78114 Closing the deadly gap Investing in women’s between what we know reproductive and what we do health Karen A. Grépin Assistant Professor of Global Health Policy New York University Jeni Klugman Director of Gender and Development The World Bank It’s time. Deliver for girls and women. Acknowledgements The authors wish to thank Emma Samman, Nistha Sinha, Susanne Schwartz, Sarah Twigg, Sarah Haddock, Josefina Posadas Samia Khan, and Jing Dong for their assistance with background research and in the preparation of the manuscript. Contents Chapter 1 Chapter 4 Introduction 3 Understanding underinvestment 13 in reproductive health Chapter 2 Women’s agency 13 Why investing in reproductive 5 Mutually reinforcing constraints 14 health is smart economics Figure 2: Limited progress in reproductive health 14 Labour productivity effects 5 is explained by mutually reinforcing constraints Financial well-being of households 6 Limited accountability 15 Future generations 6 Figure 3: Number of countries with data on 15 Broader economic returns 7 skilled birth attendance, five-year intervals Box 1: Impact of maternal mortality declines 7 Chapter 5 on female schooling in Sri Lanka What are the policies that can improve 17 Chapter 3 reproductive health outcomes? What factors contribute to poor 9 Measures to expand women’s and girls’ agency 17 reproductive health outcomes? Efforts within the health system 19 Figure 1: Relationship between MMR 10 Improving accountability for reproductive 20 and income, education, and gender inequality health outcomes Table 1: Determinants of MMR from 11 Box 2: How Peru improved reproductive health 21 1990-2010 across 158 countries with outcomes through expanding entitlements and comparable international data strengthening public sector management Chapter 6 Conclusions 23 References 25 1 Investing in women’s reproductive health Pregnancy and the consequences of childbirth remain the leading causes of death and disability among women of reproductive age in developing countries today. 2 Introduction 1 Gender equality has moved to the forefront of development The main contributions of this report are fourfold: first, debates. A key message from the 2012 World Development it brings together the evidence highlighting the economic Report (WDR2012) is that gender equality is a core develop- benefits of investing in reproductive health, which confirms ment outcome in its own right: Greater gender equality can that addressing the reproductive needs of women is indeed enhance productivity, improve development outcomes for smart economics; second, it explores the determinants of future generations, and make institutions function better.1 poor reproductive health outcomes by making a quantitative Addressing the reproductive health needs of women is a comparison across countries; third, it examines the multi- prerequisite to achieving gender equality, but despite plicity of factors that contribute to poor reproductive health international commitments, actual progress on this front outcomes at the country level, building on the insights of has been slow and remains a blight on global development.2 the analytical framework put forth in the WDR2012; and Improving reproductive health outcomes and gender lastly, it reviews the evidence that demonstrates the effective- equality outcomes are inextricably linked. ness of key policy levers to accelerate reproductive health improvements in developing countries. Pregnancy and the consequences of childbirth remain the leading causes of death and disability among women of The basic message is that investments in reproductive health reproductive age in developing countries today.3 Millions are a major missed opportunity for development. Effective of women lack the means to prevent unwanted pregnancies, and affordable interventions are available to improve and to prevent and address complications and disease during reproductive health outcomes in developing countries,7 and pregnancy. Despite some global progress in tackling maternal the challenge is less about identifying these interventions mortality, the risks remain unacceptably high: there were an but rather in implementing and sustaining policies to put estimated 287,000 maternal deaths in 2010—close to 800 proven packages of interventions and reforms into practice.8 per day on average—more than half of which occurred in This requires addressing women’s agency, making improve- sub-Saharan Africa, and just under a third in South Asia.4 ments in the delivery of health services, and increasing The maternal mortality ratio (MMR) in developing regions accountability in health systems. However, the lack of (240 deaths per 100,000 live births) is a massive 15 times reliable data and evidence continues to hamper progress and higher than in developed countries. This comparison must be seriously addressed at the country and global levels. between the burden of maternal mortality in developed Only then can a virtuous cycle be initiated that will lead and developing countries has long been cited as the to sustained improvements in reproductive health. “widest disparity in all statistics of public health.�5 In several countries—Chad, Guinea-Bissau, Liberia, Niger, Sierra Leone, and Somalia—at least one of every 25 women will die from complications of childbirth or pregnancy. Moreover, maternal mortality is just the tip of the iceberg— for every woman who dies another 30 suffer long-lasting injury or illness.6 3 Investing in women’s reproductive health Women are important contributors to the global economy: About 40% of the global labour force and more than 60% of workers in agriculture in sub-Saharan Africa are women. 4 Why investing in reproductive health is smart economics 2 To date, most research on the benefits of reproductive health High fertility can also affect female labour supply. Over time, programmes has focused on the health benefits—in terms as fertility has declined, women have tended to become a of reducing maternal mortality and morbidity—which are larger share of the labour force, most notably in Latin substantial, and which alone justify greater investment.7,9–11 America and the Caribbean where female labour force The economic benefits of such investments, from the participation rates surged by 15% over the past decade, and individual through to the economy-wide levels, are less at the same time the total fertility rate decreased by almost well recognised. Below we review the evidence of how 11%.18,19 Understanding the direction of causality between investments in reproductive health can generate economic fertility and labour force participation is a challenge. benefits. Because there is likely to be feedback between To address this empirical problem, a recent cross-country these measures—in that improvements in economic status analysis uses an instrumental variable approach. The paper may improve reproductive health outcomes—we focus on finds that the effect of fertility on female labour supply is studies that attempt to control for this potential endogenous strongest during the fertile years. Globally, female labour response.12 Measuring the economic returns on investments force participation decreases with each additional child: in reproductive health is difficult, and we do not attempt to by about 10 to 15 percentage points among women aged aggregate the costs, given the scarcity of studies and the risk 25 to 39, and about 5 to 10 percentage points among of double counting. It is nonetheless clear that the scope women aged 40 to 49.20 for economic gains should be a critical consideration for policy makers around the world. Various factors help to explain this correlation, including rising levels of education. In Timor-Leste, for example, total fertility rates vary from 6•1 births per woman with Labour productivity effects no education to only 2•9 births for women with secondary Women are important contributors to the global economy: schooling or above.21 At the same time, some evidence About 40% of the global labour force and more than 60% of suggests that the relationship between fertility and labour workers in agriculture in sub-Saharan Africa are women.13 supply is at least partially causal. In the U.S., exogenous Poor health reduces labour supply and contributes to lost increases in fertility are associated with a 10% lower wages,14,15 so improved reproductive health outcomes can probability of the mother being in the labour force.22 increase female labour supply and productivity and there- Qualitatively similar results have been found in Argentina fore should be of great concern to policy makers. and Mexico, countries with very different labour markets and institutions.23 Legalisation of abortion, which leads For every maternal death, many more women suffer to reduced total fertility, has also been linked to higher maternal morbidities, especially in developing countries.16 female labour force participation.20 Evidence from Bangladesh has documented the losses in productivity from maternal health conditions, in particular It is not only the total level of fertility that affects labour for severe complications of pregnancy.17 Maternal mortality supply but also the timing of births. Around 16 million and morbidity imposes costs in terms of foregone earnings, adolescent girls aged 15 to 19 give birth each year—almost and also means that family members need to absorb the 95% of whom live in low- and middle-income countries. work done by women inside the home, which might reduce In Latin America there are almost 74 births per 1,000 girls their own ability to exploit outside economic and educa- aged 15-19, in South Asia around 77 and in sub-Saharan tional opportunities. Africa there are as many as 120 births per 1,000 girls in that 5 Investing in women’s reproductive health age group.24An estimated three million girls in this age families bear significant financial burdens due to