Part 2 UNMET CONTRACEPTIVE NEED AND FAMILY PLANNING PROGRAMMATIC CHALLENGES IN SOUTH ASIA. Barkat-e-Khuda, PhD* Samiha Barkat, MPH.

Transcription

1 Part 2 UNMET CONTRACEPTIVE NEED AND FAMILY PLANNING PROGRAMMATIC CHALLENGES IN SOUTH ASIA Prepared by: Barkat-e-Khuda, PhD* Samiha Barkat, MPH * Professor, Department of Economics, University of Dhaka, Bangladesh. barkatek@yahoo.com. The earlier version of the paper was presented at the UNFPA-ICOMP Workshop on Operationalizing the Call for Elimination of Unmet Need for Family Planning in Asia and the Pacific Region, Bangkok, September 18-19,

2 Acknowledgements The authors are grateful to several individuals for making various documents available. They include Mr.Abdul Malik Faize (Afghanistan), Mr. Anand Tamang (Nepal), Dr. Talat Jabeen (Pakistan), and Mr. Hairudin Masnin and Ms. Neera Shrestha (ICOMP). The authors also benefitted from additional valuable information received at the UNFPA-ICOMP Workshop on Operationalizing the Call for Elimination of Unmet Need for Family Planning in Asia and the Pacific Region, Bangkok, September 18-19,

3 Background The population of Asia, excluding China, is approximately 2.8 billion. The South Asian countries together have a population of around 1.6 billion, with India alone contributing around 1.2 billion (Population Reference Bureau 2010).The population age structure is disproportionately youthful. Those aged years account for around one-third of the total population. The female age at marriage is relatively low, and hence, the increasing need for improving access to quality family planning (FP) services. Also, there is a need to provide appropriate FP methods to those who have completed their family size. Although socio-economic progress has been achieved over time, the progress has been slow. While overall literacy rates have risen in all South Asian countries, they are still low. Except Sri Lanka and Maldives with over 90 percent literacy, it ranges from around 40 percent in Afghanistan to 68 percent in Nepal. The per capita income has increased, though still low. Only Sri Lanka and Maldives are middle-income countries. The per capita GDP ranges from around US$ 400 in Afghanistan to US$ 800 in Bangladesh, India and Pakistan. However, there is considerable inequity in income distribution. Also, a significant percentage of the population lives below the poverty line. There is increasing urbanization and growth of slums, further adding to the problems of organised service delivery in health and family planning. Over the past three decades, well-planned FP policies and programmes in the South Asian region contributed to considerable increase in the contraceptive prevalence rate (CPR) and decrease in unmet contraceptive need (UCN), resulting in significant reduction in the total fertility rate (TFR). However, over the past decade or so, the rate of increase in CPR has slowed down and UCN has not declined appreciably in the region due to limited access to quality FP services, organizational problems, and declines in priority and investment in FP by national governments and donor agencies. The high levels of pregnancy termination in the region indicate considerable UCN. Also, there are disparities with regard to TFR, CPR and UCN in most South Asian countries by education and wealth status of the population. The Call for Elimination of Unmet Need for Family Planning which emanated from the UNFPA Asia Pacific Regional Office (APRO) and the International Council on Management of Population Programmes (ICOMP) Regional Consultation on Family Planning in Asia and the Pacific- Addressing the Challenges held at Bangkok in December 2010 identified 10 strategies to reduce UCN within the context of sexual and reproductive health. There is now a renewed recognition among governments and donor agencies of the need for enhanced efforts to reposition FP as part of the development agenda. Political commitment from the key policy makers, parliamentarians and civil society leaders is essential for effective implementation of the strategies and action plans agreed upon at the 2010 UNFPA APRO-ICOMP Meeting for repositioning FP programme in the South Asia region. 1 Objectives and Design The purpose of this study is to develop a feedback report on assessing the implementation of the Call for the Elimination of Unmet Need for Family Planning agreed at the December 2010 High Level Regional Consultation held in Bangkok. The focus of this study is on five South Asian countries: Bangladesh, India, Maldives, Nepal and Pakistan. Also, relevant information is provided for the other three South Asian countries-- Afghanistan, Bhutan and Sri Lanka--to make the study more comprehensive. The terms of reference for this study require the writing of a comprehensive report on the status of UCN in the five South Asian countries under the focus of this study. Three study questions are listed: i. Overview of the Status of Unmet Need for Family Planning How do the levels of unmet need differ among the sub-population groups (i.e. woman, young people, marginalized groups, migrants, people living with HIV, women in the lowest quintiles, women with low education levels, etc) within the country and across the countries in the sub-region? How has unmet need changed over time? Has there been any significant change/progress in reducing unmet need following the December 2010 High Level Regional Consultation? ii. Policies and Programmes How have governments, donors and civil society responded through policies and programmes to reduce the unmet need? Analyze and synthesize key interventions and actions implemented at country level, at policy and programme level, in terms of their effectiveness, efficiency, accessibility, affordability and sustainability. Has the goal of family planning policies and programmes shifted from an emphasis on increasing contraceptive prevalence (and reducing fertility) to satisfying unmet need (and reducing unintended fertility)? What are the strategies used to improve method mix and quality of care? Are there any examples of good practices from the focus country/countries? iii. Recommendations Evaluate available programme options and recommend strategic actions that will have 1 See UNFPA-ICOMP Call for Action, Dec

