three public members, and three Cabinet members. The total should come to 26. So it is an ongoing, permanent, bipartisan, independent commission representing all levels, but responsible in a sense to all of them. We have instituted an effort in the last 3 years to procure funding from the State and local side to get an intergovernmental mix fiscally, and we have had some measure of success in that area. Mr. WILSON. Fine. Perhaps through your Senate Members we can get a little assistance in getting the Senate committee to move our mid-decade census bill if we do get such a bill out of the House. That has been our problem in the past. Mr. WALKER. We will try our best. Mr. WILSON. We know there have been problems, but we do not have any postal legislation this year-other than the little bill I introduced to repeal the Postal Reform Act-so if we can get Senator McGee to hold hearings and take action on this matter, I think we can get a 5-year census bill this year. Mr. WALKER. That would be good news. Mr. WILSON. Do you have any questions, Mr. Chappell? Mr. CHAPPELL. Mr. Walker, are there facilities available which could be reasonably utilized on the part of the State which would contribute substantially to a feasible reduction in the cost of such a census? Mr. WALKER. I cannot speak for the States on this. You heardMr. CHAPPELL. I was not asking whether they would be willing, I was asking what facilities would be available. Mr. WALKER. As Mr. Olson indicated, there is a considerable body of expertise in State planning departments. In many-not all-but in many State planning units at the present time, there is a considerable body of expertise that I think could be of assistance. They probably would be willing to provide assistance to the Bureau of the Census in any quinquennial census. Mr. CHAPPELL. What about technical services, or computers, for example? Mr. WALKER. Some have this and some do not. There is a wide range between and among the States in their reliance upon computers. Some put heavy reliance upon them. Mr. CHAPPELL. What part do you think the computers, for example, could play in such a census? Mr. WALKER. I could not really, as a layman in this particular area, answer that question competently. I will have to beg off and plead ignorance. Mr. WILSON. As far as computers go, I think the Census Bureau. probably has about as advanced a system as there is in the country. Mr. WALKER. Right. Mr. WILSON. I would doubt the Census Bureau would have to de pend very much upon utilizing computer equipment over what is already available to them. Mr. WALKER. Yes, sir. Mr. WILSON. Thank you very much, Mr. Walker. Mr. WALKER. Thank you, Mr. Chairman. Mr. WILSON. We appreciate your being with us this morning. Our next witness again is a gentleman who has been with us on other occasions and who has always contributed greatly to what ever hearing he appears as a witness. He is Dr. Carl H. Madden, the Director of Economic Analysis and Study for the Chamber of Commerce of the United States. STATEMENT OF DR. CARL H. MADDEN, DIRECTOR OF ECONOMIC ANALYSIS AND STUDY, CHAMBER OF COMMERCE CF THE UNITED STATES Dr. MADDEN. I am Carl H. Madden, Chief Economist, Chamber of Commerce of the United States. It is a privilege to appear before this distinguished committee as it considers mid-decade census proposals. The Chamber of Commerce of the United States is the Nation's largest business federation. We have 39,000 business firms as members, around 2,700 local and State chambers of commerce, and more than a thousand national trade association members. The Chamber of Commerce of the United States supports a program of mid-decade censuses beginning in 1975 and to be conducted every 10 years thereafter, a program in addition to the regular decennial census, to include population, employment, and housing. There are many needs for a mid-decade census as we see it. Only if population in small areas is accurately known, can repreentation proportioned to population be achieved in State legislatures, County and city councils, school boards, and other local bodies. The Supreme Court doctrine of "one-man, one-vote" makes representation an important continuing question in a mobile society. More than $10 billion annually is now allocated by the Federal Government to the States or their subdivisions according to formulae using population or housing data as a major component. Additional large amounts are distributed by States to their counties and municipalities largely on the basis of population. Both Federal and State Governments use population, employment, and income data in a variety of ways to develop, analyze, and administer government policy. In States, the legislative power, taxing authority, and sometimes even the form of government that may be adopted by counties and cities depend on the population of the jurisdition. Some types of banking regulation, including bank reserve and capital requirements, the power of a political subdivision to license or tax businesses, occupations, or products, school policy in purchasing and many other exercises of governmental authority may deInd in some States on population. At the Federal level, a multitude of policies ranging from tax policy to antipoverty policy depend on accurate calculation of population estimates. At both Federal and local levels, the need for small-area data analysis, and the ability to manipulate data in analyzing the hypothetical Consequences of policy actions, are converging through the demand for gher standards of competence in policymaking. Much better and more frequent information on both internal and external migration is needed. The results of the 1970 census, for example, dissipated the false assumption, based on sampling and estimation, that the migration north to the cities of poor and less educated people had ended. Given the increasing ease of communica tions and travel, legislators at all levels of government need correct information on internal population movements to make good policy. External migration data also need strengthening to measure changing values of people. Is it correct that more people are moving from the U.S. to Canada than the reverse today, as news media report? In a country which throughout its history has taken pride in serving as a haven of freedom and opportunity, such reversals, if they exist, in the trend of external migration are sensitive indicators of the state of opinion which political leaders must consider both pragmatically and philosophically. There is a need for better social data. There is a growing need for better data and more frequent data useful in measuring many types of social change so that lawmakers and government administrators make informed decisions in the field of governmental social policy. The advent of the computer makes it possible for a city, local jurisdiction, combination of local jurisdictions, or even a State or the Nation to calculate the hypothetical impact of changes in many different areas of policy. For example, it is now possible to calculate in detail the impact on a city of a change in taxes or in the supply of public housing, or in many other variables, through the use of a mathematical, computer-based model of the dynamics of change in the social entity under study. The need for small area data on population, income, and employment or of crime, health, and so forth, associated with population, is growing. Very large economies can be achieved in government by modeling hypothetical policy changes-economies derived from avoiding technical