paying for group undergo unsafe abortions each year, and in low- and maternal health care.37 middle-income countries, complications from pregnancy and childbirth are the leading cause of death among girls While poverty is a contributing factor to maternal mortality, aged 15 to 19.24 If women are able to better control their there is also evidence that it further impoverishes house- fertility through, for example, access to family planning, holds. A study in three rural provinces of China compared then as girls they can also stay in school longer, accumulate households with a recent maternal death to matched more skills, and eventually earn higher wages. households with a recent birth but where there was no maternal death.38 The direct health expenditures of a In Bangladesh, the age of menarche influences how long maternal death were more than six times higher than those girls remain in school by delaying marriage.25 In Colombia, with a successful birth outcome. Families also had to pay when family planning programmes were expanded from funeral expenses, which were substantial. Affected families the mid-1960s, women were able to delay their first birth, also had substantial indirect financial costs, including enabling more schooling and making them more likely to substantial days of missed work and lost wages and many be employed in the formal sector.26 Hence early pregnan- reported having to borrow to fill the gaps. The total cies—especially among school-aged girls—may reduce economic burden of a maternal death was estimated at future earning potential. However, studies that try to more than a full year of household income. ascertain the causal effects of early pregnancy have had mixed results and conclusions seem to be sensitive to the methodological approach used.27 Future generations Poor reproductive health can adversely affect the economic As noted above, the relations between fertility, schooling prospects of the next generation. The most extreme impacts and economic opportunities run both ways. Not only does arise when a woman or her baby dies in childbirth, but lower fertility potentially improve economic outcomes, but maternal ill-health can also affect her children’s well-being also better economic opportunities may reduce fertility and schooling. rates. This is illustrated by a recent randomised experiment in India that helped to provide work opportunities for young Many studies have investigated how the loss of a parent women to enter the call centre industry. This led to fewer impacts human capital investments in children, and women getting married and to more women reporting studies in low-income countries have generally found that wanting fewer children.28 orphans have worse health39 and less schooling than other children.40 Notably, the studies also find that the death of a mother tends to affect children more adversely than the Financial well-being of households death of a father.41,42 In South Africa, when a mother dies, A large economics literature has studied the repercussions of her children are less likely to be enrolled in school and poor health for household consumption and other measures complete fewer years of schooling than children whose of household’s financial well-being around the world.29,30 mothers are still alive.41 In Tanzania, the estimated effect A key channel is the associated costs and consequences of of a maternal death could be as large as one full year of paying for health care. It is well established that catastrophic schooling.43 However these studies have largely focused on health costs can impoverish families in developing countries the impact of deaths resulting from human immunodefi- —households frequently report having to borrow or sell ciency virus (HIV)/acquired immune deficiency syndrome assets to pay for health services.31,32 Evidence is accumulating (AIDS), which may differ from maternal illness, which is that poor maternal health imposes important economic an area where more research is needed. burden on households. Families might also increase investments in their A study in Bangladesh has documented large reductions children if they expect them to live longer. In Sri Lanka in household resources associated with poor maternal between the 1940s and 1960s, in response to a major health outcomes, most of which were driven by large health reduction in maternal mortality, rates of female schooling expenditures.33 In Ghana and Benin, the costs of complica- increased, suggesting that families increased their invest- tions at delivery cost represented an estimated one-third ments in their daughters in response to expected longer of the annual household cash expenditures.34 In Burkina lifespans (Box 1).44 Faso, women with more severe obstetric complications reported more frequent sale of assets, more borrowing, and Poor maternal health can inhibit the development of slower repayment of debt in the following year.35 In Yemen, children through a mechanism known as the foetal origins in a survey of women who had suffered previous reproduc- hypothesis.45 Evidence suggests that children, even in tive morbidity, 43% reported having to sell assets or take out wealthier countries, benefit greatly later in life if their a loan in order to pay for their care.36 Qualitative research mothers are well fed and receive adequate health care also in Bangladesh further supports the evidence that during pregnancy: benefits include their subsequent 6 health,46,47 schooling,48 and labour market outcomes.49 Broader economic returns A long-term study in Bangladesh found that children born Even if economic benefits can be generated at the individual to mothers who received tetanus vaccines during pregnancy and household levels, it remains to be seen whether subsequently received more schooling.50 These studies population-level programmes produce economic gains in suggest that investing in routine reproductive health the aggregate when general equilibrium effects are taken services, in particular antenatal care and maternal nutrition, into account.15 In sub-Saharan Africa at least, it has been can also have long-term benefitsfor future generations. shown that there is a negative relationship between high levels of maternal mortality and gross domestic product.53 In classic economic models of fertility, since families have One mechanism that has been shown to link population- limited resources and it is costly to raise children, the more level improvements in reproductive health to broader children a family has, all else being equal, the fewer economic returns is the demographic dividend. This suggests resources that are available per child.51 It follows that family that countries can benefit economically from rapid shifts in planning might help families have fewer children and the agecomposition of the population, triggered by improve- to invest more in each child. Family planning can also ments in population health and fertility decline.54 This effect influence the ability of families to better control the spacing has been empirically documented in countries from Asia between births. Evidence from the U.S. suggests that to Europe.55,56 Where fertility rates remain high and there increases in birth spacing between siblings can improve is a large unmet need, greater access to voluntary family test scores of older siblings, possibly for physiological planning programmes might help countries speed up the reasons or by allowing families to devote more resources demographic change to capture the benefits from the and time to individual children during childrearing.52 demographic transition. Impact of Maternal Mortality declines on Female Schooling in Sri Lanka box 1 The case of Sri Lanka is well known as government turned towards improving literacy improved by 0.7 percentage a success story for improving maternal the quality of care through improve- points (2%) or 0.1 years of schooling health, proving that with the proper ments in obstetric care and family (3%) relative to males. investments dramatic and quick planning programmes. By 1999, 66% improvements can be made even in a of births occurred in an institution Notably, we are able to learn from the poor country.57 Sri Lanka was able to staffed with an obstetrician.58 case of Sri Lanka because—from the reduce its maternal mortality from very beginning—the government over 2,000 deaths per 100,000 live The case of Sri Lanka also shows prioritised the collection of high births in 1930 to just 35 today. The that improving maternal health quality data on maternal mortality and government achieved this through a has important benefits for future good monitoring and evaluation of its comprehensive strategy to expand generations. From 1946–1953, one of programmes. Administrative data on universal coverage of health services, the most dramatic declines in maternal maternal mortality in Sri Lanka during including to rural areas. By 1948 the mortality that has ever been recorded the 1940s still exceeds the quality of government-funded health care took place and maternal mortality data available in most developing delivery system had reached the declined by 70%.44 As maternal health countries today. entire island.58 Malaria and hookworm improved, families invested more in control, together with modern medical the education of girls as they would advances, contributed to declines in now be expected to live longer and maternal mortality during this period. more productive lives. This is shown The government meanwhile invested empirically by looking at districts with heavily in the training of midwives and different levels of initial maternal development of a country-wide mortality and comparing changes in delivery network, such that by 1950, girls’ enrolment to boys’ (who did not 58% of births had skilled attendants.58 benefit directly from the improvement During the 1960s and 1970s the in life expectancy). For every year of increased life expectancy, female 7 Investing in women’s reproductive health Previous research has suggested that income per capita, female education, the rate of skilled attendance at birth, and health expenditure are all correlated with levels of maternal mortality. 