Reducing Maternal and Neonatal Mortality in Indonesia: Saving Lives, Saving the Future (2013)

Appendix

Fertility Decline in Indonesia and Its Relationship to Maternal Mortality

Peter Gardiner PT Insan Hitawasana Sejahtera (IHS)

High fertility, particularly when it involves conception either too early or too late in the fertility cycle or when short birth intervals are involved, is known to pose higher risks for both mothers and infants. High fertility also is generally associated with a larger proportion of births at higher parities that lead to increased risk. In general, teenage pregnancies and those among women over age 40, as well as birth intervals under 24 months, are considered to represent higher risks. Birth intervals of between 36 and 60 months are the most preferred (Rutstein, 2011).

Since the late 1960s, Indonesia has seen fertility, as measured by the total fertility rate (TFR), decline by more than 50 percent, from around 5.6 to 2.3 based on census and intercensal survey sources. This decline has stemmed in part from the launch of a vigorous family planning program in the mid-1970s and from social and economic change, particularly massive improvements through the 1970s and 1980s in basic education for girls. These developments resulted in significant declines in fertility at the youngest and oldest ages in the reproductive span, which are the riskiest. At the same time, there were marked increases in the average intervals between births, and a decline in the percentage of births with intervals under 24 or even under 36 months.

In fact, by the late 1990s Indonesia had reached levels and patterns of fertility and levels of application of family planning that, if not entirely meeting unmet needs, (1) were quite favorable when compared with levels in other countries in the region and (2) were at a stage at which further reductions would likely have only a small independent impact on levels of maternal and neonatal mortality.

A Note on Family Planning

Family planning programs that seek to provide contraceptive services to women (and sometimes to men) to allow couples to better regulate the overall number and timing of pregnancies (and thus also reduce the number of unwanted pregnancies) can be an important adjunct to other, largely service-based efforts to reduce maternal and newborn mortality ratios. Impacts can be even further enhanced where family planning programs (as in Indonesia) have actively sought to address the needs of pregnant women, particularly at the community level.

Although there were a few largely private initiatives in the 1950s and early to mid-1960s during the post-independence, pro-natalist regime of President Sukarno, family planning did not take off until the late 1960s and early 1970s, when it was embraced as a national priority by the New Order government of President Suharto. An independent National Family Planning Coordinating Board (BKKBN) was established under the Second Five-Year Development Plan in 1973. Benefitting from strong leadership and active international support, BKKBN was able to build a national network of user groups extending down

to village level (starting in Java and Bali and gradually moving outward from there). It included extensive logistical and monitoring support structures.

Decentralization, starting in 2001, severely disrupted the largely vertical, centrally driven structures set up by BKKBN by placing responsibility for family planning services solely in the hands of local (district-level) governments and, like other aspects of basic health services delivery, outside of direct central government control. It appears that this approach has had mixed results, with some programs suffering in some areas, while others have tried to maintain the relevant institutions and activities.

Although it is important to recognize the critical role played by family planning and the national program in Indonesia, it is actually the mechanism of reducing the incidence of high-risk pregnancies that ultimately lowers maternal and neonatal mortality. Thus what follows deals not with Indonesia’s family planning program per se (e.g., contraceptive use rates) but rather with overall trends in age-specific fertility, birth spacing, high-parity pregnancies, and unwanted fertility, as well as the implications of further improvements for future reductions in mortality.

Fertility Trends

Time series of age-specific fertility rates (ASFRs)¹ for Indonesia have been calculated based on census and major intercensal survey data since 1971 and on various rounds of the Indonesia Demographic and Health Survey (IDHS) since 1987.² These sources depend on different methods to calculate fertility— the censuses use the own-child methodology and the IDHS uses birth histories. Thus the data are not directly comparable. However, they provide a consistent story about fertility dynamics over the period covered. Table A-1, which is based on census data, shows that declines in fertility since the late 1960s have led to a significant reduction in younger- and older-age fertility. However, the bulk of these declines occurred before the 1990 base year being used for the measurement of trends in maternal and neonatal mortality. The data also show the virtual stagnation in fertility decline by both age group and overall since the late 1990s, a period also synonymous with decentralization and a marked reduction in central control over family planning services and activities.

