The use of contraceptives in Bangladesh has risen steadily over the last two decades. In 1975, the contraceptive prevalence rate (CPR) was 8 per cent and rose to 40 per cent in 1991. The country experienced the steepest increase in CPR between 1975 and 1985: the rates rose from 8 per cent to 25 per cent, a more than three-fold increase in only 10 years.

From a total fertility rate (TFR) of almost 7.0 births per woman in the 1980s, TFR fell to almost 5.0 in the early 1990s, a decline of 28 per cent in a single decade. Demographers have attributed this change in TFR to a decline in marital fertility and the success of the family planning programme rather than rising age at marriage (Huq and Cleland, 1990). The decline has been uniform for all strata of people. Further, the poverty-fertility link is at odds with the evidence (Cleland, Phillips, Amin and Kamal, 1991).

The introduction of grassroots-level female family planning workers, i.e. family welfare assistants (FWAs), in 1978 caused a dramatic shift in the CPR. These FWAs are young, educated female workers recruited by the Government of Bangladesh from local areas; they offer services at the doorstep to women nationwide. Each FWA is assigned to a geographical area comprising approximately 4,000 women (Koenig and others, 1992); however, FWAs are not present in every rural cluster. The role of the FWA is to visit each household in her area that has an eligible woman and encourage that woman to use family planning methods if she does not yet know of the benefits of family planning, or to provide advice and supplies of the method that the woman prefers to use. In very remote areas of Bangladesh, these FWAs are the only contacts with the family planning programme that village women ever have. The FWA is expected to visit every woman in her area at least once every two months; however, the remoteness of some rural areas and other constraints make it impossible to reach every woman in the country. In certain instances, the FWA has to refer her clients to a satellite clinic in her area. Satellite clinics are located in areas where there is a high demand for the insertion of IUDs, or for sterilization. The other alternative is for the FWA to refer the patient to the nearest family welfare centre (FWC). Also constructed by the Government, FWCs are available in only a few clusters in Bangladesh. (In the full paper by the author of this note, the predictive power of the presence of an FWA and FWC in a cluster is assessed as it relates to the contraceptive use of Bangladeshi women, over and above the effects of other socio- economic and demographic background factors.)

This analysis considers 7,764 women resident in rural areas; the data were obtained from the 1989 Bangladesh Fertility Survey (BFS). The analysis is confined to rural women since the community variables, presence of an FWA and FWC in the cluster, were collected for rural clusters only. The data were collected using a two-staged sampling method. (Two models are considered in the full paper: modern reversible methods, and male and female sterilization.) Since sterilization is a permanent method, couples may have a different set of predictor variables, thus necessitating a separate model. The modern reversible methods include the oral pill, condom, IUD and foam. Pregnant women were classified as non- users.

The principal independent variables are presence of an FWA and an FWC in a cluster. These variables are cluster-specific and are therefore separate from other additional variables which are individual ones. These include age, parity, socio-economic score, geographical area of residence and religiosity. To examine the effect of the presence of an FWC and FWA on contraceptive use, a logistic regression model was used. Two different models were constructed: one for modern reversible method users and one for sterilization acceptors.

The analysis showed that women in Chittagong have a 42 per cent lower probability of use and women in Rajshahi have a 65 per cent higher probability of use compared with women in Dhaka; for women in Khulna, there is no difference in probability of use. Whereas religiosity has no significant effect in determining the use of modern reversible methods, the presence of an FWA increases by 54 per cent the probability of a rural woman being a user. It also accounts for more than one-third of the inter-cluster variation. The presence of an FWC in the cluster does not have any significant effect on modern method use, and it does not explain any inter-cluster variation. The results match the findings of previous research. Koblinsky (1989) and Phillips (1985) found that visits by the FWA lead to increased use. Koenig and others (1992) remarked that static FWCs did not contribute to increased use in the Matlab comparison area. This analysis also concludes that static FWCs in Bangladesh are not significant predictors of contraceptive use. Cleland and others (1991) remarked: "There is unassailable evidence that routine household visits by family planning workers can have a lasting impact on reproductive behavior", and "A visit to a health center for treatment of a sick child or to obtain family planning supplies is a major feat of logistics and persuasion" for a Bangladeshi woman.

It is, therefore, not surprising that the FWA is a significant predictor of contraceptive use whereas the FWC is not. This leads to the conclusion that "doorstep" services are very important in Bangladesh (Cleland and others, 1991). It also implies that the introduction of the FWAs has been a successful addition to the family planning programme. Since Bangladesh is a male-dominated, predominantly Muslim country, it is highly unlikely that women will voluntarily seek family planning methods, even if they are made available by the existence of FWCs close to where women live.

