimation, seemed particularly vulnerable to an undercount.

7. The methods used in taking the census in East Chicago, Indiana, were the same as those used nationally. Those officers and employees of the Bureau responsible for taking the census in East Chicago exercised reasonable care and achieved as accurate a result as the Bureau's techniques permitted.

8. While the Bureau's final population figures for the City of East Chicago undoubtedly fail to include some of its residents, that failure is the result of the inherent impossibility of achieving a perfect count without depriving citizens of their constitutional rights, and not of any material deficiency in the methods used by the Bureau or their application by its employees.

9. The allocation of additional resources to the Bureau of the Census would not have produced any significant improvement in the accuracy of its count.

Are the methods used by the Census Bureau to count Spanish-origin persons amenable to the same characterization as the methods used to count the largely non-Spanish-origin population of East Chicago? Substantially the same procedures were followed by the Bureau in Santa Clara and San Benito Counties as in East Chicago, except for some special measures taken to obtain Spanish-speaking enumerators in San Benito County.^[1] What this Court must consider, then, is whether the special educational^[2] and language barriers to counting Spanish-origin persons render constitutionally inadequate enumeration methods which otherwise are adequate as applied to other groups in the country. An ancillary matter which the Court must also consider is whether the question which was intended to identify members of the Spanish-origin minority group is so misleading *912 and ambiguous as to lead to a substantial failure to label all the members of the group as such.^[3]

Plaintiffs have presented essentially two types of evidence in support of their contention that the methods used by the Bureau were substantially likely to lead to an undercount. The first type consists of estimates by governmental entities, other than the Census Bureau, of the size of the Spanish-origin population in the nation or in the two California counties from which plaintiffs hail. While all but one of these estimates show the Spanish-origin population as larger than the Census Bureau figures indicate, the Court does not find this evidence sufficient to establish that a violation of due process has taken place. Either these estimates are not significantly greater than the Bureau's statistics,^[4] or else they were made on the basis of methods which this Court either cannot determine or finds unreliable.^[5]

The second type of evidence consists of affidavits from individual Spanish-origin persons as well as results of informal surveys taken in Spanish-origin and Anglo neighborhoods to determine the quality of individuals' contacts with the census. This evidence purports to show that because of the language barrier between many Spanish-origin persons and the census representatives, or because of such representatives' failure to assist individuals who had difficulty filling out the English-only forms, many Spanish-origin persons either threw the forms away without communicating the information to anyone connected with the census, or else filled the forms out incompletely without obtaining the assistance of a census representative.

In his deposition, however, Associate Director of the Census for Demographic Fields, Conrad Taeuber, explains how many of these recorded complaints do not necessarily establish that the people in the complainants' households were not counted or were not counted as of Spanish-origin. While his explanations do not neutralize all the lapses uncovered in *913 plaintiffs' affidavits and surveys, they at least demonstrate that plaintiffs would have to undertake a much more sophisticated set of surveys in order to prove that the census did not and could not have accurately counted a substantial fraction of the Spanish-origin population.

The final item of evidence offered by plaintiffs to establish that the Bureau's methods violated due process belongs in a class of its own. Apparently in October, 1971, the Census Bureau released a "definitive" report on the Spanish-origin population which showed a 274,000 decrease in the population over a sixteen-month period. This decrease is belied by the increasing birth rate among Spanish-Americans relative to Anglos as well as by the increasing immigration from Mexico, neither of which defendants dispute.

In fact, Dr. Taeuber concedes that there may have been no decrease at all in the Spanish-origin population. But he does not thereby admit any defects in the Bureau's enumeration techniques. The totals for all groups which were counted, he notes, were subject to the standard deviations which may always result from use of statistical sampling techniques; but so long as the sampling techniques were no less likely to obtain raw data about one group than about another, no one was disadvantaged by their use.

The only ground for criticizing the October, 1971 release which Dr. Taeuber acknowledges is the ground of failure to note in a prominent place that the statistics are subject to standard deviations. This objection has recently been met, however, by methodological background information provided in the Bureau's current releases. See Defendants' Exhibit G, Appendix C.

Accordingly, this Court finds that the evidence which plaintiffs have assembled does not cast sufficient doubt on the accuracy of the count of Spanish-origin persons in the 1970 census to warrant a conclusion that plaintiffs are being deprived of any rights without due process of law.