4 implications for policies and programmes to address unmet need in the countries under review. There are two aspects of this study which need to be noted. First, the question whether significant change/progress has been made in reducing UCN since the 2010 December High Level Consultation (i.e., in just a year-and-a-half ), is an impossible question to answer not only because the time period is too short but also because no survey has been undertaken in the countries under study to measure any such change 2. The study, therefore, examines data from the DHSs available for these countries to get a longer-term perspective on the changing situation of UCN and related issues rather than just preparing a feedback report on the status of UCN since the 2010 December High Level Consultation. Second, other than Bangladesh where the author resides, no visits were made by the author to any of the countries to obtain information on policies and programmes adopted in those countries since the 2010 High Level Consultation, and their effectiveness, efficiency, accessibility, affordability and sustainability, and therefore, the author had to rely on various relevant documents, strategies and assessment reports of those countries (all such documents have been referenced in the paper) received from colleagues from some of those countries and from ICOMP and through internet search. In addition, some relevant information was obtained through personal communication with concerned individuals from the selected countries. The paper is organized into eight sections. Section 3 examines the data on family planning and fertility in South Asia. Section 4 reviews the concept of unmet contraceptive need, and examines the extent and differentials in unmet need in South Asia. Section 5 examines barriers to contraceptive use in the region. Section 6 presents a critical review of the key family planning programmatic challenges in the region. Section 7 looks at policies and strategies adopted by the South Asian countries to address such programmatic challenges. Section 8 puts forward some concluding remarks and makes some recommendations. Family Planning and Fertility in South Asia Family Planning Family Planning (FP) programmes started first in India in the early 1950s, followed by Bangladesh and Pakistan in 1953, Nepal in 1959, and Maldives in 1984; and the programmes in the different countries went through various phases 3. The Bangladesh FP programme began with voluntary efforts in 1953, and after the independence in 1971 became an integral part of the national programme. India was the first country in the world to initiate an organized FP programme in 1952, which went through various phases since its reorganization in 1966 with the creation of the Department of Family Planning within the Ministry of Health and Family Planning. In Maldives, with the official launching of the child-spacing project in 1984, FP services were formally introduced. Pakistan started implementing a national population programme in 1965, although it is no longer a national priority. Further, it has been decentralized without adequate support from the federal government. In Bhutan, the Village Health Workers (VHW) Programme was initiated in The VHWs are the important link between the community and the government in improving FP service delivery and other health services. The RH programme prioritized FP as one of its key elements. In Afghanistan, the CPR was 10 percent in 2006, ranging from 2-12% in different provinces (Table 1). Use of oral contraceptive and injectable were 10% and 2% respectively (Govt. of Islamic Republic of Afghanistan 2006). In 2010, the CPR was 22 percent (20% modern methods). The most commonly used methods are injectables (7%), pill (5%) and LAM (4%) (Govt. of Islamic Republic of Afghanistan 2011). From about 4 percent of Bangladeshi married couples using FP in the 1960s, the CPR slowly increased during the 1970s to around 8 percent (see, e.g., Khuda 1981 and 1984). The CPR increased about six-fold to around 45 percent in compared to Thus, it is clear that the FP programme in Bangladesh achieved commendable success until the mid-1990s. Thereafter, the rate of increase slowed down. The CPR increased to 61 percent in 2011 from 56 percent in 2007 (Table 1). The slowing down in the rate of increase in CPR since the mid-1990s is due to erosion in political will and commitment to the programme and various organizational problems. In Bhutan, the CPR in 2010 was 65.6 percent almost entirely of modern methods (Table 1). Injectables were most commonly used method (28.9%), followed by male sterilization (12.6%), IUDs, oral pills, female sterilization and condoms (Royal Govt. of Bhutan, UNICEF and UNFPA 2011). 2 The preliminary findings from the Bangladesh and Nepal 2011Demographic and Health Surveys (DHSs) have been made available which show changes in unmet need between 2006 and 2011; however, such findings are not available for the other South Asian countries. 3 See, Khuda, 1981 and 1984; Khuda et al, 1992, 1993 and 1994; Khuda and Anwar 2004; Khuda and Barkat 2010; Cleland et al 1994: Jain and Jain 2010; Nirmula 2010; Sathar and Zaidi 2010; Tamang et al

5 The CPR in India was only about 13 percent in Between 1992/93 and 2005/06, it increased from 51 percent to 64 percent (Table 1). However, the increase has not been uniform. Northern India is lagging behind. The odds of contraceptive use increases with women s age, parity and number of living sons. Contraceptive use is positively associated with education, higher among employed women who received cash, and increases with household wealth quintiles ( Jain and Jain 2010). Maldives had a CPR of only about 10 per cent in By 1999, it had increased to 42 per cent; however, it declined to 39 percent in 2004 and further declined to 35 percent in 2009 (Table 1). The weakness in the FP programme is largely due to erosion of the primary healthcare system and more emphasis on medicalization (Nirmula 2010). Nepal experienced sharp increase in CPR between 1996 and 2011, with its CPR increasing from 29 per cent to about 50 per cent, and the use of modern methods increasing from 26 per cent to 43 per cent (Table 1; and Ministry of Health and Population, Nepal, New ERA and ICF Macro 2011). The CPR is positively associated with wealth status. A large segment of the Nepalese Table 1 Trend in Contraceptive Prevalence Rate in South Asian Countries Afghanistan 2006 (Govt. of Afghanistan, 2006) 2010 (Govt. of Afghanistan) Bangladesh 1975 BFS 1983 CPS 1985 CPS 1989 BFS 1991 CPS BDHS BDHS BDHS 2004 BDHS 2007 BDHS 2011 BDHS Country Any Method Method Any Modern Method Bhutan 2010 (Govt. of Bhutan, 2011) India NFHS-3 ( ) NFHS-2 ( ) NFHS-1 ( ) Maldives RHS 1999 RHS 2004 MDHS 2009 Nepal 1996 NFHS 2001 NDHS 2006 NDHS 2011 NDHS Pakistan PCPS PDHS PCPS PFFPS PRHFPS 2003 SWRHFPS PDHS Sri Lanka 2006 (Govt. of Sri Lanka 2009) NA

