Over and above the UN itself, the U.S. Statistical System contributes to other intergovernmental bodies in the statistical field. To name only two major bodies, the United States is substantially involved with the Organization for Economic Cooperation and Development (OECD) and the Inter-American Statistical Institute (IASI).

Added to the long list of official international and intergovernmental statistical bodies in which the U.S. Government plays a role is a companion list of professional scientific associations of international character in which U.S. Federal statisticians participate as a routine part of their professional activity. These professional societies such as the International Statistical Institute and its affiliates abound in every statistical subject matter and methodological field. They provide a kind of cement between the public and private sectors in international statistics, bringing together the official community, the academic community, and the business and industrial community. They are an important element in the international statistical community, as in any other scientific community, providing nonofficial forums in which ideas can be exchanged and the field advanced.

Still another essential ingredient of the international technical assistance system is the international donor community. These are the agencies such as the United States' own AID and the counterpart agencies in other developed countries such as: the Swedish International Development Agency (SIDA), the Canadian International Development Agency (CIDA), the British Overseas Development Ministry (ODM), the Norwegian Agency for Development (NORAD), and the foreign assistance arms of other developed nations. A number of national development assistance agencies and/or foreign ministries conduct bilateral assistance programs with LDC's and also are the financial mainstays for UN bodies such as the UN Development Programme (UNDP) and the UN Fund for Population Activities (UNFPA). These , agencies form a somewhat separate community of their own, interacting regularly with one another, providing the funds for technical assistance to developing countries in a broad range of subject fields, many of which involve statistics. Just as there is fluctuation in the meshing of goals between the 'The central importance of bringing together academic and official statisticians was underscored by the International Statistical Institute which created a Committee on the Integration of Statistics to study this problem in 1976. The Committee will develop recomendations for consideration at the 1979 biennial meeting. The American Statistical Association conducted a seminar of experts in March 1978 on the subject "Transfer of Methodology between Academic and Government Statisticians."

Federal statistical community and AID in the United States, so there is a companion interrelation between the goals of the international statistical community and the international donor community.

A particularly significant meshing of goals between the international donor community and the international statistical/demographic community occurred in the early 1970's with the establishment of the World Fertility Survey under the auspices of the International Statistical Institute. The survey is funded chiefly by AID and the UN Fund for Population Activities but it is also supported in principle by the majority of the international donor community. The World Fertility Survey is now entering its sixth year, the largest single social science research program ever undertaken. At its conclusion in the early 1980's, the World Fertility Survey will have caused approximately 65 national fertility surveys to be carried out-40 LDC's and 25 developed countries which are internationally coordinated and comparable in content and methodology, and concentrated around the single time period of roughly 1974 to 1979.

The history of the World Fertility Survey merits study for the lessons it offers for future large-scale international statistical programs to: (1) provide major technical assistance to LDC's and (2) coordinate the efforts of developed countries. The World Fertility Survey illustrates the kinds of massive financial and intellectual resources which can be marshalled for the scientific study of a major social problem transcending national boundaries, especially when a donor community and a scientific community perceive the problem with like urgency and harmony.

In considering the international statistical technical assistance system, one must also take account of the private donor community. Just as the international statistical community includes official national and international agencies and nonofficial international scientific associations, so the international donor community's official agencies have their counterparts in the private foundations. The large number of foundations in the United States alone includes a significant number which channel funds into international development assistance activities. Two of the most notable are the Ford Foundation and the Rockefeller Foundation; the former, in particular, maintains regional offices and field representatives around the developing world. Again, the linkages between the private foundations and the official donor community are as close as those between the official statistical agencies and the statistical associations.

Finally, the recipient community consists of those agencies and personnel in LDC's who are the targets of assistance for all of the above. The core of this community is the national statistical offices, but international statistical training and technical assistance are important for the full range of LDC governmental ministries. Officials from these target agencies flow into and out of the UN statistical system as employees, advisers and experts; participate in international scientific associations according to their governments' means; interact with the official and private donor community both as consultants and as recipients of assistance; and comprise an increasingly vocal segment of the international statistical community. An indicator of the recipient community's recent success in getting its voice heard is afforded by the fact that the strongest resolution recommended by the UN Statistical Commission at its Nineteenth Session in November 1976 dealt with building up a national household sample survey capability in developing countries.

