and satisfactory estimates of the population for States and in some smaller areas.

The CHAIRMAN. Senator Ervin?

Senator ERVIN. I understand that you are compiling a record of the votes of each county in the United States?

Mr. ECKLER. Yes, sir, Mr. Senator, we are obtaining that information. It is available in official form in Washington. The Governmental Affairs Institute has long been active in assembling the official reports from the States on the vote cast for counties and we are getting the information through their courtesy and we expect to supply to the committee in response to the chairman's request information on that subject.

Senator ERVIN. Do you not know that there are two States out of the six involved which would be covered by this formula in this bill; namely, the States of Mississippi and Virginia, which have their State elections and the local elections in odd years rather than the presidential years?

Mr. ECKLER. I am not personally acquainted with the details on the voting arrangements in the several States, Mr. Senator. I do know that this varies from State to State. What we have is the presidential vote, vote for the presidential election, in 1964. Senator ERVIN. Where do you live?

Mr. ECKLER. I beg your pardon?
Senator ERVIN. Where is your home?
Mr. ECKLER. New York State.

Senator ERVIN. But you live in the District?
Mr. ECKLER. That is right, sir.

Senator ERVIN. I thought maybe you might live over in Virginia. That is the reason for that question. I know that Virginia has its election for Governor and the State legislature, all local offices, in odd years and not in the presidential year. Do you believe that it is fair to place upon Virginia a test that applies only to a presidential year when nobody is running except candidates for President and candidates for U.S. Senate and candidates for the House of Representatives?

Mr. ECKLER. Mr. Senator, I regard that as something which the Attorney General ought to comment on. It is outside of my competence to comment on the applicability of this particular feature. Senator ERVIN. Well, do you know that from your observation that the people who really get voters to come out to vote are the ones running for local office? They are the ones who spend the time and energy and the money to get voters to come out and vote, aren't they?

Mr. ECKLER. Again, I think it is outside my competence, although it would seem to me

Senator ERVIN. That is within the competence of any intelligent

man.

Mr. ECKLER. It would seem to me that the amount of interest which attaches to a presidential election is extremely great and that this is an Outstanding event. I assume that these other elections, in many cases, have extremely important local issues and do bring out a great deal of interest and a great deal of local enthusiasm.

The CHAIRMAN. Of course, you know, do you not, that the vote is gotten out by the local candidates. You know that, do you not?

Mr. ECKLER. I assume that-again, I am not an expert on voting or on what brings out the vote. I assume that in some cases, the local candidate is extremely important. In other cases, a Governor may be extremely important, it may be an important fight. It seems to me always the President is an important election.

Šenator ERVIN. Do you not know that local candidates and their friends haul people out to their polls. You never saw a President haul people out to the polls on election day, did you?

Mr. ECKLER. No, sir.

Senator HART. Mr. Eckler, for your comfort, in Michigan, the uniformity of the election in a Presidential year brings out many more people than in the off year, although we have the governor running both times.

Mr. ECKLER. I believe that is also true in New York State.

Senator KENNEDY. Mr. Eckler, that is substantially the same in Massachusetts.

Senator HART. Mr. Eckler, you are familiar with the charge that would be imposed upon the Bureau of the Census by the Senate bill 1564, the legislation we are considering.

Mr. ECKLER. Yes, sir, Mr. Senator.

Senator HART. It would require that you determine the percentage of persons of voting age residing in a State or a county and the percentage of persons who voted in the November election last year? Mr. ECKLER. Yes, sir.

Senator HART. Do you have any doubt about the capacity of the Bureau accurately to report those figures?

Mr. ECKLER. We believe that we can provide that information in acceptable form for the purposes of the bill. There will be some work we need to do which we have not done. We have provided the House an estimate of the cost of the additional work, but we believe we can do it satisfactorily; yes, sir.

Senator HART. What, if anything, have you done under the charge of title VIII of last year's civil rights bill?

