Abstract
Since its inception in 1984, the Demographic and Health Surveys (DHS) project has become the pre-eminent source of comparative demographic information for the developing world. Under the second five-year phase of the project, DHS-II (1988-1993), which is evaluated in this report, major achievements include the following. (1) DHS-II has become highly efficient in implementing and processing surveys in a wide range of developing country contexts. (2) It has set new standards for turnaround time, with a preliminary report appearing within 3 months of fieldwork and a final report within 12 months, making possible completion of a survey in 2 years from initial contacts. (3) It has been very responsive to data requests from both academic and program users, and has managed to satisfy a range of data-using constituencies in areas of questionnaire content, basic tabulation and reporting plans, and distribution of data files. (4) It has had great success in dealing with implementing agencies in the participating countries, in some cases smoothing long-standing inter-agency quarrels. The program has also been successful in negotiating data release agreements with host countries. DHS data are being used widely. The United Nations and the World Bank population projection activities use DHS data both to provide estimates of fertility in the late 1980's and to provide estimates of child mortality. Publications of comparative world indicators, such as The United Nations Children's Fund's "State of the World's Children," the UN's "Social Indicators," and the World Bank's "World Development Report," include data from DHS in basic tabulations. The data are also being used extensively for more analytical purposes. For instance, the National Academy of Sciences' Panel on Population Dynamics of Sub-Saharan Africa is making extensive use of DHS data for analyses of fertility and child mortality. The DHS World Conference in 8/91 attracted over 200 papers describing analyses using DHS data. The data have also been used for program purposes. For example, the Zimbabwe data, particularly the community module, were used to show that the coverage of the community-based distribution program was far greater than supposed, a finding with far-reaching implications for project design. In Indonesia, data from the 1987 DHS were used as a yardstick to calibrate the existing service statistics, and the data from DHS-II in Indonesia are being used to evaluate regional program performance. The breadth and depth of use of DHS data are impressive and are still increasing. A consequence of the highly efficient and increasingly standardized procedures of DHS has been some lack of flexibility shown in completion of non-standard tasks, such as in-depth surveys or analytical programs. (Author abstract)