6 society, especially marginalized groups, however, remains underserved. Also, it is argued that other RH programmes had been prioritized at the cost of the FP programme. As a result, the pace of increase in the CPR has slowed down, and the use of modern methods has remained unchanged at around 44 percent between 2006 and The CPR in Pakistan remained below 10 per cent throughout most of the 1970s and 1980s, and was only 12 per cent in The CPR increased to 24 per cent by 1996 and further to 32 per cent in 2003, but then, declined to 30 per cent in 2006/07 (Table 1). The stagnation in CPR is more apparent in urban areas. During the 1990s, the CPR in urban areas increased more rapidly than in rural areas; however, this trend subsequently reversed and since 2001 the rate of increase in CPR is greater in rural than urban areas. Despite the consistent increase in rural CPR, absolute levels of contraceptive use in rural areas continue to be extremely low (NIPS and Macro International Inc. 2008). In Sri Lanka, the CPR was 68 percent in percent modern methods and 15 percent traditional methods (Table 1). Female sterilization (17%) and injectables (15%) were the most widely used methods (Govt. of Sri Lanka 2009). Fertility During the past five decades, fertility has declined worldwide. Between 1950 and 2000, the TFR in Asia declined by over 50 percent, to reach an average of 2.8 births per woman. Among the South Asian countries, Afghanistan has the highest fertility (TFR of 5.1), followed by Pakistan (TFR of over 4), while the TFR ranges between 2.3 and 2.7 among the other countries (Table 2). A point of concern relates to the slowing down in the rate of fertility decline in South Asia. Afghanistan recorded some decline in fertility during the past seven years, especially after According to the 2010 Afghanistan Mortality Survey, the TFR in 2010 was 5.1( Govt. of Islamic Republic of Afghanistan 2011). However, higher levels of fertility are reported by PRB (2011) TFR of 5.7 in 2010; and by the UN Population Division (2011) TFR of 6.6 for the period The TFR was higher in rural areas (5.2) than in urban areas (4.7); higher in South zones (5.3) than North and Central zones (5.0); higher among those with no education (5.3) than those with higher education (2.8); and higher among those in the lowest wealth quintile (5.3) than those in the highest wealth quintile (4.8) (Govt. of Islamic Republic of Afghanistan 2011). Bangladesh had quite high fertility during the 1960s, with the TFR of over 7; and it remained above 6 until 1981 (Cleland et al. 1994). Between 1989 and , the TFR declined sharply from 5.1 to 3.4. However, the decline was much slower thereafter, and stalled at around 3.3 during the period. Then, it declined slowly to 3 in , 2.7 in 2007 (Khuda and Barkat 2010), and 2.3 in 2011 (NIPORT, Mitra and Associates, ICF International 2012). There are sharp fertility differentials. Fertility is lower in urban than rural areas, although the difference has narrowed over the decade from 1.1 births in to 0.5 births in Four of the seven administrative divisions (Khulna, Rajshahi, Dhaka and Rangpur) have reached replacement level fertility (TFR of 2.2) or below, while it is 3.1 in Sylhet division and 2.8 in Chittagong division (NIPORT, Mitra and Associates, ICF International 2012). Fertility is lower among educated women than women with no education, and lower among women in the highest wealth quintile than those in the lowest wealth quintile (NIPORT, Mitra and Associates and Macro International 2009). Also, fertility is lower among women with access to mass media than those without, working women, women belonging to landed households, and those belonging to households with electricity connection-- an indicator of not only economic condition of the household but also its level of monetization (Khuda 2004; Neaz et al 2004). India had high fertility during , with a TFR of 5.7. Since then, it declined to 2.7 births: by about 2 births between 1971 and 1991 and by another 0.7 births between 1991 and The timing and pace of fertility decline, however, varied across states. Seven states are at or below the replacement level and four are close to replacement level; however, the North is lagging behind. There has been a slowing down in fertility decline after the elimination of the FP target system in 2004 ( Jain and Jain 2010). India as a whole has substantial fertility differentials. The TFR is higher for women with no education than for those with education, higher among Muslims than Hindus, and higher among women living in the lowest wealth quintile households than those living in the highest wealth quintile households (IIPS and Macro International 2007). Maldives started experiencing fertility decline in the early 1990s, a trend which became more rapid in recent years. Indeed, Maldives experienced one of the most rapid fertility transitions in the region, with the TFR declining from 6.4 in 1990 to 2.1 in 2006 (Niraula 2010). There are fertility differentials. Women in Malé have the lowest fertility, while women in the South Central region have the highest fertility. Fertility varies little by the woman s education; however, it is inversely associated with wealth quintile (Ministry of Health, Maldives and ICF Macro 2010). Nepal halved its fertility from 6.3 children per woman in 1976 to 3.1 in According to the preliminary results of the 2011 NDHS, it further declined to 2.6 (Ministry 6