Mr. ECKLER. Our work under that has been limited to the preparation of a statement of the cost of doing certain work which the Civil Rights Commission transmitted to us as a request. Title VIII does provide that we do work in areas specified by the Civil Rights Commission. They did so specify. We prepared a statement which was submitted to the House and submitted it to the Congress. We had hearings before the House about a month and a half ago and we are awaiting further action.

Senator HART. Thank you.

Senator KENENDY. Mr. Eckler, if the Bureau of the Census were asked to survey a given political subdivision to determine the registration on the basis of color, what factors would determine the cost? Mr. ECKLER. The cost?

Senator KENNEDY. And time necessary to make the compilation, and is there a significant difference between a subdivision, a city, or a county, of 22,000 and one of 9,000?

Mr. ECKLER. Do I understand the question to refer to a survey or a census in essence?

Mr. ECKLER. The cost would depend very directly upon the population of the area to be covered. That is the most important determining element in such a job. If we had the necessary pretesting of this, the developing of procedures and so forth, which might take a number of months, the carrying out of the survey once that were established-and this is assuming not that this were on a wholesale basis, but scattered-if this became a very extensive load so that it was going on in a great many areas at one time, then we face a problem of organization that is quite substantial. But if these were scattered operations similar to the special censuses which we conduct, presumably the work in a particular area would be-perhaps we could work out a way that we could get to it within the period of 60 to 90 days after the time that the request was made.

The measurement of this information may involve certain complications. If it refers to November 1964, and if, under the provisions of this particular bill, something like registrars had started to operate and registration was going on, there might be confusion as to whether the infromation referred to the 1964 situation or to registration that had taken place since then. I think there are some problems of that sort that might require a special study and testing.

Senator KENNEDY. I am thinking in terms of the number of political subdivisions. You have been charged in title VIII, which Senator Hart mentioned, in the 1964 bill-you have extensive responsibilities under this proposed legislation. I am wondering what is going to be the time factor, when are we going to be able to have these figures in order to make the triggering devices of this legislation active? If it is a question of cost, do you have recommendations which you ought to be considering in order to make this a realistic kind of proposal? And is it a question of additional personnel? We should know this as well.

Or are you completely satisfied that under the 1964 act and the charges under 3 (A), that you will be able to fulfill this mission without additional personnel or additional appropriation?

Mr. ECKLER. Senator Kennedy, as far as the 1964 act is concerned, there is a specific request so that we assume that whatever work is done there would be on the basis of appropriations made and on the basis of timing which we indicated to the Appropriations Committee. The ability of the time required, of course the time of the availability depends very directly upon when the money becomes available. As was pointed out at the time we appeared before the committee, something like 6 months of testing time is required and then after that, several months for the collection. So that would put the availability of that information into the early part of 1966. I think that is the most favorable assumption of timing.

As far as the present Voting Rights Act is concerned, we have assumed thus far and the discussions seem to confirm this, that what is involved is a determinaton of a relationship on the basis of projections of population for counties in States that are involved. The six States in which-which come under the voting device criteria and which have less than 50 percent, it seems to be an interpretation that those as a whole would be certified and that work on the individual counties might not be necessary. Outside of that, there would be a

considerable number of counties in other States that met their criteria where we would need to prepare these projections of county populations.

The sum required, which we estimated and made available to the House Committee was a total of about $75,000. It would take 2 or 3 months, perhaps, to do that work. So this is not such a long-drawnout job as would be involved if surveys were to be taken to determine a population figure.

Senator KENNEDY. Do I understand you correctly that you suggest that with the political subdivisions as defined in this legislation, that you will be able to ascertain accurately what the breakdown is with regard to white and nonwhite in a period of approximately 90 days? I do not want to pin you down, but I want to ascertain that. Mr. ECKLER. By means of a census?

Senator KENNEDY. By means of a census.

Mr. ECKLER. This would be true only if we had done the necessary testing in advance of the techniques for doing this. This is a kind of survey which we have not previously done in exactly this form.

