LIMIT CATEGORIES OF QUESTIONS IN DECENNIAL

CENSUSES

TUESDAY, OCTOBER 24, 1967

HOUSE OF REPRESENTATIVES,

SUBCOMMITTEE ON CENSUS AND STATISTICS OF THE

COMMITTEE ON POST OFFICE AND CIVIL SERVICE,
Washington, D.C.

The subcommittee met, pursuant to notice, at 10:14 a.m., in room 346, Cannon House Office Building, Hon. William J. Green (chairman of the subcommittee) presiding.

Mr. GREEN. The hearing will come to order.

Our hearing this morning is on a bill submitted by Congressman Betts, H.R. 10952.

(H.R. 10952 follows:)

[H.R. 10952, 90th Cong., first sess.]

A BILL To amend title 13, United States Code, to limit the categories of questions required to be answered under penalty of law in the decennial censuses of population, unemployment, and housing, and for other purposes.

Be it enacted by the Senate and House of Representatives of the United States of America in Congress assembled, That (a) subchapter II of chapter 5 of title 13, United States Code, is amended by inserting immediately after section 141 thereof the following new section:

"§ 141A. Limitation on categories of information required under penalty of law in certain censuses

"(a) In the conduct of any census under section 141 of this title, information required to be furnished under penalty of the provisions of section 221 of this title shall include only matter within the following categories:

"(1) name and address;

"(2) relationship to head of household;

"(3) sex;

"(4) date of birth;

"(5) race or color;

"(6) marital status; and

"(7) visitors in home at the time of census.

"(b) Refusal or neglect to furnish information not within the categories listed in subsection (a) of this section in connection with any census conducted under section 141 of this title shall not be an offense under section 221 (a) of this title; but nothing in this section shall be construed to prevent the Secretary from requesting any such information in connection with such census on a voluntary basis.".

(b) The table of contents of subchapter II of chapter 5 of title 13, United States Code, is amended by inserting

"141A. Limitation on categories of information required under penalty of law in certain censuses."

immediately below

"141. Population, unemployment, and housing.".

SEC. 2. Section 221 (a) of title 13, United States Code, is amended by striking out "(a) Whoever" and inserting in lieu thereof "(a) Subject to section 141A of this title, whoever".

(1)

Mr. GREEN. The purpose of his bill is to amend title 13 of the United States Code to limit the categories of questions required to be answered under penalty of law in the decennial censuses of population, unemployment, and housing, and for other purposes.

Congressman Betts first appeared before this subcommittee on June 21 of this year when he requested an opportunity to be heard on this bill, as we were conducting a hearing regarding the questions to be asked on the 1970 census. Since then he has made a number of speeches on the floor of the House and has been active in a number of different ways to develop support for his bill. We plan to hear from several witnesses this morning and to engage in a discussion of the need for this bill. Congressman Betts has asked to submit a supplementary statement in support of his bill, and we are most pleased to have him here this morning.

This committee, I feel sure, shares his concern for the right of privacy and the inclusion of unnecessary questions in the census. In fact, prior to this committee's review, it was planned, or at least it was in the discussion stage, for the 1970 census to ask everyone for his social security number, religious affiliation, and political affiliation. Those three questions, through the work of this committee, will not appear on the 1970 census.

We are most interested in hearing your views, Congressman Betts, and are delighted to welcome you this morning. Please proceed.

TESTIMONY OF HON. JACKSON E. BETTS, A REPRESENTATIVE IN CONGRESS FROM THE STATE OF OHIO, ACCOMPANIED BY G. RUSSELL PIPE, LEGISLATIVE ASSISTANT

Mr. BETTS. Thank you. First, I want to tell you and the committee how grateful I am for arranging this hearing. I appreciate it very much. You have been very kind and courteous to me and I want that to be on the record.

Mr. GREEN. Thank you.

Mr. BETTS. Mr. Chairman, when I appeared before this distinguished committee in June, I presented several reasons for urging a change in 1970 census plans. I hold the view that the constitutional intent of the census that of counting the people to determine congressional districting is not being followed, particularly with the extensive, mandatory questions now included on census forms. Because many of the questions asked in a decennial census are of a very personal nature, I contend this violates the privacy of our citizens and in most instances for no public purpose. I believe Congress should take an active role in determining what types of personal information will be sought from the American people. For that reason I introduced H.R. 10952 as a means to focus attention on this important subject.

