Understanding how family planning programs work : findings from five years of evaluation research

This report synthesizes findings of research conducted under the project from early 1992 to mid-1996. Project studies, though observational rather than experimental in nature, have been technically strong, often exemplary, in terms of conceptualization of program structure and processes, adequacy of data, measurement, methodological rigor, and balanced discussion of results. And while the diversity of FP interventions makes it difficult to render a single, overarching finding regarding their effect on FP behavior, statistically significant findings have emerged regarding their influence under a given set of conditions. (1) FP service provision. FP programs have broadened the composition of service providers. Of the programs studied here, community-based, interpersonal outreach in Bangladesh explained a rise in contraceptive use from 13% to 40% between 1984 and 1993 and prevented discontinuation by 43% in the mid-1980s and by 65% in the late 1980s. Conversely, Indonesian clients supplied with modern contraceptives by public clinics were more likely to avoid pregnancy than those supplied by private providers. Pregnant women"s exposure to prenatal care increased the likelihood of postpartum adoption of modern contraception in Tunisia and Morocco, suggesting the benefits of integrating maternal/child health and FP care. (2) Quality of FP care and contraceptive prevalence. A study in Peru found that prevalence can rise by as much as 16% if all women have access to high quality care, while a study in the Philippines suggests that making one more FP method available can raise the likelihood of using public clinics by almost 6% and of using private clinics and hospitals by 10%. A Philippines-based study of unmet need for FP finds the need rooted in traditional concerns about husbands" pronatalist attitudes, contraceptive side effects, perceived risk of pregnancy, and strength of fertility preferences, but not in problems with access to FP services. (3) FP promotion. Some of the strongest effects of FP programs on contraceptive behavior may be derived from mass media and other promotional activities. A 3-year follow-up of Nigerian women found that 34% of those who saw a FP media message in 1990 were using modern contraception in 1993 compared with 18% who were not exposed. If the women discussed FP with others, 44% of those not contracepting expressed an intention to adopt modern contraception in the future, compared with 26% of those who did not discuss FP. In studying how women informally discuss FP and diffuse FP information, one pair of investigators found that current users in Kenya had an average of 4.1 conversations with friends and relatives about FP compared with 2.8 for non-users. (4) Gains in program and service capacity. Two key studies focused on measuring FP program capacity and effort over time at the cross-national and national levels. A new round of FP effort scores shows large increases in program strength since 1972 but overall modest rises since 1989. Of 77 countries with four rounds of data, 40 showed effort beyond 50% of the maximum score in 1994, compared with 20 in 1972. A follow-up of Tanzanian health and FP facilities between 1991 and 1994 demonstrated anticipated gains in contraceptive logistics support and higher volumes of new and returning clients, but no increase in trained staff. A study of six countries found the magnitude of service availability effects to be highest when contraceptive use was initially low and fertility high. Finally, trends in use of contraception and abortion showed the former out- pacing the later in lowering fertility between the mid-1970s and mid-1980s in Colombia and Mexico. Details on the results of many other EVALUATION studies, particularly those addressing im- provements in measurement not highlighted here, are contained in this report. The appendices present a summary of study findings in tabular form. A complete bibliography of study reports is also provided. (Author abstract, modified)