sectors) rather than on bringing large numbers of 

 developing country nationals to the United States for 

 technical training. U.S. training should be reserved 

 for those involved in highly specialized prograirs that 

 cannot be easily and economically established in 

 developing countries. Cooperative relationships 

 between U.S. universities and other institutions in 

 developing countries to transfer needed planning skills 

 have also been successful in some instances. 



Basic Data 



National efforts, assisted in many instances by the 

 U.S. Bureau of the Census and the United Nations, have 

 considerably improved the collection of cen sus data in 

 many parts of the developing world, although technical 

 assistance and research and development are still 

 needed, especially to find lower cost systems. Viith 

 respect to the registrati o n of births and dea ths, the 

 situation is less favorable. Although vital statistics 

 is an important field, neither the United states nor 

 the United Nations has enough special expertise that is 

 wholly and directly applicable to statistical systems 

 in developing countries. 



Three approaches to the collection of vital 

 statistics might lead to major improvements: (1) 

 increased emphasis on registration of vital events; (2) 

 choice of representative sample areas where special 

 care can be taken to register all vital events as in 

 some of the projects assisted through the International 

 Program of Laboratories for Population Statistics of 

 the University of North Carolina (selection of such 

 areas for statistical purposes should not necessarily 

 mean that they receive additional inputs for health, 

 nutrition, and other developmental programs) ; and (3) 

 development of dual record systems which tend to 

 provide considerably more complete reporting than can 

 either single system alone. In dual record systems, 

 vital events enumerated in the first source but missed 

 in the second and vice versa are identified by matching 

 (record linkage) , and an estimate is made of those 

 missed in both sources assuming independence of 

 sources. 



Household sample surveys are used for a variety of 

 purposes, particularly in the population field (for 

 example, to estimate levels and trends in employment 

 and unemployment, migration flows, fertility patterns, 

 and school enrollment) . Similarly, they are used for 

 establishing health status, anthropometric 

 measurements, and use of services. The United states 

 has found periodic sample surveys to be an important 

 guide to improvement of services, and other nations are 



106 