Problems for the Coming Decade

International Technical Assistance and Foreign Policy

Several conclusions derive from surveying international statistical training and technical assistance in the U.S. Government. First, the Federal statistical community is not presently organized or unified in such a way as to adequately contemplate, much less recommend, action on its international aspects. International statistics is truly an issue which touches many offices, bureaus and agencies, each in a slightly different manner. For those agencies and programs which deal with statistical training and technical assistance for LDC's there is some degree of unification, but that unity is the by-product of the fact that the activities and programs are almost all sponsored by a single agency, AID. Because they have a common sponsor, the personnel involved in these programs come to know one another and to relate formally and informally.

Second, the foreign policy community appears neither unified nor sensitized to the role of statistics in international relations. The Department of State appears to assume that statistical matters are beyond its competence or purview. AID is presently organized in such a way that statistics tend to be considered discretely either by subject field, geographical region, or country. It is difficult if not impossible to find a single locus of administrative authority within AID which has cognizance for the total statistical activities sponsored by the Agency. In AID's defense it must be pointed out that the organizational forms of the

foreign assistance program are largely the direct consequences of shifting congressional mandates and directives.

Third, it is difficult to conceive of a major social or economic foreign policy issue which does not have serious statistical implications. The 1976 overview of United States Foreign Policy stresses that "for America, involvement in world affairs is no longer an act of choice but the expression of reality." It states that "food, population, energy, trade, and monetary policy-all are interrelated elements of the fundamental reality of our age, global interdependence."

Ultimately, it must be realized that a foreign policy which stresses global interdependence entails international training and technical assistance in statistics as one facet of its implementation. For example, the above-quoted document states with respect to food reserves:

Nations with a history of radical fluctuations in import requirements have an obligation, to their own people and to the world community, to participate in a system which shares the responsibility more widely. And exporting countries can no longer afford to be caught by surprise. They must have advance information to plan production and exports. In short, there must be an international reserve system.

The requirement for advance information for planning purposes translates in part into a requirement for the development of statistical systems. Countries which are characterized as less developed are countries which generally have less developed statistical systems. If the developed nations such as the United States desire that developing nations participate fully in globally interdependent systems, whether in food, energy, trade or whatever, the developed nations must eventually address the problem of assisting the LDC's to upgrade their statistical systems to the point of being able to undertake full participation. U.S. foreign policy in social and economic areas needs statistical (as well as other) resources and data to implement policy. For many LDC's the statistics simply do not exist, nor has provision been made for bringing them into existence. Hence, development and maintenance of statistical infrastructure must become a more specific objective as international development policy evolves.

'Department of State, United States Foreign Policy: An Overview/January 1976, Washington, D.C.

Need for a Continuing Resource Base

The concept of global interdependence entails the consideration that the lines between domestic and foreign policy are becoming blurred. As the Department of State puts the matter, "when our factories and farms and our financial strength are so closely linked with other countries and peoples, our prosperity is tied to world prosperity." Increasingly, our knowledge of, and ability to influence, domestic social and economic problems is bound up with our knowledge of, and ability to influence, the problems of other nations.

A considerable argument can be made for the thesis that international training and technical assistance for developing countries, in statistics as well as many other fields, are in the domestic as well as the foreign policy interests of the United States. At present, statistical training and technical assistance to LDC's is funded almost exclusively by the foreign assistance program and hardly at all by domestic programs. This is a condition which results in serious inefficiencies in the acquistion of statistical knowledge about LDC's which is important to U.S. domestic programs. For example, AID is currently organized to provide assistance to LDC's mainly by subject fields such as population/family planning, agriculture, health, education, nutrition, rural development, and urban development. While no one will dispute the importance of these subject fields for the world in general and any given country in particular, the problem is that funds are made available principally on the basis of subject fields alone. The increasingly articulate and thoughtful spokesmen for LDC statistical offices have difficulty in convincing AID and the rest of the international donor community that all such subject fields or sectors would be better serviced if some funds were set aside for building up the basic infrastructure of the country's statistical system. As one such spokesman has recently stated:

The role of coordinated socioeconomic statistical information systems in promoting developmental efforts in developing countries has not been given the recognition it merits, a weakness which is partly attributable to the fragmented nature of the social sectors. Developmental programs and projects at the sectoral and regional levels are still not based on even moderately accurate statistical information, which could provide an adequate analysis of socioeconomic problems.