8 What factors contribute to poor reproductive health outcomes? 3 This section explores the determinants of the maternal First, to visualise the relationship we plot MMR in 2010 and mortality ratio (MMR) across countries. We focus on this measures of income per capita, female primary education measure as a proxy, albeit the most extreme manifestation completion rates, and the gender inequality index during of, poor reproductive health outcomes. Previous research roughly the same time period.iii These plots are shown has suggested that income per capita, female education, in Figure 1 (page 10). We can see that all three of these the rate of skilled attendance at birth, and health expenditure factors appear to be associated with MMR in similar ways: are all correlated with levels of maternal mortality.59–62 Countries with higher levels of income, higher female Methodologically, we improve upon these models by education, and higher gender equality all have lower MMR, analysing reproductive health outcomes in a wider set of higher skilled birth attendance rates, and lower total fertility countries and over time, and exploring the contributions rates. However, none of these factors alone are sufficient of gender inequality. to fully explain the variation in reproductive health outcomes—at any given level of income, education, or For this analysis we use the most recent estimates of measured gender inequality, large differences in reproduc- MMR produced by the UN Maternal Mortality Estimates tive health outcomes remain. The impact of these factors Inter-agency Group (May 2012 estimates). These estimates also appears to level off at higher levels of income and have been generated from 1990–2010 and are available at education and in countries with greater gender equality. five-year intervals. Data on gross domestic product per capita, the female primary education completion rate, the Next, we improve upon these simple relationships and total fertility rate, the percentage of parliamentary seats in a develop a multivariate model to explore the determinants single or lower chamber held by women, and the proportion of MMR across countries and over time. None of the of the population living in urban areas are from the World previously mentioned studies employed panel data to Bank’s World Development Indicators (WDI). Data on investigate the determinants of maternal mortality. For adult HIV prevalence is from the Joint United Nations each country, we observe MMR in those countries at Programme on HIV/AIDS (UNAIDS).i We also utilise the five-year intervals from 1990 to 2010. Since we are using Gender Inequality Index, a composite measure of the estimates of MMR where some of the data have been inequality between men and women in terms of education, generated from regression models that use some of the political representation, labour force participation, and same variables we are interested in, we first develop a base reproductive health outcomes produced by the United model that includes these variables (per capita income, Nations Development Programme (UNDP).ii The sample total fertility rate, skilled birth attendance, and HIV/AIDS of 158 countries includes all countries with population prevalence) since it is not surprising that they are at least greater than 500,000 for which data is available. partially correlated. We also include a dummy variable in i HIV prevalence data were missing for a small number of countries. Where possible additional sources of data were used to estimate prevalence but in the absence of other estimates, it was assumed that the HIV prevalence was 0.1% in approximately 30, largely Middle Eastern, countries. ii The gender inequality index includes three dimensions of gender disparities: labour market (women’s labour force participation), empowerment (share of parliamentary seats held by women and the secondary and higher educational attainment of women), and reproductive health (maternal mortality and adolescent fertility). It varies between zero (full equality) and one. We use the most recent measure of the gender inequality index, which was based on 2011 data. In order to facilitate the comparisons, we compare reproductive health iii  outcomes to the inverse of the gender inequality index. Measures of income per capita and female education are from 2010. 9 Investing in women’s reproductive health Source: Authors’ analysis based on World Development Indicators and UNDP data. Relationship between MMR and income, education, and gender inequality FIG.1 MMR and income 1200 1200 1000 1000 800 800 MMR 600 600 400 400 200 200 0 0 6 7 8 9 10 Log GDP per Capita 6 7 8 9 10 MMR and education 1200 1200 1000 1000 800 800 MMR 600 600 400 400 200 200 0 0 20 40 60 80 100 120 Female Primary Education Completion Rate 20 MMR inequality 60 and gender40 80 100 120 1200 1200 1000 1000 800 800 MMR 600 600 400 400 200 200 0 0 0.2 0.4 0.6 0.8 1.0 1-Gender Inequality Index 10 0.2 0.4 0.6 0.8 1.0 our models for countries with incomplete vital registration education. When both the percentage of women parliamen- systems in order to capture the effect of model-based tarians and female education are included into the model, estimates. We then augment the models with additional both effects remain significant and relatively unchanged variables, including data on female education and the share from the earlier specifications suggesting that both have an of seats in parliament held by women, a commonly used important effect on maternal mortality. measure of gender equality that has good internationally comparable data since the early 1990s. Many factors are associated with MMR in our cross-country models. Similar to previous studies, this analysis confirms The results of these regressions are presented in Table 1. that across the countries included in our sample, countries In column 1 we see that income per capita, skilled birth with higher levels of income, more urban populations, attendance, total fertility rate and adult HIV prevalence higher levels of utilisation of skilled birth attendants, and are all associated with maternal mortality, consistent with lower fertility rates have lower levels of maternal mortality. previous studies. When we also include girls’ primary In addition, this study also finds that countries where education completion rate and women’s parliamentary women hold a higher proportion of seats in parliament representation, we see that these factors are also important and where a higher proportion of girls complete primary determinants of maternal mortality. Countries with a education we also see lower levels of maternal mortality. higher proportion of female legislators and where more Although it is impossible to rule out other factors that might girls complete primary school also have lower maternal be confounding this relationship, it does suggest that the mortality rates. Interestingly, when education is included processes of improving gender equity and reproductive in the model, the effect of income is no longer significant, health outcomes are inextricably linked. Below we further suggesting that the association between income and MMR expand on these finding by reviewing micro-level evidence is largely driven by investments these countries make in that also supports these findings. Determinants of maternal mortality, 158 countries with comparable international data, 1990–2010 Tab.1 VARIABLES (1) Base (2) Parliament (3) Education (4) Both -2.41** -1.86** Parliament Seats Held By Women (%) (0.643) (0.587) -2.92** -2.88** Female Primary Education Rate (0.381) (0.377) -50.55* -55.24** -15.66 -15.58 Log GDP Per Capita (19.624) (19.491) (19.711) (19.512) -4.90** -5.10** -3.69** -3.85** Urban Population (%) (1.515) (1.503) (1.397) (1.384) -2.47** -2.34** -1.89** -1.81** Skilled Birth Attendance Rate (0.561) (0.557) (0.522) (0.517) 72.37** 74.01** 42.88** 43.82** Total Fertility Rate (10.883) (10.797) (10.279) (10.179) 13.74** 14.50** 12.30** 12.90** Adult HIV Prevalence (%) (1.994) (1.986) (1.746) (1.738) -408.59** -403.66** -295.01* -280.48* Lacks Vital Registration Data (128.835) (127.659) (119.236) (118.120) 990.01** 1,050.78** 873.74** 889.76** Constant (206.566) (205.324) (204.181) (202.181) Observations 661 657 585 585 R-squared 0.951 0.952 0.959 0.960 Standard errors in parentheses ** p<0.01, * p<0.05 All models also include country and year fixed-effects 11 Investing in women’s reproductive health The pervasiveness of child marriage, and its association with high fertility, reflect the crucial need for expanding comprehensive sexual and reproductive health services that reach married adolescents. 