Senator KENNEDY. Now, would you describe this? What is the necessary testing that must be done in advance?

Mr. ECKLER. We should want to develop a series of questions, a questionnaire, in order to get this information that is needed, and we would want to use it in the field in several different situations in order to evaluate the ability of this particular questionnaire and the techniques to elicit reliable information. One of the problems, as has been brought out before this in other testimony, is that there is some tendency in surveys of this sort for people to state that they are registered or to state that they voted to a greater degree than the actual registration would bear out. We need to do the best we can to find procedures which get the most nearly accurate response possible.

Senator KENNEDY. Well, now, under existing procedures, what would be the time it would take to ascertain the necessary information under 3 (a)?

Mr. ECKLER. Under the procedures and the responsibilities that we assume fall upon us in the present wording, this would not involve any canvassing of population. This would involve using the 1960 records and birth and death figures, internal migration, other information, in order to prepare statements of the voting age population of the necessary counties.

Senator KENNEDY. How long would that take?

Mr. ECKLER. Perhaps 2 or 3 months after the funds were provided. Senator KENNEDY. Well, so, under any estimate, you feel the Bureau of Census can make an accurate determinaiton 2 or 3 months after the funds are available in these political subdivisions of 3(a). Mr. ECKLER. Yes, sir.

Senator KENNEDY. Now, if this legislation were to be expanded to include other political subdivisions which have not been outlined in this bill, but would include areas in which there was a significant question as to the number of nonwhites which were registered to vote. would you feel that these additional areas or political subdivisions which perhaps might be included in this legislation-would you feel that a similar determination could be made of these political subdivi

as I imagine it would to some extent-personnel and appropriations to do the job?

Mr. ECKLER. Senator Kennedy, I assume that these political subdivisions in some instances would be counties rather than smaller areas. Senator ERVIN (presiding). If I may interrupt, that is one of the things which shows that this bill was brought in great haste. No one took the pains to define a political subdivision. A political subdivision is not confined to counties. It would include all the municipalities, the cities and towns, the school districts, it would include the wards and where the town or city was divided into several wards, it would include each of those wards. In North Carolina, it would include hundreds of special tax districts and school districts where people vote. If it would take 2 or 3 months to get the figures for the counties, it would probably take you 2 or 3 years to get the rest of them for the smaller subdivisions.

Senator KENNEDY. Mr. Eckler, we have had the Attorney General define political subdivision, certainly to my satisfaction and I feel to others as well, and I think for purposes of your response, you have outlined it to my general satisfaction. What I was attempting to ascertain is whether you felt if there were a modest increase in the total numbers of areas of counties or political subdivisions that were included in this legislation, you feel that you would be able to meet these additional responsibilities?

Mr. ECKLER. Well, Senator Kennedy, in terms of that question, I believe that the additional number would not create a very serious further load. While there might be some additional sum that would be required, I do not think that the timing that I suggested before would be particularly changed. I think this would be feasible, because these are more scattered situations, I take it, pockets where problems exist and I think we could include those.

You did mention nonwhite. I assume you meant merely that the total would be determined and that the low percentage is due to the fact that a significant number of nonwhites are actually not registering and not voting. We would not, by means of this projection process that I described, be able to get a color subdivision.

Senator KENNEDY. That is correct.

Mr. ECKLER. But I think the answer to your question is that this would not seem to add significantly to the problem outlined.

Senator KENNEDY. Could I ask you, what has been your experience on the validity of figures which have been made available by several of the States-as to their accuracy and as to their breakdown on race and other factors?

Mr. ECKLER. Senator Kennedy, we have given, of course, a great deal of attention to this matter of improving these projections. We were looking at the record. In terms of the average difference between the actual census and the projection on the basis of the current methods, it appears to be something like a 1-percent average deviation over a period of 10 years. Now, this is a period shorter than 10 years, so that the average deviation should be smaller, significantly, than the 1 percent. Furthermore, I think that many of these States which are involved in the estimation work here are States which have had a significantly better than average record on estimates in the past.