Firstly, the issue of contraception is considered to be extremely private and confidential, and women in rural Bangladesh would feel shy to express a desire to limit family size. This is more so because a women cannot visit the health centre on her own. Conservative social values do not allow a woman to visit any place on her own without a male escort. Older women are more likely to have greater freedom and may be seen in public accompanied by their children, but young women (below 30 years of age) are expected to be out of the public eye or even practice purdah (covering with a veil) when in public. Prior to the introduction of FWAs, the Government employed male family planning workers and installed FWCs in various villages from 1969 to 1978. The existence of family planning facilities or the knowledge of family planning were not enough to motivate rural women to use contraception. It was only after the introduction of the FWAs in 1978 that the family planning programme gained momentum and the CPR began to rise and TFR decline.

It is not the location of the FWCs or the distance that women must travel to reach one of them that impedes the women from using the services offered by the FWCs. An understanding of women's role in a conservative Muslim society can explain the reason why a woman will not seek assistance, even if she feels motivated to use contraceptives and knows that an FWC is situated nearby. One study by Kamal and Sloggett (1993) shows that use of modern methods increases gradually as women's autonomy in the family and her mobility improves.

In the initial days of their recruitment, young educated FWAs faced much resistance from society. They were branded as "characterless", evil women who worked with men, did not observe purdah, and advocated something that was totally against God's will. Over the years, the gradual diffusion of ideas has brought about a change in attitude and today the FWAs occupy a special respectable position in village society (Simons and others, 1992). Caldwell and others (1992) remarked about the success of family planning in Bangladesh by saying: "The only adequate explanation is the legitimisation of the concept of the small family and of contraceptive use, and the greater diffusion of the idea that controlling family size may prove to be economically advantageous".

As for sterilization, the multilevel logistic regression model showed that, as expected, older women are more likely to be acceptors of sterilization compared with younger women (below 30 years of age). It also showed that women in higher parity groups are more likely to accept sterilization. However, there is a drop in probability from women with parity 3 or 4 to women with a parity of 5 and above, which could be because women in that age group (parity 5 and above) have already completed their childbearing, perhaps more in the social than biological sense.

Rahman (1984) remarked that in rural Bangladesh, women who are already grandmothers would consider it shameful to bear children. Being a grandmother in Bangladesh does not imply being in an old- age group. The age at first marriage in Bangladesh can be as low as 12 or 13 years, so many women in parity 5 who have married daughters and grandchildren may be relatively young. After age 30 and parity 5, the majority of women cease to have active conjugal relationships. As a result, they do not feel the need to use contraceptives to limit their fertility, although they are still married. Therefore, women who are currently in parity 3 or 4 are more likely to limit fertility than women in parity 5 or above, because they are still likely to be in an active sexual relationship and, being in a younger age group, are more likely to accept fertility-limiting measures. This is because the motivation to adopt family planning and the technologies for this purpose have been quite recent in Bangladesh, and more women in the younger age groups have responded positively to family planning. In any population, the diffusion of new ideas and technologies usually begins with the younger generation. That indeed has been the case in Bangladesh, as has been remarked by Huq and others (1990) in their initial analysis of 1989 BFS data.

Women in lower socio-economic groups are more probable acceptors of sterilization, whereas women in higher socio-economic groups opt for modern reversible methods. In Bangladesh, couples receive a compensation payment of approximately US$4 for being sterilized. The attractiveness of this cash payment has been especially attractive to couples living in poor economic conditions (Cleland and Mauldin, 1991).

Religiosity plays a strong role in the acceptance of sterilization, as opposed to the use of modern reversible methods. Women who are less religious have a higher (43 per cent) probability of being an acceptor compared with those who are strictly religious. Women who consider themselves average in terms of religiosity do not differ from their more religious counterparts. In a predominantly Muslim society like Bangladesh, this differential is not at all surprising.

Strict religiosity has often been identified as an impediment towards the adoption of contraception. Bernhaart and others (1990) remarked that religious beliefs are widely perceived and practised in Bangladesh, but there was no association between contraceptive use and devoutness. However, that study was limited because of its small sample size. (The data used by the author of this note, which are nationally representative, show that, after controlling various demographic and socio-economic attributes, decreased devoutness of a woman leads to increased acceptance of sterilization.) Compared with Dhaka Division, Chittagong Division had a 55 per cent lower probability of adoption of sterilization, and Rajshahi Division had a 36 per cent higher probability; for women in Khulna, adoption of sterilization did not differ significantly.