12 Understanding underinvestment in reproductive health 4 Despite advances in gender equality, potential major attended by a medically trained provider than those with economic gains, manifest health benefits and intrinsic more autonomy.66 Policies promoting women’s awareness, value of investments in reproductive health, alongside knowledge and autonomy, linked to efforts to promote the well documented technical solutions, too little economic opportunities, are therefore crucial to improving progress has been made in addressing reproductive reproductive health outcomes. health needs. Why do these gaps persist? The constraints on younger women’s agency may be more severe and thus may require special attention. Child Women’s agency marriage remains common, with an estimated 60 million Agency emerges as a very important part of the story. In women globally aged 20–24 years being married before the capabilities approach, as conceived and developed by they were 18 years old.67 This represents a global average of Nobel Prize-winning economist Amartya Sen, agency is around 36% of all girls marrying before they turn 18, but the ability to pursue goals that one values and has reason regionally the rates are as high as 45% in South Asia,68 with to value,63and an agent is “someone who acts and brings a peak of more than 82% of girls in Bangladesh.68 Child about change�.64 The concern is with both processes marriage has been linked to psychological and health risks, (intrinsic) and outcomes (instrumental). Agency, as defined including vestico-vaginal fistulae and a higher likelihood in the WDR2012, is the ability to use endowments to take of acquiring HIV.68 However, the practice has also been advantage of opportunities to achieve desired outcomes. linked with increased risk of violence and decreasing rates Constraints on this ability can include social and cultural of schooling, and with increased risk of teenage pregnancy, norms, lack of services, institutional discrimination, and which can have a range of detrimental impacts as so on. Given measurement challenges, the focus tends discussed below. to be on the institutional conditions for the exercise of agency, a practice that we follow in this paper. Concepts Recent work in India has confirmed extremely high rates of agency and constraints on its exercise help to cast of child marriage—almost half of women aged 20–24 years useful new light on observed patterns of reproductive had been married before 18 years of age, almost 23% were health outcomes, and the performance of various types married before age 16, and almost 3% were married before of programmes. 13 years. Among women aged 20–24, the study found that child marriage is significantly associated with decreased use The WDR2012 highlighted women’s voice and agency as of contraceptives prior to the birth of the first child, high a critical area where progress over time has been especially fertility (three or more births), multiple unwanted pregnan- slow—and underlined the critical links between agency cies, short birth spacing and female sterilisation.69 The and reproductive health. In Papua New Guinea, for pervasiveness of child marriage, and its association with example, almost three out of every ten women cite lack of high fertility, reflect the crucial need for expanding compre- knowledge as the main reason for not using contraception. hensive sexual and reproductive health services that reach In 37 countries with recent data, 12% of women state that married adolescents. violence is justified when a wife refuses to have sex with her husband—and was as high as 45% in Burkina Faso.65 In Bangladesh, less autonomous women were less likely to receive antenatal services or to have deliveries 13 Investing in women’s reproductive health Mutually reinforcing constraints the country level (as shown in Figure 1). There are also large gaps in outcomes within countries.75 Women disadvan- Figure 2, adapted from the WDR2012, illustrates how taged by class, caste, location and ethnicity experience far good reproductive health outcomes result from the interplay worse reproductive health outcomes than other women.72,76,77 of a number of factors. It shows that the key outcomes of In Viet Nam, for example, 60% of women from ethnic gender equality—in terms of endowments, agency and minorities give birth without prenatal care, twice the rate opportunities—are inter-related, and are influenced by the as that for the majority Kinh women.1 interplay of markets, formal institutions (like public health systems) and informal institutions (including norms and Formal institutions matter a lot and poorly functioning attitudes), as well as decision making within the household. health services contribute to poor reproductive health These factors interact in important ways: for example, poor outcomes. In 2008, fewer than two thirds of births globally women have lower access to high quality health care services and less than 47% in Africa had skilled birth attendants.78 than wealthier women,70-72 while low quality of maternal In Nepal, only 44% of pregnant women receive antenatal health services often deters utilisation.73,74 care, and fewer than one in five deliver with the assistance of skilled health personnel.75 Here, as elsewhere, there are Sometimes these factors move in the wrong direction and major differences across the distribution—in India, nearly vicious cycles can impede progress, but the implementation nine out of ten women in the richest quintile have assistance of innovative policies can help to break and reverse their during delivery while only two out of ten in the poorest direction, beginning virtuous cycles that can improve quintile do.75 reproductive health. This visualisation helps to underline that women often face dual or compounding disadvantages Barriers to accessing health care vary across countries and where poor reproductive health is associated with other regions, but financial and physical barriers, inadequate sources of disadvantage. transportation options, low quality health services, and the lack of knowledge about where to deliver are frequently cited Economic growth is necessary, but not sufficient to improve as reasons for not seeking care.79–82 Unaffordability was cited reproductive health outcomes: There is great variation in by three out of four Cambodian women with problems reproductive health outcomes at similar levels of income at Source: Adapted from World Bank 2012. Limited progress in reproductive health is explained by mutually reinforcing constraints FIG.2 Changing social norms and attitudes Changing household dynamic and decision-making Da ta INFORMAL  INSTITUTIONS economic opportunity Reproductive MARKETS HOUSEHOLDS endowments health outcomes Increasing economic agency opportunities FORMAL INSTITUTIONS  ce id en Ev Improving health system delivery 14 accessing maternal health care and 38% of women in there is little accountability in the system to reinforce the Pakistan. In Burkina Faso, 46% of women cited long momentum towards better reproductive health outcomes.87 distances as a main barrier in accessing health care, and We see better data and evidence as oil that can help grease another 40% cited lack of transportation.75 the wheels of both the administrative and political systems, to help achieve better outcomes. Part of the institutional challenge is the lack of resources, in particular human resources, in the health system. In a Fortunately, there have been some improvements on the cross-section of countries, low density of health-related data front. Data on skilled birth attendance, for example, human resources was found to be correlated with high levels is available for the past five years for about 145 countries— of maternal mortality.83 Tanzania has extremely low numbers compared to fewer than 80 countries in the preceding five of health workers (only 0•008 physicians and 0•24 nurses/ years and fewer than ten countries in the 1990s (Figure 3). midwives per 1,000 people) and Chad, the country with the However, reporting frequency is low for many countries, highest MMR globally, has only 0•28 nurses/midwives per since only about half of those countries update annually. 1,000 people.75 This is exacerbated by low quality of care Cross-country comparisons also remain an issue, given the from health care providers and high rates of absenteeism.84,85 variability across countries in data sources. There needs to be a much greater investment in collecting accurate and timely data on reproductive health. Limited accountability Another part of the problem is lack of accountability for improvements in reproductive health, which is exacerbated by the paucity of quality data to measure, monitor and track maternal mortality. A recent report suggests that only 11 of 75 high priority maternal and child health countries have data on 11 core maternal and child health indicators and many have no data at all.86 Without strong data systems and better evidence on the effectiveness of programmes, Source: Authors’ analysis based on World Development Indicators and UNDP data. Number of countries with data on skilled birth attendance, five-year intervals FIG.3 145 150 120 90 78 60 30 7 3 0 2007–2011 2002–2006 1997–2001 1992–1996 15 Investing in women’s reproductive health The extent and nature of women’s and girls’ agency, including the power to make decisions and to challenge authority, has major ramifications for their reproductive health. 16 What are the policies that can improve reproductive health outcomes? 