Let me also on this occasion make clear several possible misconceptions as to my interest in the census. First, I am not against providing adequate information for the Federal Government. I do not oppose the Bureau of the Census or plans to conduct a mid-decade census for I respect the competence of the Census Bureau and voted for establishing a mid-decade census. Finally, I am not fighting to weaken census plans and procedures; on the contrary, I want to streamline the 1970 census forms so the maximum number of persons will

participate in providing accurate data on themselves and their households.

The crux of the issue we consider today is embodied in H.R. 10952. My bill will keep the mandatory provision under which the census has been conducted for many years. However, citizens would be required to answer questions on only seven subjects: Name and address, relationship to head of household, sex, date of birth, race or color, marital status, and visitors in home at the time of census. H.R. 10952 removes the penalty provision for any additional questions the Director of the Census wishes to present on a second voluntary portion of the questionnaire. This is a simple, workable approach, it seems to me. I have submitted to the committee a statement on the feasibility of this mandatory-voluntary plan. This statement shows that the Census Bureau itself already receives a high level of cooperation from citizens in its many periodic surveys and samplings which are voluntary. A study of the 50 State statistical gathering programs revealed that in only two States was population information sought under penalty of fine or imprisonment for noncompliance. The States engage in considerable data collection and have found no need for compulsion to secure satisfactory results. Inquiries were mailed to major private market research organizations in the United States. Private companies, of course, must rely on the voluntary cooperation of the public for the success of their survey work. Presidents of these commercial organizations were nearly unanimous in their support for the voluntary approach to securing information from the public.

As we ponder the problems the Census Bureau will face in conducting the 1970 census of 205 million Americans. I would advance three additional reasons for limiting the mandatory questions and designating any other inquiries as important but not subject to punishment if not completed.

1. In 1970 some 60-65 percent of the people will receive their census forms by mail. The Toledo Blade of Toledo, Ohio, in an editorial on October 13, 1967, made this pertinent analysis:

Of course, Americans have been accustomed to mandatory questions in the past Federal censuses. But an important difference in the 1970 one is that it will be conducted mostly by mail. Without the subtle pressure of the personal interview by the door-to-door census-taker of the past, many citizens may indeed just fail to send the 1970 form back to the Census Bureau daring it to come find them and haul them into court.

Chances are that it will get more cooperation—and therefore more reliable information on all those personal matters if, for a change, it asks rather than tells people to answer the questions.

Mr. Chairman, public cooperation will play an exceedingly important part in the success of this mail-out census. I can see every advantage to asking rather than demanding people to answer the questions. 2. The element of distortion becomes more vivid when people complete the census questionnaire without an enumerator present and return it to census officials by mail. Some market research experts caution against compulsion for all questions as a negative reaction may result in considerable false or erroneous information being provided by citizens.

3. The necessity for a complete count and Congress interest in keeping census costs to a reasonable level is another reason for adopting the type of census questionnaire I propose. A short, simple form will

get maximum results completely and promptly. This will require a limited number of interviewers to make household visitations or telephone for further details.

These are some of the reasons for my requesting favorable consideration of H.R. 10952 by this committee. I want to underscore the point that this bill is simply a vehicle for examining other possible approaches which might solve the objections to present 1970 census plans as I see them. In the 4 months since I met with you, the interest in the census, its questions and mandatory features has far exceeded my expectations. Let me reflect on the areas of major concern which have been brought to my attention from people in nearly every State in the Nation.

1. You are familiar with the privacy issue. Questions like these taken from the New Haven pretest are of most concern:

(If a woman) How many babies has she ever had, not counting stillbirths? Have you been married more than once?

Did your first marriage end because of death of wife or husband?

Where did you live in April 1962?

What was your major activity in April 1962?

Place of birth of parents?

What is the value of this (your) property?

What is your rent?

Last year, 1966, what did sales of crops, livestock and farm products amount to?

Did you work at any time last week?

These and other proposed subjects are private matters and should not be part of a survey, particularly a national census.

