census than in 1960, unless the public is deliberately influenced to be uncooperative.

The propaganda directed at the question on bathroom facilities in an effort to point up the impertinence of some census questions is a case in point. The fun that reporters and cartoonists have had with this question just goes to show what can happen when the facts are not checked. Taking on faith that the question is worded: "Do you share your shower?", they have concluded, and rightly, it is no business of anyone's that you do or do not shower with another.

If they had seen the census schedule, they would have learned that the question was misstated and misrepresented to them. The exact wording is: "Do you have a bathtub or shower?

Yes, for this household only

Yes, but shared with another household

No bathtub or shower"

The purpose is not to satisfy bureaucratic curiosity but to provide one of the best available indices of housing quality. It is almost identical to the corresponding question included in the 1940, 1950, and 1960 censuses, and it yields very useful statistics for government housing programs.

ALL QUESTIONS ARE IMPORTANT

Critics of the 1970 census plans have supplied a wide choice of "improvements"; almost as many and as varied recommendations have been offered up as there are critics.

They would eliminate many of the questions; as many as one-half of the questions on population, and as many as three-fourths of the questions on housing. The critics would allow only a minimal number of questions, something under ten, on a 100-percent mandatory basis, and rule out many of the rest. For the few remaining, they would reduce the sample size below the limit of statistical usefulness, and complete the destruction by placing them on a voluntary, throw-away basis.

Omitting what the critics term nonessential questions, or relegating them to inadequately small samples, would result in chaos for the administration of many Federal programs. As an example of the dependence on census statistics by administrative agencies, here is what the Secretary of Agriculture says:

"The Department of Agriculture determines the eligibility of rural areas to receive loans, grants, and other benefits made available by the Area Redevelopment Act which provides that areas be made eligible which ‘are among the highest in numbers and percentages of low-income families, and in which there exists a condition of substantial and persistent unemployment or underemployment.' It also specifies that the 'number of low-income farm families be considered, as well as the extent of migration out of the area.' "The Manpower Development and Training Act of 1962 has specific provisions for unemployed and underemployed workers in low-income farm families. Basic figures on the occupational and education characteristics of farm population are important to the furtherance of the objectives of this act." And the Secretary of Housing and Urban Development has this to say about that department's use of census statistics:

"There is at present no other source for the detailed information on housing characteristics and the correlation of these characteristics with population data outside the Bureau of the Census. Hence the Federal Housing agencies draw heavily from census data and require applicants for Federal housing and community development aid to document their cases with census data.

"The most urgent statistical need for housing and community development planners and administrators is for data for small areas including census tracts and blocks. It would be undesirable to conduct a census of housing that failed to yield the information most needed by its major users." Some critics of the 1970 plans, in the interest of reducing costs and burden, would defer some of the questions to a series of voluntary intercensal sample surveys and other questions to surveys conducted by State or local governments. As for intercensal sample surveys, it is well known that they can serve useful purposes, but they cannot serve as a substitute for a census because they can provide only limited data for large geographic areas. Thus they would fail in their prime purpose-to meet the statistical needs of Federal and local governments for such areas as cities, counties, Congressional districts, census tracts,

city blocks. Nor would they provide the detail required for the Nation, States, and large metropolitan areas that the much larger 25-percent sample within the census can provide.

The truth is that when questions are dropped, or sample size reduced, in the national census and then covered in separate surveys or censuses, the usefulness of the data is severely reduced because there is no standardization and the data from the different surveys or censuses cannot be interrelated. Nor is there any "savings of millions of dollars." The real situation is that a series of sample surveys, of any adequacy at all, carried out separately from the census would cost considerably more than the all-in-one census and would, in addition, be far more burdensome to the public.

The Census Director of the State of Massachusetts, which is one of the two States that take a limited mid-decade population census, said in his testimony before the Subcommittee during hearings on the Federal Mid-Decade Census: "The method we use to take a census in Massachusetts, in my opinion, does not compare with the method which the Federal Government uses. The Federal Government is better equipped to conduct a census."