5 A number of policies have been shown to be effective in girls’ and women’s aspirations. In Bangladesh, the expansion improving reproductive health outcomes. Below we review of job opportunities for women in the garment sector was some of the most promising policy options that can be associated with a rapid increase in girls’ schooling—between used to initiate more virtuous cycles in reproductive health 1983 and 2000, those villages within commuting distance outcomes. We group these interventions into those that to garment factories saw a 27% increase in girls’ school expand women’s agency, those that target improvements enrolment rates.93,94 Similarly, reforms to inheritance within the health sector, and those that increase account- laws in India, which gave women greater economic power, ability for reproductive health outcomes. similarly resulted in delays in marriage for girls, an increase in years of schooling by 11–25%, and lower dowry payments.95 Measures to expand women’s and girls’ agency The extent and nature of women’s and girls’ agency, In many developing countries, women’s lack of agency, while including the power to make decisions and to challenge exacerbated by limited education, can be traced to how the authority, has major ramifications for their reproductive formal legal system is designed and administered. Even in health. While the effect of interventions tends to be the 21st century, legal barriers—such as laws requiring highly context specific, some policies and programmes parental notification of their daughters’ intended abortion have expanded agency effectively, either by expanding or laws that allow husbands or partners to veto wives’ or opportunities, or through increased awareness, knowledge girlfriends’ use of contraception—prevent girls and women and aspirations. Education is a primary vehicle for from accessing family planning programmes and necessary delivering both such opportunities and knowledge, safe delivery services.96 These barriers contribute to delays and indeed mothers’ education and maternal and child in accessing essential maternal health services.97 health are robustly correlated. As a result, legal reforms can play an important role in Educational gains for women can be directly linked to improving reproductive health outcomes, particularly with reproductive health: In the U.S., in areas where women respect to access to abortion services. Nearly 22 million gained access to colleges, women were more likely to use unsafe abortions occurred globally in 2008, many of them antenatal care.88 Female education also appears to affect in developing countries—over half of all abortions in subsequent fertility: In Indonesia, increases in female developing countries are unsafe, compared with just 6% of education was associated with falls in both fertility and abortions in developed countries.98 While highly restrictive child mortality,89 and in Nigeria, one additional year of abortion laws do not tend to lower abortion rates, they female schooling lowered fertility by 0•26 births.90 More typically do make it unsafe. The 82 countries with the most schooling among teenage girls in Malawi has been linked to restrictive abortion legislation are also those with the highest lower rates of teenage pregnancy.91 While most studies have incidence of unsafe abortion and abortion mortality rates.99 focused on primary education, it is likely that post-primary By contrast, where abortion is permitted on broad legal education is also important. This points to the potential for grounds, it is generally safer.100 In Bangladesh, Romania and tremendous future gains, since the gap between boys and South Africa, for example, abortion policy liberalisation, girls at the secondary level is more stark.92 coupled with implementation of safe abortion services and other reproductive health interventions led to dramatic Greater economic opportunities can alter traditional declines in abortion-related mortality. In Romania, definitions of gender roles, duties and responsibilities, and following policy reform in 1989, the abortion-related 17 Investing in women’s reproductive health mortality ratio dropped from a high of 148 deaths per increasing education, health and employment prospects. 100,000 live births in 1989, to 5 deaths per 100,000 live These interventions include targeted information and access births in 2006.101 Similarly in Bangladesh the proportion to contraceptives, which in turn directly increase the range of maternal deaths due to abortion dropped by more than of choices teenagers have available to them.27 In Peru, half from 24% in 1976–1985 to 11% in 1996–2005 in CCTs increased school attendance, which was identified areas receiving specific interventions in addition to legal as a mechanism for reducing fertility among beneficiaries liberalisation. In South Africa, the annual number of of Juntos.27 In Chile extended school-hours programmes abortion-related deaths fell by 91% between 1994 and 2001 were found to reduce teen pregnancy.103 following liberalisation of the abortion law.102 Because teenage pregnancy is negatively correlated with a Data suggests that the incidence of child marriage has fallen number of socioeconomic outcomes, understanding the over time, however this has taken place relatively slowly, causal effect of early childbearing is challenging. Part of the and the drivers of those reductions are not well known.68 negative outcomes attributed to teenage pregnancy might be Child marriage is prohibited by law in many countries, but due to the poor educational and economic opportunities often with limited effect. In India for example, marriage that young women (and young men) face. In this sense, early before the age of 18 has been illegal for about three decades, pregnancy can be a consequence rather than a cause of yet about half of all girls still marry before the age of 18. socio-economic disadvantage. Knowing that poverty and Similarly in Nigeria, legal limitations on the age of marriage lack of opportunities are key determinants of early child- have not fundamentally altered the practice.68 As explored bearing confirms the importance of focusing holistically on further below, policies targeting education—such as expanding young women’s range of available opportunities conditional cash transfers (CCT)—may also be effective and strengthening their agency to make effective choices.27 in reducing child marriage. While fertility is often associated with school dropout, for example, other factors such as marital status, aspirations and In Latin America, the risk of adolescent pregnancy has anticipated economic returns to education, and family and been addressed through policy interventions that focus on social attitudes, are also related to school continuation and 18 the likelihood of experiencing adolescent pregnancy.104 In technologies and pharmaceuticals such as sulfonamides, this light, a more holistic approach to reducing adolescent as well as changes in the organisation and delivery of fertility may be needed that focuses on addressing the maternal health services.107,108 underlying causes as well as providing reproductive health services. A recent review of interventions found that those Over the past decade, many governments have sought to that encouraged school attendance proved more effective in reduce the financial barriers to accessing maternal health reducing overall adolescent fertility, while those that focused services. Although some results are mixed, and no studies on increasing knowledge and changing attitudes about have investigated the direct impacts on health outcomes, sexual and reproductive health—such as through peer there is sufficient evidence to conclude that these policies education or school-based workshops—were successful in are promising, especially in contexts where demand-side the short term (within the first year), but with no sustained constraints are major barriers to service utilisation. change in attitudes over the longer term (two years post-intervention).104 A number of countries, responding to concerns about the regressive impact of user fees, have exempted pregnant Experience is accumulating around the world about the women or certain maternity services. Such policies have ways in which programmes designed to promote education, been introduced in Ghana, Senegal, Mali, Morocco, Sierra can have positive benefits in terms of delaying pregnancy. Leone, and elsewhere.109-113 Evaluations do generally find In Kenya, for example, the provision of free uniforms to increased utilisation among targeted populations, although students was found to decrease the likelihood of dropping many of these studies lacked strong control groups, and a out of school and reduced the probability of being pregnant recent review found that their overall quality is too poor to by 17%, and these effects persisted for years after the subsidy draw firm conclusions.114 The poor implementation of these programme ended.105 In Malawi, a programme that provided policies,115and a lack of accountability to the beneficiaries,116 cash transfers to current schoolgirls and recent dropouts has likely limited the effectiveness of these approaches. conditional upon staying in school or returning to school reduced early marriage, teenage pregnancy, Vouchers to encourage the use of maternal health services and self-reported sexual activity after just one year of in Bangladesh, Cambodia Kenya,Pakistan, Tanzania and programme implementation.91 Similarly in Chile and Peru Uganda have been evaluated more rigorously, and with increased school attendance and extended school-hours more encouraging results.117 In Bangladesh, women who programmes respectively were found to reduce teen gave birth in areas targeted by the vouchers were more likely pregnancy.103 Meanwhile evidence from Bogota, Colombia to use an accredited provider and were 13 percentage points suggests that performance-based incentives are critical to more likelyto give birth in a facility than those in compar- the success of a given CCT programme in reducing teen ison areas.118 In rural Kenya, women living in voucher areas pregnancy.106 The study finds that the Subsidio educativo were more likely to give birth in health facilities and with programme, for which renewal is conditioned on the skilled health care providers,119and similar results were beneficiary girl being enrolled in the following year, found in the urban slums of Nairobi.120 causes a sizeable reduction in teenage pregnancy, while the Familias en Accion, which does not contain a perfor- Another approach involves cash payments to women mance condition, has no effect. conditional upon their use of maternal services. A systematic review has shown that CCTs can effectively improve health outcomes, including maternal health.121 The most rigorously Efforts within the health system evaluated CCT programme, known as the Progresa/ Both history and contemporary patterns of maternal Oportunidadeds programme in Mexico, targeted the health suggest that improvements within the health sector health-seeking behaviour of poor households.122 Women can translate into improvements in reproductive health.107 exposed to the programme, especially younger mothers, Major reductions in maternal mortality in what are now were more likely to both access antenatal care and select developed countries did not generally occur until the late more skilled health care providers for delivery.123,124 India 1930s to the end of the 1960s, primarily as a result of new has also recently introduced the Janani Suraksha Yojania Data suggests that the incidence of child marriage has fallen over time, however this has taken place relatively slowly, and the drivers of those reductions are not well known. 