Results for various years of the Indonesia Demographic and Health Survey are shown in Table A-2. Although they show overall slightly higher levels of fertility,³ the trends are consistent. The trends also confirm the significant declines through the mid-1990s and the ensuing relative stabilization at levels that suggest that, while still significant, pregnancies too early or too late are unlikely to be a major factor in the current unacceptably high levels of maternal mortality in the country.

TABLE A-1 Estimated Age-Specific Fertility Rate (ASFR) and Total Fertility Rate (TFR) per 1,000 Women by Age Group and Source of Data, Indonesia

SOURCES: Badan Pusat Statistik (2007); data for 2010 from Badan Pusat Statistik (2011).

¹Estimated annual births per 1,000 women by five-year age group between the ages of 15 and 49.

²These data include the precursor to the IDHS, the National Contraceptive Prevalence Survey (NCPS), conducted in 1987.

³Hull and Hartanto (2009) argue that this is partly due to the tendency of the IDHS surveys to underestimate the number of never-married women and therefore to overestimate fertility rates, particularly at younger ages.

TABLE A-2 Estimated Age-Specific Fertility Rate (ASFR) and Total Fertility Rate (TFR) per 1,000 Women by Age Group and Source of Data, Indonesia

NOTE: NCPS = National Contraceptive Prevalence Survey; IDHS = Indonesia Demographic and Health Survey. Estimates are based on birth histories and refer to a period of 1-36 months before the survey.

SOURCES: National Contraceptive Prevalence Survey, 1987; Indonesia Demographic and Health Survey, 1991, 1994, 1997, 2002-2003, 2007, 2012.

Birth Intervals

Similar arguments can be made for birth spacing. Although census data are not suitable for this purpose, the birth histories collected in the IDHS series allow such analysis. Table A-3 shows birth intervals from IDHS surveys conducted in Indonesia from 1991 to 2012. The relatively rapid decline in shorter birth intervals, particularly those between 18 and 23 months and even up to 35 months, is clearly evident. Most of these declines occurred up to the late 1990s. More recently, change has been much slower, in line with the stagnation in fertility decline that occurred after decentralization.

A recent paper by Rutstein (2011) compares Indonesia with other South and Southeast Asian countries where demographic and health surveys have been conducted since 2000. Table A-4 reveals that Indonesia has a relatively low percentage of women experiencing either very short birth intervals (< 24 months) or moderately short intervals (< 36 months). What is interesting, however, is the relatively high percentage experiencing long birth intervals (> 60 months), particularly since, similar to short intervals, excessively long intervals also carry an increased risk of birth complications and perinatal and neonatal mortality.

Also based on Rutstein (2011), Table A-5 shows a comparison of preferred birth intervals that again places Indonesia among those countries in which the more dangerously short intervals are clearly not preferred. Taken together, Tables A-4 and A-5 indicate the degree to which Indonesian women are meeting their spacing preferences—not completely, but to a considerable degree.

Higher-Parity Births

Indications of trends in high-parity births can be gained from the IDHS data on children ever born, particularly those born to older women who have completed or nearly completed childbearing. Table A-6 shows trends in numbers of children ever born to all women and currently married women aged 15-49 and to all and currently married women in the oldest recorded age group, 45-49—an age group in which virtually all childbearing has been completed. The mean number of children ever born, a useful summary measure, is also included.

The data clearly show the consistent pace of decline in both average numbers and the percentage of high-parity births (particularly at parity 6.0 and over). Even among women currently at the end of childbearing, the declines have been significant and are likely to decline even further because these women (even in 2012) did much of their childbearing 15 or more years earlier during an overall higher-fertility regime.

TABLE A-3 Percent Distribution of Non-First Births in the Five Years Preceding IDHS by Number of Months since Previous Birth, Indonesia: IDHS, 1991-2012

SOURCE: Indonesia Demographic and Health Survey (IDHS), 1991, 1994, 1997, 2002-2003, 2007, 2012.