At the same hearings, the Mayor of Durham, North Carolina, made this statement:

"The Federal Government has already recognized the national character of many of the problems which face local governments and the stake of the Nation in preserving vital municipal centers. Congress has wisely, we think, enacted housing legislation to stimulate new home construction, urban renewal, slum clearance, public housing, and residential conservation and rehabilitation. Federal laws now aid municipal governments in the fields of airport construction, hospital construction, community facilities planning and construction, sewage treatment, and other programs.

"There can be no doubt that prudent investments of vast sums of public money should be made with the best possible data in mind. Simple economics requires that we build facilities which will serve their intended purpose for the maximum time. Adequate census data is the cornerstone of this effort."

CENSUS IS CONFIDENTIAL

And now I want to scotch the worry we have heard the loudest cries aboutinvasion of privacy.

It may not flatter you to know that the Census Bureau is not interested in you as an individual, but such is the case. Your name, address, and phone number are on the census form for your use and to enable the census taker to call back if necessary to complete the census. When the census form is returned to the Bureau, the information is put on computer tape, but your name, address, and phone number are omitted: thus, the data included in the files are wholly impersonal and unidentifiable.

Every employee of the Census Bureau, including temporary workers, must take the oath of confidentiality, and is subject to imprisonment and fine for violation of the oath. Penalties for disclosure are severe. There is no known case of any census employee violating his oath of confidentiality.

Your census report cannot be used for purposes of taxation, investigation, or regulation. The information may not be released to other government agencies or to anyone else for any purpose whatsoever.

The copies of the census forms are kept under lock and key. The only person who can obtain the information about you is yourself or your designated legal representative. This census service has proved a life-saver for millions of Americans who had lacked necessary vital documents. In the last two years about oneand-a-half million persons turned to the Census Bureau, asking that a search be made in past census records to prove their age for Social Security or Medicare Benefits, their place of birth for citizenship, their family relationship for a rightful inheritance, and for other purposes.

I do not want anyone to misunderstand me. I am confident that the critics of the census are sincere in their concern about invasion of privacy. I am worried about it, too. It is something we should be concerned about. But the critics of the Census Bureau have the saddle on the wrong horse. Its reputation for the preservation of confidentiality, both in the letter and the spirit, is a byword.

The Congress has already expressed its concern in this area as regards the real violators of privacy: the credit bureaus, financial houses, employment agencies, lie detector operators, private investigators, psychological testers, who not only obtain the most personal information, but exchange it with other firms. This is

where our efforts to protect the individual's privacy should be directed, not toward the Census Bureau, which maintains inviolate the privacy of the individual.

MAIL METHOD

With the mail-out mail-back method to be used to enumerate most of the population in 1970, the census gives even further protection to private information as most households will receive their census forms by mail to fill out in the privacy of their homes, and mail back to the Census Bureau.

This mail method is a logical extension of Census Bureau technology. It has been used for years to take the economic censuses and many other surveys. It was used to a limited extent in the 1950 population and housing census and more widely in the 1960 census. But it has never been used on the scale planned for 1970 simply because complete, reliable resident address lists were not previously available.

Some critics say, however, that this mail method spells nothing but trouble, and is just another reason why results of the census could be chaotic. They evidently have not heard about the testing that has taken place. The Census Bureau has been testing mail methods for the past seven years. A series of population and housing censuses trying out questions, techniques, methods-in fact, every aspect of census taking-has been conducted in various kinds of areas ranging from small towns to large metropolises.

Experience has been gained and techniques polished in the successive tests, until now plans for the Nineteenth Decennial Census are in the final stage and are being tested in three "dress rehearsal" censuses planned this year: Sumter and Chesterfield Counties, South Carolina; Dane County (Madison metropolitan area), Wisconsin; and Trenton, New Jersey.