7 of Health and Population, Nepal and ICF Macro 2011). Fertility is higher in rural than urban areas (Ministry of Health and Population and ICF Macro 2011; Ministry of Health and Population 2005; Pradhan and Pant 2007). Fertility is inversely associated with household wealth status (Tamang, Subedi and Packer 2010). Pakistan is the last among its neighbours to experience fertility decline, and continues to have the highest fertility in the region, except Afghanistan. The TFR in Pakistan is more than one birth compared to that of India and Bangladesh (Sathar and Zaidi 2010). Fertility is lower in rural than urban areas, lower in Punjab than in the other provinces, higher among women with no education than those with education, and is inversely associated with wealth (National Institute of Population Studies and Macro International Inc. 2008). Sri Lanka s TFR was 2.4 in It was slightly higher in rural areas (2.6) than urban areas (2.2) (Govt. of Sri Lanka 2009). Adolescent Fertility There is a large adolescent population in South Asia 4. In 2008, adolescent women in the developing countries had an estimated 14.3 million births, 91 percent of which occurred in the developing countries --six million in South Central and Southeast Asia. Each year, adolescent women account for 12 percent of those in South Central and Southeast Asia. Although adolescent fertility has declined somewhat over the past three decades, they still vary widely by country and by region within country. South Central Asia recorded the largest decline in adolescent fertility, where births dropped from an estimated 90 per Table 2 Fertility Trend in South Asian Countries Country Year TFR Afghanistan Bangladesh 2006 (Govt. of Afghanistan, 2006) 2011 Govt. of Afghanistan,2011) Bhutan 2010 (Govt. of Bhutan, 2011) 2.6 India Nepal Pakistan NFHS-1( ) NFHS-2 ( ) NFHS-3 ( ) (NFFS 1986) (NFHS 1991) (NFHS 1996) (NDHS 2001) (NDHS 2006) (NDHS 2011) 1984 (PCPS) (PDHS) 1994 (PCPS) (PFFPS) (PRHFPS) (SWRHFP) (PDHS) Sri Lanka 2006 (Govt. of Sri Lanka, 2009) There are currently over one billion people aged years, by far the largest childbearing cohort in history. Sexual activity among youth places them at risk of unintended pregnancy and STIs, including HIV/AIDS. Meeting the reproductive health needs of this underserved population is critical, because their contraceptive choices will have a major impact on the size of the future world population. 7

8 1,000 women aged years in 1975 to 73 per 1,000 in There is extremely high level of adolescent fertility in Bangladesh, being one of the highest in the world. One-quarter of adolescent married women have begun childbearing in 2011, down from 33 percent in 2007 (NIPORT, Mitra and Associates and ICF International 2012). Adolescent fertility is higher in rural than urban areas, higher among those with no education than those with education, and higher among those in the lowest wealth quintile than those in the higher wealth quintiles (NIPORT, Mitra and Associates and Macro International, 1994, 2001, 2005 and 2009; Khuda and Barkat 2010). In Afghanistan, adolescent fertility is 16 percent. By age 19 years, one-third of the women have started childbearing. It is higher in rural areas, in Central and Western zones, among the less educated, and the poor (Govt. of Afghanistan 2011). In Bhutan, adolescent fertility is 5.9 percent (Royal Govt. of Bhutan, UNICEF and UNFPA 2011). Overall, 12 percent of adolescent women in India have become mothers and 4 percent are currently pregnant with their first child. That is, one in six adolescent women has begun childbearing (IIPS and Macro International Inc., 2007). Adolescent fertility is quite low in Maldives: only 2 percent of adolescents have started childbearing (1% are mothers, and less than 1% are pregnant with their first child (Ministry of Health, Maldives and ICF Macro 2010). In 2011, one in six adolescent women became mothers in Nepal, down from 20 percent in 2006 (Ministry of Health and Population, Nepal, New ERA and ICF Macro 2011). Adolescent fertility has declined in Pakistan from about 16 percent at the time of the PDHS to 9 percent at the time of the 2006 DHS. The percentage of adolescent women who have begun childbearing is higher among those without education than those with education, and higher among those from poorer households than those from wealthier households (NIPS and Macro International Inc. 2008). In Sri Lanka, adolescent fertility is 6 percent (Govt. of Sri Lanka 2009). Fertility Preference In Afghanistan, the desired number of children is 4 per family; however, no information is available regarding the percentage of women who wish to space and limit childbearing (Govt. of Islamic Republic of Afghanistan 2006). About two-thirds (65%) of currently married women in Bangladesh want to limit childbearing in 2011 (up from 60% in 2006): 59 percent want no more children, and an additional 6 percent have been sterilized. The proportion of currently married women who either want no more children or have been sterilized increases sharply with the number of living children, from 16 percent among those with one living child to 82 percent among those with two children and 93 percent among those with 4 or five children. There has been a slight increase in the proportion of women who want to limit their childbearing, up from 63 percent in 2007 to 65 percent in 2011, while the desire to have another child has declined to 31 percent in 2011 compared to 34 percent in 2007 (NIPORT, Mitra and Associates and ICF International 2012). Thus, the vast majority of currently married Bangladeshi women want to either space their next birth or limit childbearing altogether. There is hardly any difference in the proportion of women who desire to limit childbearing in urban and rural areas. The desire to limit childbearing varies among administrative divisions. About half of currently married women with two children in Sylhet division do not want to have another child compared to 84 percent of women in Khulna and Rajshahi divisions. There are major differences in women s fertility preferences by level of education. The desire to limit childbearing is higher among women with no education (79%) than among those with completed secondary level of education (50%). There are differences in the desire to limit childbearing by household wealth-- women in the lowest wealth quintile are more likely to want no more children (69%) than those in the highest wealth quintile (60%) (NIPORT, Mitra and Associates and Macro International 2009). There have been rapid changes in fertility preferences among Indian women. During the 1992/ /04 period (from NFHS-1 to NFHS-3), the percentage of currently married women with one living child who want no more children doubled from 14 percent to 28 percent. The percentage with two living children