Measuring the Thai Family Planning Program's Impact on Fertility Rates: A Comparison of Computer Models

Abstract

Recent interest in the measurement of the impact of family planning programs on birth rates has focused on the use of component projection computer models. This is shown by the revival of the PROJTARG model developed by Bogue, Edmonds, and Bogue (1973) in applications to the situation in Thailand and Colombia (Teachman, Hogan, and Bogue, 1978) and the development of the TABRAP/CONVERSE models described in Nortman and Bongaarts (1975) and Nortman et al. (1978). These models were designed initially for target-setting purposes-to project the level of family planning acceptance necessary for a specified decline in the birth rate-and, as a corollary, to determine demographic changes resulting from a specified level of family planning acceptance or contraceptive protection. Because of the latter feature, they also can be used retrospectively to estimate the effect on fertility rates of a family planning program, based on the actual number of family planning acceptors. This paper presents some estimates of the national family planning program's impact on fertility in Thailand, using the TABRAP/CONVERSE computer models (Nortman et al., 1978). The main objective is to obtain the most accurate estimates possible of the extent to which the fertility rate has been curtailed by births averted through use of program contraceptives. Since the analysis in this case is retrospective and both the number of family planning acceptors and the changes in the fertility rates are already known, the program's impact can be estimated without invoking the entire projection model. It is only necessary here to calculate the births averted by family planning users and then relate them to the observed changes in the fertility rates; thus, the smaller program, TABLE4, is used. Parallel runs are also carried out using the CONVERSE program to see if the results differ substantially when the larger component projection model is used. A secondary objective of this paper is to compare the estimates obtained here with those obtained by Teachman, Hogan, and Bogue (1978), who carried out a similar analysis for Thailand using the PROJTARG model (the analysis is described in more detail in Hogan, 1978). In 1968, a three-year plan was initiated to provide family planning services throughout Thailand as a part of government health clinic activities. According to the plan, services were to be extended to 18 provinces in 1968, to an additional 24 provinces in 1969, and to 25 more provinces in 1970. Prior to 1968, a few pilot programs such as the Potharam Project (from 1964 to 1966) and the International Postpartum Program (beginning in 1966) offered family planning services to a small segment of the Thai population. In 1970, the Thai government officially adopted a policy to reduce the rate of population growth, and family planning services became available nationwide through government clinics and the private sector. Estimates of the extent to which programaverted births contributed to the fertility decline in the 1964-75 period are presented. Four estimates are obtained based on the combination of two sets of pill and IUD continuation rates obtained in surveys in 1971 and 1974-75 and two estimates of acceptors' potential fertility derived from 196465 and 196869 marital fertility rates.