68 19 Investing in women’s reproductive health programme, a CCT to promote the use of maternal health concealable forms of contraception was thought to be a services through financial incentives for women to deliver reason why a randomised intervention to increase the in health care facilities. A recent evaluation found that uptake of family planning programmes failed to increase the programme increased rates of institutional delivery,125 utilisation.136 Recent work in Peru found that where however, the quality of care delivered has been questioned machismo, or strong male stereotyping, is present, reaching and the impact of the programme on mortality rates is not out to men may be a prerequisite for programmes targeted yet known.126,127Nonetheless, if well implemented, as in the at women.137 Mexican case, CCT programmes can be an effective route to increase the utilisation of maternal health services. At the same time, as underlined by Figure 1, it is unlikely that any single intervention is sufficient to make significant Some countries have also experimented with policies improvements in reproductive health outcomes. In Sri that target the supply side—that is, providers. For example, Lanka, where rapid maternal mortality declines were result-based financing schemes are a relatively recent documented from the 1940s tothe 1960s, numerous health innovation that has been tried in several countries. In system improvements were implemented (see Box 1).44 Rwanda, small incentives were paid to providers conditional Similar strategies have been adopted in Morocco, where on ensuring that their patients received prescribed maternal maternal mortality has also dropped rapidly. More recently, and child health services.128After 23 months, the study found the Saving Mothers, Giving Life programme has piloted a a 23% increase in institutional deliveries and increases in ‘whole of health systems’ approach to dramatically reduce the quality of prenatal care. In Zimbabwe, a similar scheme maternal mortality in eight districts in Zambia and Uganda. provides subsidies to rural health clinics and hospitals based While the official evaluations are still underway, preliminary on their performance in delivering a package of free health evidence suggests that the programme has influenced key services to pregnant women and children under five years of indicators such as institutional delivery rates. age. While the programme is still in its infancy, initial results are promising: the number of women who had four or more prenatal visits increased by 65% from a year before.129 Improving accountability for reproductive health outcomes Efforts to scale-up the availability of human resources, Increasing accountability to patients can help improve in particular less qualified health workers who can be health service delivery and health outcomes.138 Options more rapidly trained and more easily deployed into rural include creating community participation mechanisms,139 communities, also hold promise but require more evalua- enhancing the quality of health information for consumers, tion. Ethiopia, a large and mostly rural country with one of establishing community groups to empower consumers to the lowest ratios of physicians per population in the world, take action, and including non-governmental organisations undertook a massive expansion of health human resources, (NGOs) to expand access to care.138 Information, dialogue targeting improvements in maternal and child health. and negotiation have been widely identified as important Some studies have shown that the programme has been elements that enable accountability mechanisms to address well received by women and suggested that the use of family problems and to foster better service provision, in particular planning and antenatal care have increased.130,131 However, in the context of reproductive health services. But power other evaluations found the impact of the programme on relations, social contexts and the policy and service delivery maternal health indicators to be limited.132 Evaluations systems within which they are applied must be taken into suggest that in rural Pakistan, the Lady Health Worker— account in determining how any accountability mechanisms local female residents who have receive a short pre-service should be applied.140,141 and in-service medical training to deliver a range of health services from immunisations to family planning, to provide Evidence from Uganda suggests that introduction of a basic health education and to identify and refer more local accountability mechanism, which included efforts complicated cases—programme led to increases nationally to stimulate beneficiary control alongside provision of in the use of contraceptives.133 information about staff performance, resulted in 21% fewer births over the period 2006–2009 and a drop in the rate Survey evidence suggests that a large unmet need for family of stillbirths by almost one percentage point. By contrast planning persists, suggesting the scope for the expansion when the community lacked information about perfor- of these programmes. In sub-Saharan Africa, a quarter of mance, the same type of intervention had no measurable women report an unmet need for family planning, and this impact on the quality or uptake of medical care.142 In Peru, has remained relatively constant since the early 1990s.134 a citizen surveillance programme increased the number of However, studies also underline the importance of under- births in health facilities by almost one-third over one year, standing household preferences and constraints on female and dramatically increased access to culturally appropriate autonomy in the design of family planning programmes. delivery options. The programme centred on training of A randomised controlled trial in Zambia found that a indigenous women to visit health centres, hospitals and programme was effective at reducing unwanted pregnancy pharmacies to monitor the quality and acceptability of when women were able to access family planning without care provided. Through the reports that were shared and their husbands present.135 In Ethiopia, a lack of high quality, 20 discussed with health providers, the process promoted appropriately in the short term in the absence of direct accountability of the local health authorities and increased representation, but continuation may be threatened if awareness of rights and responsibilities between health those who need and use the services are not directly service providers and users.143 Box 2 reviews the additional involved. This was demonstrated in Brazil, where the reforms implemented to strengthen accountability in Peru. incoming Mayor of Santa Barbara d-Oeste saw no benefit in continuing a vasectomy programme introduced by the Recent work from Orissa, India identifies three processes former Health Secretary. Community representatives on that underpin the effectiveness of social accountability: the executive committee of the project strongly defended generating demand for rights and better services; leveraging the programme and its services, and ultimately changed intermediaries to legitimise the demands of poor and the Mayor’s mind.145 marginalised women; and sensitising leaders and health providers to the needs of women. In this context, social Social accountability mechanisms can also empower accountability mechanisms such as public hearings provided marginalised groups when they are encouraged as active new ways for women to collectively voice their concerns and participants in the accountability mechanisms. This was demands in a supportive environment. These demands are the case among Sangha women in Andhra Pradesh, India, then reinforced and legitimised by intermediary partners who were given training on gender, health and social action such as local elected officials and the media, who also as part of a pilot project to make the health system more participate in the hearings, leading to increased receptivity accountable by improving interactions between providers to women’s needs.144 Evidence suggests that some account- and lower caste women.141 ability mechanisms may address beneficiaries’ needs How Peru improved reproductive Health Outcomes through Expanding Entitlements and Strengthening Public Sector Management146 box 2 By the early 2000s, although an user fees and entitlements to specific to improve institutional birth rates. A improvement from the previous services were provided conditional on performance-based budgeting process decades, the rates of perinatal and particular health conditions. The was put in place that was contingent maternal mortality in Peru were high proportion of births covered by SIS on meeting certain institutional birth by regional standards and institutional increased dramatically. The Ministry targets across districts. delivery rates had plateaued at a of Health also attempted to address Although these efforts have not been national average of only about 70% non-financial barriers to institutional rigorously evaluated, and identifying but with less than half of all births in deliveries by developing and which particular interventions were rural areas taking place in a clinic or promoting culturally–informed birth effective would be challenging, there hospital. As part of a broader set of practices in government facilities, have been significant increases in reforms aimed at improving health which included allowing for more institutional