who want no more children increased from 60 percent to 83 percent. In every population group, more than 60 percent of women with two or more living children want no more children. Also, within each group, the proportion of women who want no more children rises sharply with the number of living children (IIPS and Macro International 2007). About one-third (32%) do not want any more children, 38 percent cannot have another child because either the wife or the husband has been sterilized, and 3 percent say that they cannot get pregnant (i.e., they are declared in-fecund ). One-quarter (26%) of women would like to have another child (13% within two years, 12% after waiting at least two years, and 1% undecided when).the proportion desiring to limit childbearing increases sharply with the number of 5 For more detailed discussion on adolescent fertility, see IPPF_Facts on the Sexual and Reproductive Health of Adolescent Women in the Developing World, April

9 living children. Only 3 percent of women with no living children do not want any more children (the woman or her husband is sterilized or the woman says she wants no more children) compared to 83 percent and 90 percent of women with two and three living children respectively. As expected, older women are much more likely than younger women to want no more children. About one-third (32%) of women aged years want no more children compared to 78 percent and 90 percent respectively of women aged years and years. The proportion of women who want no more children is higher in urban areas (74%) than in rural areas (69%). The urban-rural differential is particularly large for women with one living child. There is no strong pattern by educational attainment overall; however, the desire to stop childbearing increases steadily with the level of education for women with 1-3 children, excepting those with less than five years of schooling. The proportion of women with two living children who do not want to have any more children is much lower for Muslim women (66%) than for women in any other religious group. Among women with two living children who want to limit childbearing, the proportion rises from 65 percent for women in the lowest wealth quintile to 92 percent for women in the highest wealth quintile (IIPS and Macro International 2007). Nearly half of married women in Maldives do not want any more children (37%) or have been sterilized (11%). Among those wanting another child, the majority (26% of all currently married women) either want to wait two years or more to have the next birth or are unsure about their childbearing intentions. Slightly less than half of the women who want another child (18% of all currently married women) want a child soon (within two years). As expected, the majority (75%) of women who have no children want a child soon. Among women with more than one child, the proportion wanting to limit childbearing increases rapidly, from 47 percent among women with two children to 96 percent among women with six or more children (Ministry of Health, Maldives and ICF Macro 2010). The proportion wanting no more children varies considerably with education. Higher proportions of women with primary or no education want no more children compared to women with secondary or higher education. Among currently married women with four or more children, there are only minor differences in the proportions that want to limit childbearing. However, among women with three or fewer children, fertility preferences vary considerably across subgroups. For example, 60 percent of women with two children in urban areas want to stop childbearing compared with 38 percent in rural areas (Ministry of Health, Maldives and ICF Macro 2010). In Nepal, 87 percent of currently married women in 2011either wanted to delay the birth of their next child or wanted no more children (including those sterilized). This is similar to that reported in the 2006 NDHS. Fertility preferences are closely related to the number of living children a woman has. In general, as the number of living children increases, the desire to want another child decreases. For example, 95 percent of currently married women with 5 living children say they want to have no more children or have been sterilized compared with 5 percent of women with no children (Ministry of Health and Population, Nepal, New ERA and ICF Macro 2011). More than half of currently married women aged years in Pakistan (52%) either do not want another child at any time in the future or are sterilized. More than four in ten women want to have a child at some time in the future (21% want one within two years, 20% would prefer to wait two or more years, and 2% are undecided as to when). Since the PDHS, there has been an increase (12 percentage points) in the proportion of married women who want to limit childbearing (from 40% to 52% in 2006) (NIPS and Macro International Inc. 2008). The desire to limit childbearing (including those women who are sterilized) increases with the number of living children, reaching 55 percent and 88 percent respectively among women with three living children and those with six or more children, while the reverse is true of those who want to have another child soon or later. Urban women (57%) are more likely than rural women (49%) to want to limit childbearing. A comparison with the PDHS shows that while the increase in the proportion of women who want to limit their childbearing is negligible in urban areas, a substantial increase (from 35% in to 49% in ) is evident in rural areas. Urban women express a desire to limit childbearing at lower parities than rural women. For example, 66 percent of urban women with three children want to stop childbearing compared with 48 percent of rural women. By province, Punjabi women are the most likely (54%) while Balochi women (37%) are the least likely to want no more children. In general, differences in fertility preferences by educational attainment are not pronounced. However, at parities 3 and above, these differences are pronounced because the desire for no more children is much higher among the more educated women than the uneducated or less educated women. The proportion of women wanting no more children is positively associated with the wealth status of the household, and the differentials are more pronounced among women with two to five children. In Sri Lanka, there is considerable desire among currently 9