As for respondent cooperation in the previous seven test censuses, results speak well for public acceptance. Altogether, upwards of 70 percent of the households returned their questionnaires. In the two largest areas the response rate was very high: 88 percent for the Louisville, Kentucky metropolitan area (covering three counties); and 81 percent for the city of Cleveland, Ohio.

The Bureau knows there will always be some people who will be laggard, or whose educational level precludes their ability to answer the questions, or who are physically handicapped, such as the blind or the aged. Enumerators will call on those who do not return their questionnaires or who return incomplete ones. Experience has shown that often a telephone call will provide the missing information and save the cost of a personal visit.

The mail method will not be used in most rural areas and small urban communities in 1970 because address lists suitable for census purposes are unobtainable. Personal house-to-house enumeration will be used in these areas as in 1960.

TESTING THE METHOD

Critics also noted that there was a significant undercount of some groups in the 1960 census, and say that, after their perusal of the content and methodology planned for 1970, they can only conclude that there will be an even greater undercount.

But it was the Census Bureau itself that pointed out the problem and reported on the 1960 undercount and much of the Bureau's effort in preparing for the 1970 census has been directed toward finding means of enumerating these “lost Americans."

In several test censuses, particularly in that part of North Philadelphia, which is known locally as "The Jungle," much valuable experience was gained. Also, the Census Bureau has had the advice and suggestions of sociologists, welfare workers, minority group leaders, and others. Some of their suggestions have been tried and found helpful.

Now, if these critics who seem to be bubbling over with new ideas for taking censuses, would come up with some effective ways to gain the trust of persons who fear being counted because they are wanted by the law, or because they are living in a household where the welfare agency believes no man is present, or who are militantly antagonistic toward the Government, they would make a constructive rather than a destructive contribution to the census.

The Bureau knows there is no simple way to reach these people. But through trial and error, and through the help of public-spirited residents in these hardto-enumerate areas, the Bureau is making progress in finding ways to obtain their cooperation.

The Census Bureau is continuously searching for the best, most lucid wording of questions so that they can be readily understood by persons of limited education. Efforts are being made also to provide special assistance to the inner-city residents. However, in these areas, the final completeness and accuracy of the census does not depend upon the initial mail response, but upon the intensive follow-up made by trained field workers. Savings in personnel requirements through the use of the mail in easy-to-enumerate areas will enable the Bureau to concentrate canvassing activities in difficult areas to an even greater extent than ever before.

The questions, techniques, and procedures planned for the 1970 census have gone through a long and comprehensive period of testing and review. No question is on the schedule unless the information it will supply is needed and requested by the Government. The fact that the information is also useful to the private sector is an added benefit, and since it is taxpayers' money that pays for the collecting and processing, the Government should make the information available to anyone who can make use of it. But no question appears on the schedule simply because it serves the interests of non-government users alone.

This is something I want to underscore. Through the years, the Census Bureau has strongly and effectively resisted the addition of questions where the needs and uses nationwide, not local or special-have not been amply demonstrated and proved.

COOPERATION IS GOOD CITIZENSHIP

A Presidential Proclamation issued at the time of the 1960 census sets forth the purpose of the census and the uses to be made of the information. It said, in part:

"The sole purpose of the Census is to secure general statistical information regarding the population, its characteristics and its homes. Replies are required from individuals only to enable the compilation of such general statistics. No person can be harmed in any way by furnishing the information required

...

"Life and liberty in a free country entail a variety of cooperative actions for the common good. The prompt, complete, and accurate answering of all official inquiries made by Census officials should be regarded as one of the requirements of good citizenship."