errors and economies derived from correct diagnosis of policy needs. A mid-decade census would contribute to the data base needed for improved modeling of hypothetical social change. The need for better sampling bases Many government programs not based on data from decennial censuses are based instead on sample inquiries. But sample inquiries, in turn, depend for their accuracy on correct definition of parent populations from which samples are drawn. As the data from a census become obsolete with the passage of a decade, the value of the census as a sampling frame declines. A mid-decade census would yield benefits to all levels of government and to business-and therefore to owners of stock and to consumers-through the cost-saving which goes along with improved accuracy of sampling. Mandatory response to census questions Mandatory response to census questions is desirable for technical reasons. Without the requirement to be counted, complete enumeration would be difficult. The requirement to answer questions truthfully reduces to a minimum biases that would result from nonresponse or false answers. The mandatory requirement to respond is more equitable, but it is also statistically preferable to permitting voluntary nonresponse. If census queries are voluntary, then statistically the population enumerated is voluntary respondents to the census, not the U.S. population. Mandatory response is useful in generating persistence in enumerators in finding the hard-to-enumerate. A mandatory census seems well within the power of Congress to ask and the citizen's duty to answer. Even though the census has never breached confidentiality defined most stringently, the Government should continue to take the strongest measures to assure confidentiality of information about individual people in the census. The Government should respect and remain empathetically sensitive to the U.S. public's fears about the invasion of privacy. And although the individual may be required to answer questions reasonably related to legislated purposes, the questions and the way they are asked have to accord with people's sense of propriety, on the grounds of common civility as well as to avoid increasing the cost of census-taking. Lerel of penalty for nonresponse The level of penalty for nonresponse should be sufficient to make mandatory answers effective but not out of proportion to the relative seriousness of intentional nonresponse. Geographic level The mid-decade census should enumerate the U.S. population, and therefore it should be developed and designed by the same census process as is used by the decennial census. Scope and content of questionnaires A mid-decade census should be less extensive than the decennial census in subject matter and should give first priority to demographic elements population, population characteristics, migration patterns and behavior. The U.S. business community has developed an elaborate and highly complex system of information in analyzing the demand for and supply of products and services in a trillion-dollar economy. U.S. business uses information in vast quantities daily to reduce the risks of loss inherent in a world of uncertainty. Business uses census data as benchmarks; it uses this data as it uses any other data it can develop-because it is relevant and available. Business, however, was generally satisfied with the 1970 census questions, and its interest in either adding or subtracting questions is at a low level. Surveys of business opinion concerning the 1970 census reflected that even experts, such as marketing analysts and economists, take the census for granted but at the same time process such large amounts of information that the census as such is perceived as only one among thousands of data sources. In some ways this may not be wholly desirable, but it is the fact as we have been able to determine it. Thank you very much for this opportunity to testify. I notice that it is your feeling that the scope of the census could be lessened for the mid-decade census as compared with the decennial census, in contrast to at least one other witness who recommends as complete a census for the mid-decade series as for the decennial census. Dr. MADDEN. The chamber of commerce as such has no opinion on this. My personal opinion is not that of an expert, but for what it is worth, it would seem as if the mid-decade census, if it were to establish correct population totals for small areas and were to involve the basic questions on population and housing, income and employment, might then serve as the parent population for sample surveys which could explore the more detailed aspects which are covered in the decen three public members, and three Cabinet members. The total should come to 26. So it is an ongoing, permanent, bipartisan, independent commission representing all levels, but responsible in a sense to all of them. We have instituted an effort in the last 3 years to procure funding from the State and local side to get an intergovernmental mix fiscally, and we have had some measure of success in that area. Mr. WILSON. Fine. Perhaps through your Senate Members we can get a little assistance in getting the Senate committee to move our mid-decade census bill if we do get such a bill out of the House. That has been our problem in the past. Mr. WALKER. We will try our best. Mr. WILSON. We know there have been problems, but we do not have any postal legislation this year-other than the little bill I introduced to repeal the Postal Reform Act-so if we can get Senator McGee to hold hearings and take action on this matter, I think we can get a 5-year census bill this year. Mr. WALKER. That would be good news. Mr. WILSON. Do you have any questions, Mr. Chappell? Mr. CHAPPELL. Mr. Walker, are there facilities available which could be reasonably utilized on the part of the State which would contribute substantially to a feasible reduction in the cost of such a census ? Mr. WALKER. I cannot speak for the States on this. You heard Mr. CHAPPELL. I was not asking whether they would be willing, I was asking what facilities would be available. Mr. WALKER. As Mr. Olson indicated, there is a considerable body of expertise in State planning departments. In many-not all-but in many State planning units at the present time, there is a considerable body of expertise that I think could be of assistance. They probably would be willing to provide assistance to the Bureau of the Census in any quinquennial census. Mr. CHAPPELL. What about technical services, or computers, for example? Mr. WALKER. Some have this and some do not. There is a wide range between and among the States in their reliance upon computers. Some put heavy reliance upon them. Mr. CHAPPELL. What part do you think the computers, for example, could play in such a census? Mr. WALKER. I could not really, as a layman in this particular area, answer that question competently. I will have to beg off and plead ignorance. Mr. WILSON. As far as computers go, I think the Census Bureau probably has about as advanced a system as there is in the country. Mr. WALKER. Right. Mr. WILSON. I would doubt the Census Bureau would have to de pend very much upon utilizing computer equipment over what is already available to them. Mr. WALKER. Yes, sir. Mr. WILSON. Thank you very much, Mr. Walker. Mr. WALKER. Thank you, Mr. Chairman. Mr. WILSON. We appreciate your being with us this morning. Our next witness again is a gentleman who has been with us on other occasions and who has always contributed greatly to what |