birth rates across the services throughout the country, the culturally sensitive birth practices and country, particularly among poor rural government specifically targeted passing a law ensuring that health women. There was a more than 50% institutional delivery rates as the key workers respect cultural diversity. On increase in institutional delivery rates indicator of progress towards the supply side, facilities and human among women in the bottom income improving maternal and perinatal resources were strengthened to quintile from 2005 to 2009 (32% to health. The government undertook provide more and higher quality health 48% of all births). This case study a series of reforms that targeted the services, in particular in previously demonstrates that increased rates of entitlements of women to health underserved regions. There were also institutional deliveries through policy services and strengthened public efforts to monitor and improve the reforms in challenging contexts are sector management. quality of health services delivered at feasible but that multi-pronged government facilities notably through In 2002, through the creation of the solutions that address overall the creation of a new standards Seguro Integral de Salud (SIS), pooled financing, demand-and supply-side process. Public sector management funding was provided to allow poor factors are likely to be essential. reforms were also undertaken in order households to be exempt from health 21 Investing in women’s reproductive health Improving the reproductive health of women around the world is vitally important not just for the health benefits that will ensue but also for the substantial social and economic benefits, for women, their families and their communities. 22 Conclusions 6 Improving the reproductive health of women around the world Here, as elsewhere, the lack of data constrains our under- is vitally important not just for the health benefits that standing, and possibilities for holding ourselves to account will ensue but also for the substantial social and economic for results. Because only a third of countries have complete benefits, for women, their families, and their communities. civil registration systems, our ability to understand the To date, most discussions about improving reproductive burden of disease associated with maternal mortality is health outcomes have focused on making improvements limited.4 Without such systems, too much energy is spent from within the health sector. Our analysis suggests that on the estimation, and debating estimates of mortality, than measures of gender inequality are important predictors of actually addressing the problems and monitoring progress.87 poor reproductive health and therefore addressing these Recent global commitments to redress this situation— sources of inequality is critical to improve reproductive including the Interagency and Expert Group on the health outcomes. Proven policy options are available to Development of Gender Statistics147and the high-level realise those gains, although what works and doesn’t work Commission on Information and Accountability for will be very much driven by the country’s institutional, Women’s and Children’s Health recommendations—are political and cultural context. encouraging, and need to be followed through with efforts on the ground. Context is important, not only because of how this shapes the technical and financial possibilities of the interventions, The other broad point that emerges is that while evaluations but because social norms and attitudes are so important. have shown that many reproductive health policies are More broadly, the recognition of agency underlines the potentially effective, efforts can be plagued by poor imple- need to engage both women and men in reproductive mentation. Proper monitoring and evaluation are critical to health interventions. ensure performance and ongoing support. This in turn requires major investments in statistical capacity building, open access to data and utilisation of evidence to improve accountability and performance. 23 Investing in women’s reproductive health 24 References 1 The World Bank. World Development Report 2012. Washington, DC, 15 Jack W. The Promise of Health: Evidence of the Impact of Health on Income and The World Bank, 2012. Well-Being. In: Glied S, Smith PC, eds. The Oxford Handbook of Health Economics. Oxford University Press, 2010. Lozano R, Wang H, Foreman KJ et al. Progress towards Millennium Development 2  Goals 4 and 5 on maternal and child mortality: an updated systematic analysis. 16 Say L, Raine R. A systematic review of inequalities in the use of maternal health The Lancet 2011; 378: 1139–65. care in developing countries: examining the scale of the problem and the importance of context. Bulletin of the World Health Organization 2007; 85: 812–9. Lozano R, Naghavi M, Foreman K, et al. Global and regional mortality from 235 3  causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the 17 Powell-Jackson T, Hoque ME. Economic consequences of maternal illness in rural Global Burden of Disease Study 2010. The Lancet 2012; 380: 2095–128. Bangladesh. Health Economics 2012; 21: 796–810. 4 The World Health Organization, UNICEF, The World Bank, UNFPA. Trends in 18 Ahn N, Mira P. A note on the changing relationship between fertility and female Maternal Mortality: 1990 to 2010. Geneva, World Health Organization, 2012. employment rates in developed countries. Journal of Population Economics 2002; 15: 667–82. 5 Mahler H. The Safe Motherhood Initiative: A Call to Action. The Lancet 1987; 329: 668. 19 The World Bank. The effect of women’s economic power in Latin America and the Caribbean. Latin America and Caribbean Poverty and Labor Brief, 2012 6 Koblinsky MA. Beyond maternal mortality — magnitude, interrelationship and http://ezproxy.library.nyu.edu:5447/external/default/WDSContentServer/IW3P/IB/2 consequences of women’s health, pregnancy-related complications and 012/11/30/000386194_20121130024624/Rendered/PDF/NonAsciiFileName0.pdf. nutritional status on pregnancy outcomes. International Journal of Gynecology and Obstetrics 1995; 48: S21–S32. Bloom DE, Canning D, Fink G, Finlay JE. Fertility, female labor force participation, 20  and the demographic dividend. Journal of Population Economics 2009; 14: 79–101. 7 Campbell OMR, Graham WJ, Lancet Maternal Survival Series steering group. Strategies for reducing maternal mortality: getting on with what works. 21 National Statistics Directorate, Ministry of Finance, ICF Macro. Timor-Leste The Lancet 2006; 368: 1284–99. Demographic and Health Survey, 2009-10. Dili, Timor-Leste, National Statistics Directorate and ICF Macro, 2010. 8 Freedman LP, Graham WJ, Brazier E, et al. Practical lessons from global safe motherhood initiatives: time for a new focus on implementation. The Lancet 22 Angrist J, Evans W. Children and Their Parents’ Labor Supply: Evidence 2007; 370: 1383–91. from Exogenous Variation in Family Size. The American Economic Review 1998; 88: 450–77. 9 Ronsmans C, Graham W. Maternal mortality: who, when, where, and why. The Lancet 2006; 368: 1189–200. 23 Cruces G, Galiani S. Fertility and female labor supply in Latin America: New causal evidence. Labour Economics 2007; 14: 565–73. 10 Darmstadt GL, Bhutta ZA, Cousens S, et al. Evidence-based, cost-effective interventions: how many newborn babies can we save? The Lancet 2005; 365: 24 World Health Organization. Adolescent pregnancy. Geneva, World 977–88. Health Organization, 2012. 2012.http://www.who.int/mediacentre/factsheets/ fs364/en/index.html. 11 Cleland J, Conde-Agudelo A, Peterson H, Ross J, Tsui A. Contraception and health. The Lancet 2012; 380: 149–56. Field E, Ambrus A. Early Marriage, Age of Menarche, and Female Schooling 25  Attainment in Bangladesh. Journal of Political Economy 2008; 116: 881–930. 12 Jack W, Lewis M. Health Investments and Economic Growth: macroeconomic evidence and microeconomic foundations. In: Spence M, Lewis M, eds. Health Miller G. Contraception as Development? New Evidence from Family Planning 26  and Growth. Washington, DC, World Bank Publications, 2009: 1–40. in Colombia. The Economic Journal 2010; 120: 709–36. 13 FAO, IFAD, The World Bank. Gender in Agriculture Sourcebook. Washington, DC, 27 Azevedo JP, Favara M, Haddock SE, Lopez-Calva LF, Müller M, Perova E. World Bank Publicationshttp://go.worldbank.org/5Z9QPCC7L0. Teenage Pregnancy and Opportunities in Latin America and the Caribbean. The World Bank, 2012. 14 Gertler P, Gruber J. Insuring Consumption against Illness. The American Economic Review 2002; 92: 51–70. Jensen R. Do Labor Market Opportunities Affect Young Women’s Work and 28  Family Decisions? Experimental Evidence from India. The Quarterly Journal of Economics 2012. 25 Investing in women’s reproductive health 29 Wagstaff A. The economic consequences of health shocks: evidence from 51 Becker G. An economic analysis of fertility. Demographic and economic change in Vietnam. J Health Econ 2007; 26: 82–100. developed countries, 1960. 30 Gertler P, Gruber J. Insuring Consumption against Illness. The American 52 Buckles KS, Munnich EL. Birth Spacing and Sibling Outcomes. Journal of Human Economic Review 2002; 92: 51–70. Resources 2012; 47: 613–42. 31 Kruk ME, Goldmann E, Galea S. Borrowing and selling to pay for health care in 53 Kirigia JM, Oluwole D, Mwabu GM, Gatwiri D, Kainyu LH. Effects of maternal low- and middle-income countries. Health Aff (Millwood) 2009; 28: 1056–66. mortality on gross domestic product (GDP) in the WHO African region. African Journal of Health Sciences 2006; 13: 86–95. 32 Xu K, Evans DB, Kawabata K, Zeramdini R, Klavus J, Murray CJL. Household catastrophic health expenditure: a multicountry analysis. The Lancet 2003; 54 Lee R. The demographic transition: three centuries of fundamental change. 362: 111–7. Journal of Economic Perspectives 2003; 17: 167–90. 