10 married women to control the timing and number of births. About one-fifth (18%) would like to wait for two years or more for the next birth, and 60 percent either do not want another child or are sterilized. The remaining women are uncertain about their fertility desires or are unable to get pregnant (infecundity) (Govt. of Sri Lanka 2009) 6. Unmet Contraceptive Need Definition and Criticisms Unmet need for contraception (UCN) refers to the condition of currently married couples wanting to space or limit childbearing, but not using any FP method 7. The concept of UCN dates back to the 1960s, with the emerging evidence of a gap in the developing countries between women s fertility preferences and their use of contraception. The evidence was used to justify investments in FP programmes (Sonfield, 2006). Using data from the first set of the World Fertility Surveys (WFSs) from Asia, Westoff (1978) produced a five-country study of UCN, the phrase he substituted for KAP-gap, in an effort to develop morerefined measures of the gap between fertility preferences and contraceptive use. In his first study on the subject, Westoff, however, excluded pregnant and amenorrheic women, considering that they had no immediate need for FP. This limitation of the concept soon drew attention of others, including Westoff himself. Westoff and Pebley (1981) developed 12 alternative definitions of UCN, and showed that the different definitions produced different estimates of UCN. Also, they recommended that the concept should be enlarged to cover both the desires for spacing and of limiting childbearing. Subsequently, the Contraceptive Prevalence Surveys (CPSs) included questions about interest in spacing and limiting births, thereby, making it possible to calculate UCN for both spacing and limiting births. Nortman (1982), Nortman and Lewis (1984) and Khuda and Howlader (1990, 1988, 1986a, 1990) further broadened the definition by arguing that some pregnant, breastfeeding, and amenorrheic women should be included in the definition of UCN, because many of those women would require contraception soon after their current non-susceptible status ended. The Demographic and Health Surveys (DHSs) also attached due importance to the concept of UCN. In a comparative analysis of the DHS data, Westoff and Bankole (1995) showed that the low perceived risk of getting pregnant, indeed, accounts for a substantial fraction of UCN in many countries. Women s health and rights groups argued that the concept of UCN is too narrow, because: (i) it neglects reproductive health (RH) needs other than preventing births; (ii) it does not take into account potential clients other than married women; and (iii) the standard measure of UCN does not consider the degree to which women are dissatisfied with their present FP method (Dixon-Mueller and Germain 1992; Dixon-Mueller 1993). Women s groups, therefore, used the 1994 International Conference on Population and Development (ICPD) as the forum to shift the focus of population programmes from demographic goals and targets to women s lives, including but not limited to, their RH goals (Sen, Germain, and Chen 1994; McIntosh and Finkle1995). This resulted in the elimination of demographic targets, quotas, and goals. However, in the ICPD Programme of Action (POA) UCN received explicit mention as a core rationale for population programmes. The ICPD POA document stated Governmental goals for family planning should be defined in terms of unmet needs for information and services. All countries should, over the next several years, assess the extent of national unmet need for good-quality family-planning services (United Nations1994: paragraphs 7.12 and 7.16). Thus, reducing UCN became a target in itself, rather than a means for achieving demographic goals (Sai 1997). That UCN is real is evident from the high incidence of unintended pregnancies in developing countries, including the South Asian region. The rationale for UCN is based on its desirability of preventing such unintended pregnancies (Yinger 1998). Between one-fifth and one-quarter of births in the developing world are unwanted (Bongaarts 1997). However, this figure could be an under-estimate, given that a substantial fraction of pregnancies are terminated through induced abortion, the fraction of unwanted pregnancies must be even higher than the fraction of unwanted births, and a further fraction of recent births are reported as mistimed (Alan Guttmacher Institute 1999). The DHSs report that large number of recent births were aborted, unwanted, or mistimed, indicating the prevalence of UCN. In addition to the above criticisms about the concept of UCN, there are other criticisms. Economists would generally argue that if individuals actually wish to space or limit their childbearing, they would find the means to do so. Under the conventional economic theory, unmet need (which economists consider synonymously with unmet demand) can be viewed as a temporary disequilibrium which market forces would correct in the short run. From this viewpoint, nonuse of contraception simply demonstrates a lack of sufficient motivation (Demeny 1975; Pritchett 1994). However, this criticism originates 6 No information on fertility preference is available for Bhutan. 7 This section draws heavily from a comprehensive discussion on the concept of unmet need and related issues by Casterline and Sinding