2. The threat of a $100 fine and 60 days in jail is considered both an insult and an intimidation by many Americans. Since this penalty is more of a harassment than a method of actual enforcement, I see no reason for maintaining it for this entire census questionnaire.

3. We all recognize that the first requirement of a census is to find every person. The long, complex form, many nonessential questions and several very personal inquiries will discourage maximum cooperation. The mail-out method is new and can have some operational problems without being burdened with confusion and hostility by many citizens. The success of the 1970 census depends on early and accurate returns. Why gamble that success to get facts on some extraneous questions?

Mr. Chairman, there is little doubt in my mind that if present 1970 census plans are executed, the results can be chaotic. I do not make this assertion without having made a careful analysis of reports on the 1960 census and proposals advanced by the Census Bureau for 1970. Let me compare these two national censuses.

In 1960 the census of population and housing was conducted completely by interview which required the hiring of 160,000 enumerators to visit the households of 179 million residents of the United States. According to official reports of the Census Bureau, the 1960 census failed to count 5,702 million persons, or 3.1 percent of the population. This compares with a 5,675 million population undercount in 1950, or a 3.6-percent understatement. An official of the Census Bureau, Jacob S. Siegel, has reached these conclusions regarding the 1960 census:

The 1960 census failed to count a substantial portion of the nonwhite population in the United States. It is clear that a disproportionate share of the omis

sions occurred in large cities. A highly conjectural inference may be made that the enumeration of Puerto Ricans and other population groups concentrated in the deteriorated sections of our large cities was also rather defective.

It is quite probable that serious distortion does occur in the figures for many smaller geographic units within the country, particularly units in cities (such as census tracts, congressional districts, and enumeration districts) where Negroes, Puerto Ricans and other relatively poor minority groups are concentrated.

Let us look now to 1970 census plans. The Census Bureau has concluded that a mail-out, mail-back procedure will be used in the major metropolitan areas, which include 60 to 65 percent of the total population of the United States. The use of the postal system to deliver census forms will result in cutting the number of enumerators from the 160,000 in 1960 to 125,000 in 1970, even though the number of people to be counted will rise from approximately 180 million to 205 million. The use of data processing equipment has resulted in some changes in the form in that respondents are filling in dotted lines to expedite automatic tabulation. The number of questions in the 100-percent census has increased from 17 to 21 subjects; the 20-percent sample questionnaire affecting some 16 million households, now has as many as 67 subjects rather than 56 subjects in 1960. The 1970 questionnaires are all to be delivered in a 1- or 2-day period, maximizing promotional efforts, but will spread thin the personal attention which can be given city dwellers receiving the mail-type questionnaire.

What conclusions do these facts lead to? It certainly requires a basic education for innercity families to successfully complete the census without the help of an enumerator. I asked the Education and Welfare Division of the Library of Congress what educational level they would estimate necessary to complete the long-form questionnaire used in both New Haven and North Philadelphia. Mr. Francis Crowley of the Legislative Reference Service, after consulting an expert in the field, reported to me that few people with less than a 10th grade education could complete the form. Mr. Ellsworth Tompkins, executive secretary of the National Association of Secondary School Principals, responding to the same question, indicated that at least an eighthgrade education would be needed to understand most parts of the questionnaire. There were an estimated 17 million adults over age 25 in 1960 with only an eighth-grade education. How are the innercity poor, Negroes, and Puerto Ricans going to be counted if they can't fill out and return the census form?

I think the length and complexity of the 1970 form for the 20percent sample questionnaire is self-evident to anyone who has tried to complete it. No one who has attempted this task has been able to fill out the form completely in under 30 minutes. Thirty minutes of interest and concentration is a lot for many people, especially those who have difficulty with reading. Add sensitive, personal questions or inquiries which will take research such as the value of your property or amounts of income, and you further jeopardize maximum participation.

Mr. Chairman, if millions of citizens are uncounted in 1970, think what effect this will have on congressional districting, State legislative apportionment, allotments of billions of dollars in Federal grantin-aid programs to States and cities and the power minority groups. are calling for today. The dubious validity of the count will be minor compared with the tremendous impact it will have politically and economically.