TABLE A-4 Percent Distribution of Births by Length of Actual Birth Intervals, Selected Countries in South and Southeast Asia: Most Recent Demographic and Health Survey

NOTE: Table shows by country the percentage of birth intervals less than 24 months, less than 36 months, 36-59 months, and 60 months or more, and the median length of birth intervals, for intervals ending in the five years prior to the survey.

SOURCE: Rutstein (2011).

TABLE A-5 Percent Distribution of Births by Length of Preferred Birth Intervals, Selected Countries in South and Southeast Asia: Most Recent Demographic and Health Survey

NOTE: Table shows by country the percentage of preferred birth intervals less than 24 months, less than 36 months, and 60 months or more, and the median length of birth intervals, for intervals ending in the five years prior to the survey.

SOURCE: Rutstein (2011).

TABLE A-6 Percent Distribution of All Women and Currently Married Women by Number of Children Ever Born and Mean Number of Children Ever Born, Indonesia: IDHS, 1991-2012

SOURCE: Indonesia Demographic and Health Survey (IDHS), 1991, 1994, 1997, 2002-2003, 2007, 2012.

Unmet Need for Family Planning

IDHS surveys also include calculations of the unmet need for family planning in the areas of both spacing and an outright limitation on the number of children. Table A-7 shows that a residual unmet need for family planning of slightly less than 10 percent was reached by the mid-1990s and has remained close to that level ever since. It also shows that satisfaction of the total demand for family planning is high.

TABLE A-7 Percentage of Currently Married Women Having Unmet and Met Needs for Family Planning and Total Demand for Family Planning, Indonesia: IDHS, 1991-2012

NOTE: (1) Unmet need for spacing includes pregnant women whose pregnancy was mistimed and amenhorrheic women who are not using family planning and whose pregnancy was mistimed. It also includes fecund women who are neither pregnant nor amenhorrheic and who are not using any method of family planning and say they are unsure whether they want another child or who want another child but are unsure when they want to give birth, unless they say it would be no problem if within the next few weeks they discovered they were pregnant. The unmet need for limiting refers to pregnant women whose pregnancy was unwanted, amenhorrheic women whose last child was unwanted, and fecund women who are neither pregnant nor amenhorrheic and who are not using any method of family planning and who want no more children. Excluded from the unmet need category are pregnant women and amenhorrheic women who became pregnant while using a method of contraception (these women are in need of a better method of contraception).

   (2) The use of family planning for spacing is defined as an instance in which women who are using some method of family planning and who want to have another child or are undecided about whether to have another. Family planning for limiting is defined as an instance in which women who are using family planning and want no more children. The specific methods used are not taken into account here.

   (3) Nonusers of family planning who are pregnant or amenhorrheic and women whose pregnancy was the result of contraceptive failure are not included in the category of unmet need but are included in the total demand for contraception because they would have been using family planning had their method not failed.

SOURCE: Indonesia Demographic and Health Survey (IDHS), 1991, 1994, 1997, 2002-2003, 2007, 2012.

A new (simpler) method of calculating unmet need was introduced in 2012 by Bradley et al. (2012). The finding produced using this method is shown in Table A-7 for the survey year 2012a. This method resulted in a slightly higher estimate of unmet need and total demand and a lower estimate of satisfied demand.⁴ The relevant point here, however, is that the figures on total demand and the percentage of that demand satisfied provide an indicator of the scope for increased contraceptive use. For the revised 2012 figures, if all demand were satisfied, the contraceptive prevalence among currently married women in Indonesia would be 73 percent as opposed to the actual level of 62 percent recorded in the 2012 IDHS (Statistics Indonesia et al., 2012).

⁴Bradley et al. (2012) also show recalculated estimates of unmet need, total demand, and percentage of demand satisfied for family planning dating back to 1991. These data (shown here) reveal patterns consistent with the 2012 data in Table A-7.

    Complimentary calculations of births by fertility planning status and wanted total fertility rates are shown in Tables A-8 and A-9. These tables also indicate the potential for further reductions in fertility based on the desires and behavior of the women covered in the surveys.

Finally, Table A-10, based on Rutstein (2011), compares the need for family planning related to birth spacing in countries in South and Southeast Asia that conducted a demographic and health survey after 2000. Again, the relatively strong position of Indonesia among the countries covered in meeting the demand for family planning is clearly evident.