This statement applies equally to the 1970 census. I am appending a list of the questions planned for inclusion in the 1970 census. With each question is an explanation of the need for the data which answers to that question will provide. The 1970 census will count as many as 210 million persons and 70 million housing units, the biggest decennial census of all time. It is the Nation's basic statistical program from which most of our knowledge about the population and their housing must come. The Census Bureau needs the confidence and the support of all of us in order to make it a success, and all valid evidence indicates that it will be a success. It would be catastrophic if misguided anti-census and anti-statistics propaganda reduces the fine cooperation this Nation has historically given its census takers.

I know only too well how little time we in the Congress have to give to research; with so many demands made upon us, it is impossible for us personally or our staffs to dig out all the facts about every cause before committing ourselves to it. As a result, we are vulnerable to being misled by misinformation, no matter how well-meant.

It is for this reason that I have searched out the facts about the census and made an effort here to answer those who appear to find so much wrong with it. I hope my words will be taken in the light in which they were offered-a sincere and honest attempt to look objectively and with a broad view at the 1970 census plans.

As useful as census statistics have been in the past, all indications are that they will become even more vital in the future. America's social and economic problems are going to confront us with more and more urgency in the years ahead. We dare not create havoc in our planned census programs. America would lose in the end.

We have a simple choice: knowledge or ignorance. Experience has long since shown that it costs too much to know too little.

(The list of the questions planned for inclusion in the 1970 census, referred to above, follows:)

NEED FOR ITEMS ON THE CENSUS SCHEDULE

POPULATION ITEMS

Relationship to head of household (100 percent)

Collected in each census since 1880. This information provides the basis for classifying the population into families and other residential groupings, and for characterizing families in terms of their structure and composition. It is fundamental to all statistics showing characteristics of the family and basic to the consideration of social and economic problems which affect the family as a group. Much of the recent discussion about the association between the "broken" home and poverty is based on the information about family structure collected in the

census.

Color or race (100 percent)

Collected in each census since 1790. Information about the racial composition of the population is of fundamental importance in analyzing the data on families, school enrollment, fertility, and economic activity. All of these data become much more meaningful for the planning, administration and evaluation of programs and for demographic, economic, and sociological analyses when they are classified by race. Objective information on the numbers, location, and characteristics of persons identified with the several racial groups is essential to the Government, as well as to agencies which are concerned with ameliorating the conditions of Negroes, Indians and other ethnic groups.

Age and sex (100 percent)

Collected in each census since 1790. Most social and economic characteristics of the population-family relationship, school attendance, labor force participation, occupation and income are closely related to age and sex. Thus, in any comparison of relative social and economic status among various segments of the population, a knowledge of the age and sex composition is essential. For some subjects such as labor force, unemployment and school enrollment, the figures are meaningful only for persons in selected age groups. The age data are also directly used for planning major government programs in health, education, and social security.

Marital status (100 percent)

Collected in each census since 1880. This question shows whether an adult has married and, if so, whether he is still married or has become separated. divorced, or widowed. The data are used by research workers in Federal agencies and elsewhere to analyze such subjects as the number of single men subject to military draft, dependency involving widowhood and orphanhood, employment of wives and mothers, and the level of fertility.

State or county of birth (25 percent)

Collected in each census since 1850. The information on State of birth is useful in measuring the streams of internal migration from certain regions of the country to other regions, for example, Negroes from the South to the big cities of the North and West.

The Government finds information on country of birth essential in formulating immigration policy and in considering immigration laws. The information is also used by organizations concerned with the welfare of various ethnic groups and provides a measure of assimilation when the population, classified by nativity the country of origin, is tabulated by such characteristics as fertility, education or occupation.

Years of school completed (25 percent)

Collected in each census since 1940. The educational level of the population is an important measure of the quality of the manpower pool. Federal, State and local agencies concerned with manpower training and development need this information as the starting point in all program planning. Education exerts a major influence in the social and economic condition of the individual. It also indicates the output of the educational system. Statistics on almost every item in the census are tabulated by education (as well as age, color and sex) because of the light these cross-classifications throw on the relationship between successful completion of various levels of schooling and eventual sucess in other types of life adjustment.