Session IV:
Institutionalizing Data Collection in Host Countries
Workshop participants agreed that more efforts are needed to increase the ability of institutions in developing countries to collect, analyze, and report on health and population data. Support for additional training for participating countries' personnel, including development of curricula dealing with survey research and qualitative methods, would help promote such an effort.
Building Local Capacity
Jeremiah Sullivan discussed the efforts of the DHS program to increase the institutional capability of host countries to conduct demographic and health surveys. The DHS program has encouraged institutionalization through on-the-job training, formal training for data processing systems, the provision of equipment, workshops in data analysis, and sponsoring of fellowships and grants.
Enhanced data processing capabilities are an essential aspect of building capacity in developing countries. Sullivan reported that more than 150 people from developing countries have received training in the use of the DHS microcomputer data processing software. Another 120 are scheduled to receive the training by 1997. Furthermore, the DHS has developed mechanisms to assist with further analysis of data to aid policy development and program evaluation. DHS offers short-term technical assistance and analysis workshops for host country staff, along with fellowships and grants. In conjunction with other donors, DHS has established a small grants program for the analysis of DHS data, and, to date, 52 researchers from developing countries have been involved in this program. Workshop participants suggested that summer training programs with United Nations or USAID funding be established to teach statistics and qualitative study methods.
Morris noted that turnover rates of trained host country survey staff often inhibit progress in capacity building. Robert Bush argued that many donor programs subvert the host country statistical offices and systems, essentially displacing them from data collection activities and the decision-making process. He suggested that current DHS projects do not receive adequate funds to support institutionalization in host countries and that effective capacity building to train and use local statistical offices would double the costs of the data collection efforts. However, greater use of local capacity is possible in certain countries, such as Indonesia or the Philippines, both of which have accumulated a substantial reservoir of trained statistical staff because for the last decade they had been sending 10–20 people per year for practical statistical training and graduate education. However, Hermalin noted that as data collection becomes more complex, it may prove to be much more difficult to transfer the requisite knowledge abroad.
Improving Coordination Among Donors
Participants agreed that there is much need to coordinate the services and activities of donor organizations working in developing countries. Establishing standard tools or data systems among donors would greatly enhance host countries' ability to sustain aid projects.
Michael Vlassoff described the work of the UNFPA in supporting censuses, vital statistics systems, and civil registrations, which enhance the statistical knowledge base of developing countries. UNFPA has set up large training activities at national, regional, and global levels and has provided many fellowship and training programs, such as the African Census Training Program. National statistical offices and population planning units have participated in these efforts. UNFPA and other U.N. offices have sponsored the Household Survey Capability Program, which was a global initiative that supported more than 90 surveys undertaken by the countries themselves.
Following the 1994 International Conference on Population and Development in Cairo, UNFPA assessed its own strategic priorities, much as USAID has done. While the change will be gradual, UNFPA envisages that it will focus less on the support of vital registration systems and population censuses and more on linking existing data systems from different sources (e.g., linking administrative data from social sector ministries with census data) and encouraging appropriate geographic and health information systems for management, monitoring, and evaluation activities. Because population and housing censuses have been used to develop sampling frames for DHS and other surveys, Vlassoff suggested USAID consider assuming partial responsibility for their support.
UNFPA is also undertaking the International Reproductive Health Family Planning Survey Research Program to develop new methods and indicators for population and health programs. It is envisioned that an integral part of the project will be improving and expanding management information systems and other health information systems. Two UNFPA data dissemination services are currently available: POPIN is a U.N.-sponsored internet resource network for population information; POPMAP is a software package designed to make practical geographic information systems available in developing countries.
David Daniels reported that the United Kingdom's Overseas Development Administration (ODA) is trying to examine quantitative and qualitative measurements of program outcomes in an effort to develop benchmarks to allow comparisons among programs. Daniels supported ideas of collaborative efforts, especially in establishing a standard set of indicators; Mark Pearson noted that ODA currently uses indicators developed by the EVALUATION Project. Laura Shrestha mentioned an evaluation of the World Bank's population, health, and nutrition projects that were implemented in the 1980s, which revealed that 30
percent of the projects had no indicators and many of those that had input indicators did not have outcome indicators. Last year the World Bank produced a report identifying 10 key family planning indicators as a guide for internal use. A similar guide for reproductive health projects has been produced.