Basically, the need for statistical information systems has not been recognized as an objective in its own right and a pertinent input into the developmental process per se. Such assistance as

has been available has been conceived essentially in piecemeal fashion as an ingredient of other development projects addressed to other objectives.*

To put the matter another way, LDC statistical offices can secure funds for a demographic survey, a health survey, or an agricultural cost of production survey. They cannot secure funds for building up basic cartographic capabilities or for establishing a national sample frame or for assistance in computer systems design. If they could secure funds for these latter purposes, their efforts in demography, health, and agriculture would be substantially upgraded. They cannot do so because at present the goals of the donor community do not mesh with the goals of the statistical community in the important respect of improving statistical infrastructures.

Under these circumstances, it is recommended in this Framework that some limited resources should be available to the U.S. Federal Statistical System to provide the training and technical assistance for LDC statistical offices which the foreign assistance program is unwilling or unable to provide. There is need, in short, for some continuing resource base, available to the Federal Statistical System, for filling in the gaps left by the foreign assistance program and aimed at the fundamental purpose of assisting in the development of the statistical systems of LDC's. Provision of such a resource base is a long overdue acknowledgement and acceptance of the lengthy investment in LDC's on the part of the Federal Statistical System and of the fact that over many years, the U. S. statistical agencies have developed an international clientele to whom they have some obligations, independent of the shiftng fortunes of the foreign assistance program.

This proposal may be viewed by some as in direct opposition to current efforts to consolidate all foreign assistance activities into a single independent agency whose administrator reports to the President (the recent Humphrey bill, for example). However, the fundamental character of support of statistical infrastructure means that some level of long-term support is necessary before the short-term, program responsive statistical programs can be useful.

'Parmeet Singh, Director of Statistics, Central Bureau of Statistics of Kenya, "Technical Assistance for the Improvement of Statistics in the Developing Countries: Basic Problems and Issues," a paper prepared for the UN Statistical Commission, Nineteenth Session, New Delhi, November 1976, UN Document No. E/CN.3/472/Add.1.

Implications for Federal Statistical Agencies

Given the periodic fluctuations in the goals and resources of the U.S. foreign assistance program, the strategy for the coming decade with regard to statistical training and technical assistance for LDC's should be concentrated on the acquisition, by the statistical agencies themselves, of limited but continuing resources for these purposes. This strategy entails the building of a case within the various Departments which house the statistical agencies to the effect that budgetary appropriations for international training and technical assistance to LDC's are in the legitimate interests of the agencies and Departments themselves.

More specifically, international statistical training programs for LDC technicians, as currently operated by the Bureau of Labor Statistics, the Bureau of Economic Analysis, the National Center for Health Statistics, the Bureau of the Census and the Economics, Statistics, and Cooperatives Service of USDA,' should seek core training staff support from their own Departments and be placed on solid footing, with respect to both quality and program continuity, rather than the current hand-to-mouth existence which results from dependence on AID. Enrollment should be open to all developing countries with which the U.S. has cordial diplomatic relations, irrespective of whether tuition, travel and per diem costs for participants are borne by AID, a UN agency, or the LDC governments themselves.

A significant deficiency of current U.S. training efforts is the lack of foreign language capability among U.S. statistical personnel, especially French and Spanish. The agencies should recruit training personnel with such capability and work with established regional centers to extend the technical assistance and training which is currently provided.

International statistical research activities in this area which are U.S.-based should receive line item appropriations from the Departments and the essential scope of these activities should be defined and justified by the statistical agencies themselves. The justification lies in following through to its logical conclusions the domestic implications of "global interdependence." That is, there is a

'Training staff in the NCHS Office of International Statistics is for ad hoc, short-term, in-service training and visitor orientation. It is supported by NCHS itself rather than through AID funds. It is safe to generalize that all Federal statistical agencies provide such services to foreign colleagues as a routine professional courtesy. 'SRS training activities of the statistical unit in the Economics, Statistics, and Cooperatives Service are carried out by the Bureau of the Census by means of detail of USDA personnel to the Census Bureau's International Statistical Programs Center.

demonstrable and growing need on the part of U.S. government agencies, business and industry, the academic community, the Congress, and the public for the most up-to-date and accurate statistical information on all foreign countries. At present, the U.S. Federal Statistical System does not have organizational and staff capability of its own for acquiring, evaluating, and disseminating international statistical data on all countries of the world for U.S. domestic purposes. The United States cannot look to other systems such as the United Nations for adequately fulfilling these purposes.