TABLE A-8 Percent Distribution of Births in the Five Years Preceding IDHS (Including Current Pregnancies) by Fertility Planning Status, Indonesia: IDHS, 1991-2012

SOURCE: Indonesia Demographic and Health Survey (IDHS), 1991, 1994, 1997, 2002-2003, 2007, 2012.

TABLE A-9 Total Wanted Fertility Rate and Total Fertility Rate in the Three Years Preceding IDHS, Indonesia: IDHS, 1991-2012

NOTE: Rates are calculated on the basis of births to women aged 15-49 in the period 1-36 months before the survey.

SOURCE: Indonesia Demographic and Health Survey (IDHS), 1991, 1994, 1997, 2002-2003, 2007, 2012.

TABLE A-10 Need for Birth Spacing, Selected Countries in South and Southeast Asia: Most Recent Demographic and Health Survey

NOTE: The table shows by country, among currently married women, the percentage with an unmet need for family planning to space births, the percentage using contraception to space births, the total demand for contraception to space births, and the percentage of demand for spacing births that is unsatisfied. Table includes failure to space where available in surveys.

SOURCE: Rutstein (2011).

Contribution of Family Planning Programs to the Millennium Development Goals

The data clearly show the progress of the fertility transition in Indonesia and the movement in levels and patterns of fertility more conducive to safer pregnancies and childbirth. Although it is impossible to quantify precisely, family planning has played a significant role in Indonesia in providing access to methods (particularly modern methods) of controlling fertility across virtually all parts of the country. The current contraceptive prevalence as recorded in the 2012 IDHS is about 62 percent, and this number, as shown earlier, could rise to as much as 73 percent if the total demand for contraception could be met.

As a matter of history, one of the factors in the stagnation that has characterized fertility decline and family planning performance since 2000 has almost certainly been decentralization, which removed the high level of central control under the National Family Planning Coordinating Board and transferred responsibility for family planning to local (district-level) governments. Although some governments may have tried to maintain the extensive and previously vertical networks of user groups extending down to the community level, there has almost certainly been significant deterioration. It is interesting that not only desires but also performances in regulating fertility have remained relatively constant and have not deteriorated as well. Some potentially serious efforts (and the political will) to revitalize the national program are currently under way with the recent appointment of a new head of BBKBN. The intention is to ensure that any influence of high-risk pregnancies on maternal and neonatal mortality is further reduced and not exacerbated.

To summarize, there is the potential for at least a moderate further decline in fertility by meeting the unmet needs for family planning (as is shown in the tables). Furthermore, within the context of the already heavily moderated levels of fertility and higher-risk ages and past achievements in reducing the proportion of dangerously short birth intervals, additional efforts to reduce fertility and meet the underlying demand for family planning would be supportive of a program to reduce maternal and neonatal mortality. However, it seems unlikely that further increases in contraceptive use and the resulting declines in fertility at the levels suggested in the IDHS would alone make a major dent in levels of maternal mortality.

References

Badan Pusat Statistik (Statistics Indonesia). 2007. Proyeksi Penduduk Indonesia per Propinsi Menurut Kelompok Umur dan Jenis Kelamin 2005-2015. Jakarta: Badan Pusat Statistik.

______. 2011, Fertilitas Penduduk Indonesia: Hasil Sensus Penduduk 2010. Jakarta: Badan Pusat Statistik.

Bradley, Sarah E. K., Trevior N. Croft, Joy D. Fisher, and Charles Westoff. 2012. Revising Unmet Need for Family Planning. DHS Analytical Studies, No. 25, Calverton, MD: ICF International.

Hull, Terence, and Wendy Hartanto. 2009. Resolving contradictions in Indonesian fertility. Bulletin of Indonesian Economic Studies 45 (2009):61-71,

Rutstein, Shea O. 2011. Trends in Birth Spacing, DHS Comparative Reports, No. 28, Calverton, MD: ICF Macro.

Statistics Indonesia, National Population and Family Planning Board, Ministry of Health, and Measure DHS: ICF International. 2012. Indonesia Demographic and Health Survey 2012: Preliminary Report. Jakarta.