Neither does the Court find any substantial likelihood that Spanish-origin persons will not be identified as such by the 1970 census. Plaintiffs' objection is that question 113 on the 5% questionnaire, which asks the householder to state whether his "origin or descent" is Mexican or Puerto Rican, contains serious ambiguities. According to plaintiffs, a member of the Mexican-American minority group could very well be the American born child of two American born parents; yet he would have to answer that he was not of Mexican "origin or descent." Apparently, plaintiffs believe that the questionnaire could have been constructed to allow all members of the Mexican-American minority group to so identify themselves, perhaps by including a question which asks the house-holder whether he is Mexican-American or Chicano.

Defendants have endeavored to identify any Mexican-Americans who answered "no" to question 113, however, by compiling statistics about the languages other than English spoken in individuals' homes when they were children and about the number of individuals with Spanish surnames. None of these statistics, it should be noted, is more "official" than any other. Thus, given the availability of these alternative means of identifying Spanish-origin persons, this Court does not find it substantially likely that the census failed to label a significant number of Spanish-origin persons as such. And absent such a likelihood, there can be no deprivation of rights without due process of law.

The sum total of evidence which plaintiffs have amassed against the census to date does not rise to the status of a denial of due process of law. It follows that plaintiffs' chances of success on the merits are not sufficiently high to warrant the granting of a preliminary injunction. It is therefore ordered that plaintiffs' motion for a preliminary injunction is hereby denied.

[1] According to defendants, three different types of questionnaires were distributed in the 1970 census, and every household received one of the three. The short form questionnaire sent to 80% of all households included a limited number of questions. These questions plus more were sent to an additional 15% of all households. And the 164-question questionnaire about which plaintiffs have the most complaints was sent to 5% of all households. The last questionnaire included most of the questions in the first two plus more. All forms were in English; and the determination of which form would be sent to any particular household was made in advance in accordance with scientific sampling techniques.

Two methods were used to retrieve information from the survey questionnaires. In large metropolitan areas, including Santa Clara County, the census developed an address list and sent out an appropriate questionnaire to each address with instructions that it be filled out and returned. An individual called an enumerator was sent to any household which either failed to return the form or filled it out incompletely. In the rest of the country, including San Benito County, the postal service delivered a questionnaire to each household, with instructions that the form be filled out and retained until an enumerator contacted the household. The enumerators then went door to-door to complete and collect the forms. If after three visits the enumerator was unable to establish contact with the household, he was instructed to obtain his information from someone else, usually a neighbor or a relative living nearby.

Defendants claim that attempts were made to find local people to serve as enumerators. And interpreters were available for enumerators who were not familiar with the local language. In addition, the enumerators' qualification test was waived for Spanish-speaking applicants in San Benito County. Defendants cannot state, however, how many enumerators in San Benito and Santa Clara Counties actually were bilingual or were accompanied by interpreters. Deposition of Conrad Taeuber, Associate Director for Demographic Fields of the Census Bureau, Defendants' Exhibit A.

[2] Defendants apparently do not dispute plaintiffs' contention that almost one-half of all Spanish-surnamed adults have not gone beyond the eighth grade.

[3] Various methods were used to determine the ethnic identity of persons who were counted by the census. In both the 5% and 15% questionnaires, households were asked where each member was born. In the 15% questionnaire, households were asked where each member's mother and father was born, as well as what language other than English was spoken in each member's home when he or she was a child. With respect to persons of Spanish origin in particular, question 113 on the 5% questionnaire asked each member of the household to state whether his "origin or descent" is Mexican or Puerto Rican. And in the five Southwestern states of Arizona, California, Colorado, New Mexico, and Texas, an examination was made of every surname, and those which were identified as Spanish were separately coded.

[4] For example, defendants' 1970 count of the Spanish language population of California by County shows that such persons constitute 18% of the population of Santa Clara County and 45% of the population of San Benito County. Plaintiffs' statistics from the California Department of Education show that Spanish-surnamed persons constitute 16.9% of the population of Santa Clara County and 54.4% of the population of San Benito County. The Bureau claims that the Department of Education's figures are inflated because they are based on the percentage of school children in these counties, and Spanish-surnamed Americans are more likely than members of other groups in these counties to have large families. Plaintiffs claim that the Department's figures are an underestimation because Spanish surnamed students are more likely than students from other groups to drop out of school at an early age in order to work. Given the small discrepancies between the Bureau's and the Department of Education's percentages, as well as the lack of statistical data to support plaintiffs' interpretation of the Department's figures, this Court cannot find that the Department of Education's statistics establish a likely undercount of Spanish-origin persons.

[5] For example, plaintiffs point to a recent estimate by the President's Cabinet Committee on Opportunity for the Spanish-Speaking, which pegs the Spanish-origin population at a figure which is five million higher than the Census Bureau's most recent figure. But plaintiffs have not outlined the methods used by the Committee in reaching its estimates.