Finally, some limited funds should be made available to the statistical agencies themselves for providing short-term technical assistance to LDC statistical offices where the foreign assistance program cannot do so and the past activities of the statistical agencies have generated legitimate expectations for some continued assistance to the LDC's.

Implications for the Foreign Assistance Program

Whatever the resources Departments can make available to the statistical agencies it will remain true that the foreign assistance area is the province of the Department of State and AID. Overseas activities of the statistical agencies will necessarily continue to be conducted under the administrative authority of the State Department and AID.

Statistical agencies currently working in LDC's under the aegis of AID do so under agreements called Participating Agency Service Agreements (PASA).' These Agreements establish the framework for cooperative relationships between AID and the relevant Department. In the language of the Agreements:

AID recognizes the unique personnel resources, capabilities and experience of the Department to this task; it seeks through this agreement, therefore, to enlist as fully and effectively as possible, on a partnership basis, the pertinent resources of the Department in planning, executing and evaluating those portions of the foreign assistance program in which it has special competence.

In furtherance of broad U.S. objectives, the Department recognizes its responsibility, within its

'In recent years AID has sometimes used as a new term, Resource Services Support Agreements, but the substance of RSSA remains virtually identical to that of PASA.

authority, to contribute toward U.S. foreign policy by participation in foreign assistance programs."

The strategy proposed here simply extends these underlying premises a step further and, recognizing the long tradition of investment of expertise by the statistical agencies in the foreign assistance program solely at AID's expense, proposes that the agencies provide some funds of their own as well as their manpower. As indicated above, the contributions of the statistical agencies would be limited to core staff and basic ongoing activities. It would still be necessary for AID to provide funding for major program undertakings such as the Census Bureau's support to the 1980 World Census of Population and Housing or NCHS' increased activities for improvement of vital registration in LDC's.

The strategy should in no way be construed as undermining the authority of AID or the Department of State, but rather as contributing additional resources in pursuit of mutual objectives. The Department of State and AID would have clearance rights on overseas statistical activities which they do not finance-rights they presently have-and that statistical personnel operating overseas still fall under the jurisdiction of the appropriate embassies and AID missions.

In order to make this strategy most effective the international assistance agencies, especially AID, should be encouraged to staff their subject area divisions with more people with a background in statistics and data analysis. Also, AID should establish a small coordinating office, staffed with people with a very broad background in statistics, including data collection and analysis. This group would coordinate AID's role in statistics and serve as consultant to the statistical agencies by providing advice on the need for infrastructure investment.

Task of Coordination

None of this strategy can be accomplished without an intensive effort at coordination. Formal responsibility for coordination lies with the Office of

'This quotation is from Article II, Underlying Premises, in the standard pro forma or "boilerplate" entitled, "General Agreement Between the Department of X and the Agency for International Development."

Federal Statistical Policy and Standards. It works through direct contact with individual statistical agencies and through the Federal Committee on International Statistics. The actual involvement of both the Office and the Federal Committee has been very limited in recent years, and there is a need to develop more effective coordination.

Vigorous active coordination is needed first among the Federal statistical agencies engaged in training and technical assistance to LDC's. This community of statisticians at present interacts loosely and informally. Effective mechanisms are needed for welding this community into a cohesive force which meets regularly and formally, articulates its own. interests, establishes cooperative ties, and formulates priorities of its own.

Second, coordination is especially needed between this special interest statistical community and the Department of State and AID. At the level of the Department of State, regular machinery should be established for examining the statistical implications of U.S. foreign policy, preferably as that policy is being formulated. This coordination should be viewed as a service the statistical community supplies to the State Department.

Similarly, since the Office of Federal Statistical Policy and Standards formally serves as the focal point for United States participation in statistical activities of international organizations, and because of its role for insuring that statistical standards and definitions are implemented, it seems legitimate also for the Office to inquire into the quality of statistical programs sponsored by donor agencies, including AID. Longstanding bureaucratic machinery exists for OMB overview of domestic statistical activities and programs; virtually none exists in the foreign. assistance area. Orderly procedures should be established to insure that a productive and mutually beneficial dialogue is initiated and continued between the foreign assistance program and the Federal Statistical System.

Thus the major recommendation in the area of coordination is the creation of a working level mechanism for continuing contact with and evaluation of needs for technical assistance, results of technical assistance and establishment of ongoing training activities.