11 from misunderstandings of the concept of UCN, and the absence of a sound behavioral model in much of the mainstream research on UCN. The key concept which has not been routinely articulated in the literature is competing preferences (Casterline and Sinding 2000). Social scientists argue that strongly held preferences will often not translate themselves into behavioral changes without a time lag due to obstacles in implementation of those preferences or because other preferences overrule them (see, e.g., Dawes 1998; and Pittman 1998). Preferences to avoid pregnancy are often constrained by various obstacles, which include predominantly fear of side effects of FP methods, health considerations and social opposition; and therefore, not surprisingly, a substantial fraction of pregnancies are reported as unintended (see, e.g. Khuda and Howlader 1986 b; RAND 1998; Ashford, L. 2003; DHSs). More women today choose not to use contraception either because they are concerned about the health risks and side effects of various FP methods or they find contraception too inconvenient to use. Between 6 percent and 28 percent of married women in the late 1980s cited one of these method-related reasons compared to between 19 percent and 36 percent of women more recently, suggesting that although FP programmes have been successful in educating women about their FP options, a number of both perceived and real risks associated with some forms of contraception continue to prevent use of FP (Maki 2007). However, Feyisetan and Casterline (2000) concluded that considerable increases in CPR can be achieved in the absence of changes in the demand for children by meeting the already-existing UCN. Unmet need and latent demand for FP have often been used interchangeably (see, e.g., Westoff and Bankole 1995; Sinding, Ross, and Rosenfield 1994). However, it needs to be emphasized that satisfaction of all UCN cannot be attained in the short term. Many women with UCN are unlikely to begin using FP any time soon, not necessarily because of their lack of access to FP services but because of their unwillingness to use FP on account of their perception of a low risk of conceiving or because of social, cultural, and health concerns (see, e.g., Pritchett 1994). However, a more balanced view is that some fraction of the estimated UCN in reality represents latent demand for FP, which can be converted into FP use by ensuring quality FP services. This view is supported by the analysis of DHS data by Westoff and Bankole (1996), who considered several scenarios in which only a subset of women with UCN adopts contraception. Thus, the FP programme by satisfying a fraction of the existing UCN would still be able to achieve considerable demographic impact, especially in countries with relatively low CPR, thereby reinforcing the rationale for a focus on UCN. The rationale is further reinforced by the need for FP services for purpose of pregnancy prevention. Extent and Differentials in Unmet Need in South Asia According to an estimate, there were 137 million women in the developing world with an UCN, and another 64 million with an UCN for a modern FP method (Sonfield 2006). Between 1990 and 2009, UCN declined quite slowly from 13 percent to 11 percent globally, from 11 percent to 9 percent in Asia and from 18 percent to about 15 percent in Southern Asia (UN, 2011). Using data from DHSs, the authors of the Guttmacher report concluded that between 1990/95 and 2000/05, UCN declined only 2 percent in sub-saharan Africa. In other regions of the developing world including Latin America and the Caribbean, North Africa and West Asia, and South and Southeast Asia UCN declined between 4 percent and 7 percent (Maki, 2007). In Asia, the prevalence of UCN among married women for both spacing and limiting childbearing is about the same. Unmarried women also add to UCN, accounting for 4 percent in Asia (Ross and Winfrey 2002). Levels of UCN vary considerably among subgroups of women, both at the regional level and within countries 8. Women who are young, uneducated, poor or living in rural areas or urban slums are generally at high risk of having an unintended pregnancy; and UCN is also high among such women. Among married women, UCN is highest among those aged years. Unmet need declines with age, and then, increases among older, high parity women. In South Asia, the extent of UCN ranges between 7 percent in Sri Lanka and 28 percent in Maldives. It is around 25 percent in Nepal and Pakistan, 23 percent in Afghanistan, and 12 percent each in Bangladesh and Bhutan (Table 3). Between 2000 and 2006, UCN declined from 33 percent to 25 percent in Pakistan, from 28 percent to 25 percent in Nepal, from 21 percent to 13 percent in India, and from 18 percent to 7 percent in Sri Lanka. Between 2007 and 2011, UCN declined from 17 percent to 12 percent in Bangladesh, though slightly higher than that in 2004 (11%). In Bangladesh, UCN declines with age, from 14 percent among adolescent women (aged years) to 8 percent among women aged years. UCN is lower in urban areas (9.9%) than rural areas (12.4%). UCN is highest in Chittagong division (19%) and lowest in Khulna and Rangpur divisions (8% each). There is very little differential by education and wealth status (Table 4; and NIPORT, 8 For more detailed discussion on UCN among specific population groups, see, for example, IPPF- Facts on Satisfying the Need for Contraception in Developing Countries Updated November

12 Mitra and Associates and ICF International, 2012). The higher UCN among the younger women, those living in rural areas and those in Chittagong division suggest that those women are less well served by the FP programme. Also, this may reflect problems with the supply of FP services and/or an increase in demand for FP supplies and services. There is a positive association between participation in household decision-making and UCN. Women who participate in household decision-making have lower UCN than those who do not. In contrast, UCN is higher among women who believe that wife beating is justified than those who do not hold that opinion (NIPORT, Mitra and Associates and Macro International Inc. 2009). In Bhutan, UCN was about 12 percent, though there are some differentials regionally. Also, it is considerably higher among adolescents (27.4%) than among women aged years (6.5%). It is positively associated with education of women, although the relationship with wealth status shows no clear pattern (Royal Govt. of Bhutan, UNICEF and UNFPA 2011). Between 1992 and 2004, UCN in India increased from 16 percent to 21 percent, but then, declined to about 13 percent in Of the 13 percent currently married women in India having UCN, 7 percent are limiters and 6 percent spacers (Table 3). UCN ranges from 5 percent in Andhra Pradesh to 35 percent in Meghalaya. More than one-fifth of women have UCN in Nagaland, Jharkhand, Bihar, and Uttar Pradesh. Unmet need for spacing ranges from 3 percent or less in Himachal Pradesh, Punjab, and Andhra Pradesh to 10 percent or more in Meghalaya, Mizoram, Jharkhand, Bihar, and Nagaland. Unmet need for limiting ranges from 2 percent in Andhra Pradesh to 16 percent in Nagaland. Similar to the national pattern, the unmet need for limiting is higher