The presence of an FWA in a rural cluster increased by 66 per cent the probability of a couple being an acceptor of sterilization. As with modern methods, the presence of an FWA explained more than one-third of the inter-cluster variation in the model. But the presence of an upazila-level family welfare centre (FWC) had no significant effect on the acceptance of sterilization, nor did it explain any fraction of the inter-cluster variation.

Increasing the number of FWCs in the clusters might not be fruitful in attempting to improve CPRs in Bangladesh. A more positive input might be to increase the number of FWAs in rural areas by recruiting FWAs for every cluster. Enabling FWAs to visit each eligible woman once a month instead of once in two months, as well as increasing the effective time spent with each client would be beneficial. Although the FWAs have been successful in their work, there are various areas of Bangladesh where women have had very few visits or none at all from FWAs in the previous six months. In particular, Chittagong Division deserves mention because of its lower than average CPR. The 1991 Contraceptive Prevalence Survey (CPS) revealed that only one-fourth of the women in that District were visited by FWAs compared with two-fifths of the women in the three other divisions (Mitra and others, 1992). The Government could attempt to reach higher targets for field-worker visitation during the next five years. Increasing the frequency of visits and the number of FWAs, and providing better incentives and facilities for FWAs in the future would likely raise the CPR and make a major contribution towards achieving replacement-level fertility by the year 2000.

Bernhart, M.H. and M. Mosleh Uddin (1990). "Islam and family planning acceptance in Bangladesh", Studies in Family Planning, 21( ? ):287-292.

Caldwell, J.C. and P. Caldwell (1992). "What does the Matlab fertility experience really show?", Studies in Family Planning, 23(5):292-310.

Cleland, J. and W.P. Mauldin (1991). "The promotion of family planning by financial payments: the case of Bangladesh", Studies in Family Planning, 22( ? ):1-18.

Cleland, J., J.F. Phillips, S. Amin and G.M. Kamal (1991). The determinants of reproductive change in Bangladesh. Paper prepared for the World Bank.

Huq and Cleland (1990). Bangladesh Fertility Survey 1989 (Main Report) (Dhaka: National Institute of Population Research and Training).

Kamal, N. and A. Sloggett (1993). The influence of religiosity, mobility and decision making on contraceptive use: Secondary Analysis of BFS 1989 data. National Institute of Population Research and Training (NIPORT), Dhaka, Bangladesh.

Koblinsky, M.A., S.J.G. Brechin, S.D. Clark and M.Y. Hasan (1989. "Helping managers to manage: work schedules of field workers in rural Bangladesh", Studies in Family Planning, 20(4):225-234.

Koenig, M.A., U. Rob, M.A. Khan, J. Chakroborty and V. Fauveau (1992). "Contraceptive use in Matlab, Bangladesh in 1990: levels, trends and explanations", Studies in Family Planning, 23(6):352-364.

Mitra, S.N., C. Lerman and S. Islam (1992). Bangladesh Contraceptive Prevalence Survey 1991 (Key Findings)(Dhaka: Mitra and Associates).

Phillips, J.F., M.B. Hossain and M.A. Koblinsky (1985). The determinants of contraceptive choice in two rural subdistricts of Bangladesh: a logistic regression analysis of the effect of an organization development strategy on contraceptive use dynamics. Paper presented at the annual meeting of the East-West Population Institute, Honolulu, Hawaii, August.

Rafiquzzaman, A.K.M., A. Ahmed, A. Larson, F. Mabud and A. Rashid (1991). Contraceptive Prevalence Survey-1990 (Dhaka: National Institute of Population Research and Training).

Rahman, Mukhlisur (1984). "Tradition, development and the individual", In: Asian Population Change Series No. 1, P. Kane and L. Ruzicka (eds.).

Simons, R., R. Mita, M.A. Koenig (1992). "Employment in family planning and women's status in Bangladesh", Studies in Family Planning 23(2):97-109.

Tabachnick, B.G. and L.S. Fidell (1989). Using Multivariate Statistics (New York: Harper Collins Publishers).

Woodhouse, G., J. Rasbash, H. Goldstein and M. Yang (1992). A Guide to ML3 for New Users (London: Institute of Education, University of London).

For further information on this material please contact: loftus.unescap@un.org