than that for spacing in most states of India (IIPS and Macro International Inc. 2007). UCN declines with age, from 27 percent among adolescents to only 2 percent among older women aged years. Younger women (15-24 years) have a greater UCN for spacing than for limiting, while the reverse is true of older women. The UCN for spacing decreases very sharply from years of age to years of age, beyond which it is negligible. The UCN for limiting increases up to years of age, and then, declines continuously. Rural women have higher UCN than urban women for both spacing and limiting. The UCN for spacing rises with increasing education through 8-9 years of completed education, but the UCN for limiting is highest for women with no education. Hence, the total UCN is about the same for women with different levels of education. UCN is particularly high for Muslim women and particularly low for Sikh and Jain women. UCN for both spacing and limiting declines with an increase in wealth (Table 4; and IIPS and Macro International Inc. 2007). Between 1991 and 1996, UCN in Nepal increased from 28 percent to 31 percent, then declined to 25 percent in 2006, and increased in 2011 to 27 percent 10% for spacing and 17% for limiting (Ministry of Health and Population, Nepal, New ERA and ICF Macro 2011). The decline in 2006 was much more pronounced among spacers (34%) than limiters (11%). Over the past decade, the decline in UCN was greater in rural areas (21%) than in urban areas (9%). However, UCN increased among urban women over the past 5 years (from 16% to 20%). UCN declined in all sub-regions, except in Eastern mountains, Western hill and Eastern terai, where it actually increased between 2001 and 2006 (Pradhan and Pant 2007). In the 1996, 2001 and 2006 surveys, UCN was higher among younger women, rural women, and women who lived in the mountain zone. However, the relationship between UCN and women s education varied. In general, as women s level of education rises, their level of UCN increased up to the primary level in 1996 and 2001, and up to the secondary level in 2006, and declined thereafter. There may be several reasons for this pattern. UCN is relatively low among women with little or no education, primarily because they are less likely to express a need for FP. As the level of education increases, women are more likely to be aware of the benefits of using FP, and thus, are more likely to express greater need for FP. This, coupled with an increasing ability to access FP services, has resulted in a decline in UCN among the more educated women. The interplay between education and UCN is also influenced by the availability of the preferred FP methods, the availability of a range of FP methods, exposure to information about FP methods, and the differential expectations of women across the country, especially across the three ecological zones. UCN was inversely associated with household wealth in the 2001 and 2006 surveys; however, this relationship was unclear in the 1996 survey. Also, there was no clear relationship between UCN and the number of living children and the number of living sons (Aryal et. al 2008). UCN among currently married women in Maldives is 28 percent: 15% for spacing and 13% for limiting (Table 3). UCN declines with age, from 36 percent among adolescents to 16 percent among older women; is slightly higher in rural than urban areas; and varies from a level of 25 percent in the North and Central regions to 36 percent in the South (Table 4; Ministry of Health and Family, Maldives and ICF Macro 2010). However, there is hardly any difference in UCN by educational level and wealth status (Table 4). UCN for spacing among women who agree with no reason for wife beating is higher (15%) than 12

13 among women who agree with reasons for wife beating (6%); and UCN for limiting is lower (13%) among women who disagree with any reason for wife abuse than those who agree with reasons for wife beating (19%) (Ministry of Health and Family, Maldives and ICF Macro, USA 2010). Between 1990 and 1997, UCN in Pakistan increased from 32 percent to 38 percent. The decline started in 2000 at 33 percent and further in 2006 at 25 percent (Table 3). Of the 25 percent with UCN in 2006, 11 percent were spacers and 14 percent were limiters (NIPS and Macro International Inc. 2008). UCN declines with age, from 20 percent among adolescents to 15 percent among older women (Table 4). As expected, UCN for spacing is higher among younger women, while UCN for limiting is higher among older women. UCN varies by women s education and household wealth status. However, the relationship of UCN with the background characteristics of the women has changed over time. In 1991, women from the poorest households had the lowest UCN; over time, UCN among these women increased substantially, and they now have the highest UCN (Sathar and Zaidi 2010). Unmet need, which was initially lower in rural areas, is now higher (26%) than in urban areas (22%). By region, Punjab has the lowest UCN (23%) and Balochistan and NWFP have the highest UCN (31%) (NIPS and Macro International Inc. 2008). UCN and the proportion of births that are unplanned and the high rate of abortion suggest that a large fraction of currently married women in Pakistan are at risk of an unwanted pregnancy and potentially an unsafe abortion due to lack of access to quality FP services, especially in rural areas of Pakistan and in Balochistan and NWFP. In Sri Lanka, UCN is considerably higher in the districts of the Eastern province (ranging from 15-23%) than in all the other provinces (4-9%). UCN declines from younger ages (13% and 10% respectively among women aged years and years) to older age groups (less than 1% among women aged 40 years and above) (Govt. of Sri Lanka 2009). Table 3 Unmet Need for Family Planning in South Asian Countries Country Year Age Total Spacing Limitting Data Source Afghanistan Govt. of Afghanistan 2006 Bangladesh 1993/ DHS 1996/ DHS 1999/ DHS DHS DHS DHS Bhutan Govt. of Bhutan 2011 India Nepal Pakistan 1992/ DHS 1998/ DHS 2002/ / DHS National survey DHS DHS DHS Govt. of Nepal / DHS 1996/ National survey 2000/ National survey 2006/ DHS Maldives DHS Sri Lanka DHS DHS Source: UN 2011; Govt. of Afghanistan 2006; Govt. of Bhutan 2011; Govt. of Nepal 2011 National survey District Level Household Survey on Reproductive and Child Health