33 Powell-Jackson T, Hoque ME. Economic consequences of maternal illness in 55 Bloom D, Canning D. Contraception and the Celtic tiger. Economic and Social rural Bangladesh. Health Econ 2012; 21: 796–810. Review 2003; 34: 229–48. 34 Borghi J, Hanson K, Acquah CA, et al. Costs of near-miss obstetric complications 56 Bloom DE, Finlay JE. Demographic Change and Economic Growth in Asia. Asian for women and their families in Benin and Ghana. Health Policy Plan 2003; 18: Economic Policy Review 2009; 4: 45–64. 383–90. 57 The Center for Global Development. Chapter 6: Saving Mothers’ Lives in Sri Lanka. 35 Storeng KT, Baggaley RF, Ganaba R, Ouattara F, Akoum MS, Filippi V. Paying the In: Case Studies in Global Health. Jones & Bartlett Publishers, 2007. price: The cost and consequences of emergency obstetric care in Burkina Faso. Social Science and Medicine 2008; 66: 545–57. 58 Pathmanathan I, Liljestrand J. Investing in Maternal Health. World Bank Publications, 2003. 36 DeJong J, Bahubaishi N, Attal B. Effects of reproductive morbidity on women’s lives and costs of accessing treatment in Yemen. Reprod Health Matters 2012; 20: 59 Shiffman J. Can Poor Countries Surmount High Maternal Mortality? 129–38. Studies in Family Planning 2000; 31: 274–89. 37 Afsana K. The Tremendous Cost of Seeking Hospital Obstetric Care in Bangladesh. 60 Scott S, Ronsmans C. The relationship between birth with a health professional Reprod Health Matters 2004; 12: 171–80. and maternal mortality in observational studies: a review of the literature. Tropical Medicine and International Health 2009; 14: 1523–33. 38 Ye F, Wang H, Huntington D, et al. The Immediate Economic Impact of Maternal Deaths on Rural Chinese Households. PLoS ONE 2012; 7: e38467. 61 Buor D, Bream K. An Analysis of the Determinants of Maternal Mortality in Sub-Saharan Africa. Journal of Women’s Health 2004; 13: 926–38. 39 Ronsmans C, Chowdhury ME, Dasgupta SK, Ahmed A, Koblinsky M. Effect of parent’s death on child survival in rural Bangladesh: a cohort study. The Lancet 62 Bulatao RA, Ross JA. Which health services reduce maternal mortality? 2010; 375: 2024–31. Evidence from ratings of maternal health services. Tropical Medicine and International Health 2003; 8: 710–21. 40 Beegle K, Filmer D, Stokes A. Orphanhood and the living arrangements of children in sub-saharan Africa. World Development 2010. 63 Sen A. Development as capability expansion. In: Griffin KB, Knight JB, eds. Human development and the international development strategy for the 1990s. 1990. 41 Case A, Ardington C. The Impact of Parental Death on School Outcomes: Longitudinal Evidence from South Africa. Demography 2006; 43: 401–20. 64 Sen AK. Reason before identity. Oxford University Press, USA, 1999. 42 Evans DK, Miguel E. Orphans and Schooling in Africa: A Longitudinal Analysis. 65 Author’s calculations, computed from data taken from Demography 2007; 44: 35–57. http://www.measuredhs.com/Data. 43 Beegle K, De Weerdt J, Dercon S. The intergenerational impact of the African 66 Haque SE, Rahman M, Mostofa MG, Zahan MS. Reproductive Health Care orphans crisis: a cohort study from an HIV/AIDS affected area. International Utilisation among Young Mothers in Bangladesh: Does Autonomy Matter? Journal of Epidemiology 2009; 38: 561–8. Women’s Health Issues 2012; 22: e171–80. 44 Jayachandran S, Lleras-Muney A. Life Expectancy and Human Capital 67 World Health Organization. Early marriages, adolescent and young pregnancies. Investments: Evidence from Maternal Mortality Declines. The Quarterly Journal 2012. of Economics 2009; 124: 349–97. 68 Nguyen MC, Wodon Q. Global Trends in Child Marriage. World Bank Working 45 Almond D, Currie J. Killing Me Softly: The Fetal Origins Hypothesis. Journal Paper, 2012. of Economic Perspectives 2011; 25: 153–72. 69 Raj A, Saggurti N, Balaiah D, Silverman JG. Prevalence of child marriage and its 46 Ravelli AC, van der Meulen JH, Michels RP, et al. Glucose tolerance in adults effect on fertility and fertility-control outcomes of young women in India: a after prenatal exposure to famine. The Lancet 1998; 351: 173–7. cross-sectional, observational study. The Lancet 2009; 373: 1883–9. 47 van Ewijk R. Long-term health effects on the next generation of Ramadan 70 Barber SL, Bertozzi SM, Gertler PJ. Variations in prenatal care quality for the fasting during pregnancy. Journal of Health Economics 2011; 30: 1246–60. rural poor in Mexico. Health Aff (Millwood) 2007; 26: w310–23. 48 Almond D, Mazumder B. Health Capital and the Prenatal Environment: The Effect 71 Rani M, Bonu S, Harvey S. Differentials in the quality of antenatal care in India. of Ramadan Observance During Pregnancy. American Economic Journal: Applied International Journal for Quality in Health Care 2008; 20: 62–71. Economics 2011; 3: 56–85. 72 Victora CG, Matijasevich A, Silveira M, Santos I, Barros AJD, Barros FC. 49 Chen Y, Zhou L-A. The long-term health and economic consequences of the Socio-economic and ethnic group inequities in antenatal care quality in the 1959-1961 famine in China. Journal of Health Economics 2007; 26: 659–81. public and private sector in Brazil. Health Policy and Planning 2010; 25: 253–61. 50 Canning D, Razzaque A, Driessen J, Walker DG, Streatfield PK, Yunus M. The effect 73 Fotso JC, Mukiira C. Perceived quality of and access to care among poor urban of maternal tetanus immunization on children’s schooling attainment in Matlab, women in Kenya and their utilisation of delivery care: harnessing the potential Bangladesh: follow-up of a randomised trial. Social Science and Medicine 2011; 72: of private clinics? Health Policy and Planning 2012; 27: 505–15. 1429–36. 26 74 L Nikiema YKGCBSYM-P. Quality of Antenatal Care and Obstetrical Coverage in 94 Heath R, Mobarak AM. Does Demand or Supply Constrain Investments in Rural Burkina Faso. Journal of Health, Population and Nutrition 2010; 28: 67. Education? Evidence from Garment Sector Jobs in Bangladesh. Background paper for WDR2012. http://faculty.washington.edu/rmheath/garments%20  eproductive Health Country Profiles. The World Bank, http://go.worldbank. 75 R Heath%20Mobarak.pdf (accessed 2 Jan2013). org/6DZC2ITCQ0. 95 Roy S. Empowering Women: Inheritance Rights and Female Education in India. 76 Adamson PC, Krupp K, Niranjankumar B, Freeman AH, Khan M, Madhivanan P. Are Background paper for WDR2012; : 1–47. marginalized women being left behind? A population-based study of institutional deliveries in Karnataka, India. BMC Public Health 2012; 12: 30. 96 Cook R. International Human Rights and Women’s Reproductive Health. Studies in Family Planning 1993; 24: 73–86. 77 Wang C. Trends in contraceptive use and determinants of choice in China: 1980-2010. Contraception 2012; 85: 570–9. 97 Pacagnella RC, Cecatti JG, Osis MJ, Souza JP. The role of delays in severe maternal morbidity and mortality: expanding the conceptual framework. 78 Adegoke AA, van den Broek N. Skilled birth attendance-lessons learnt. BJOG Reproductive Health Matters 2012; 20: 155–63. 2009; 116 Suppl 1: 33–40. 98 World Health Organization. Unsafe Abortion. Geneva, Switzerland, 2011. 79 Gabrysch S, Simushi V, Campbell OMR. Availability and distribution of, and geographic access to emergency obstetric care in Zambia. International Journal 99 Berer M. National Laws and Unsafe Abortion: The Parameters of Change. of Gynecology and Obstetrics 2011; 114: 174–9. Reproductive Health Matters 2004; 12: 1–8. 80 Kesterton AJ, Cleland J, Sloggett A, Ronsmans C. Institutional delivery in rural 100 Sedgh G, Singh S, Shah IH, Åhman E, Henshaw SK, Bankole A. Induced abortion: India: the relative importance of accessibility and economic status. BMC incidence and trends worldwide from 1995 to 2008. The Lancet 2012; 379: 625–32. Pregnancy Childbirth 2010; 10: 30. 101 Benson J, Andersen K, Samandari G. Reductions in abortion-related mortality 81 Titaley CR, Hunter CL, Heywood P, Dibley MJ. Why don‘t some women attend following policy reform: evidence from Romania, South Africa and Bangladesh. antenatal and postnatal care services?: a qualitative study of community Reproductive Health Matters 2011; 8: 39. members’ perspectives in Garut, Sukabumi and Ciamis districts of West Java Province, Indonesia. BMC Pregnancy Childbirth 2010; 10: 61. 102 Jewkes R, Rees H. Dramatic decline in abortion mortality due to the Choice on Termination of Pregnancy Act. South African Medical Journal 2005; 95: 250. 82 Borghi J, Ensor T, Neupane BD, Tiwari S. Financial implications of skilled attendance at delivery in Nepal. Tropical Medicine and International Health 2006; 103 Berthelon ME, Kruger DI. Risky behavior among youth: Incapacitation effects 11: 228–37. of school on adolescent motherhood and crime in Chile. Journal of Public Economics 2011; 95: 41–53. 83 Anand S, Bärnighausen T. Human resources and health outcomes: cross-country econometric study. The Lancet 2004; 364: 1603–9. 104 McQueston K, Silverman R, Glassman A. Adolescent Fertility in Low- and Middle-Income Countries: Effects and Solutions. Center for Global Development 84 Das J, Holla A, Das V, Mohanan M, Tabak D, Chan B. In urban and rural India, a Working Paper. 2012. standardized patient study showed low levels of provider training and huge http://www.cgdev.org/content/publications/detail/1426175 (accessed 24 Jan2013). quality gaps. Health Affairs 2012; 31: 2774–84. 105 Duflo E, Dupas P, Kremer M. Education, HIV, and Early Fertility: Experimental 85 Chaudhury N, Hammer J, Kremer M, Muralidharan K, Rogers FH. Missing in action: Evidence from Kenya. Working Paper. teacher and health worker absence in developing countries. Journal of Economic Perspectives 2006; 20: 91–116. 106 Cortés D, Gallego JM, Maldonado D. On the design of education conditional cash transfer programs and non education outcomes: The case of teenage pregnancy. 86 iERG. Every Woman, Every Child: from commitments to action. Geneva, CESIFO Working Paper, 2011. Switzerland, World Health Organization, 2012. 107 De Brouwere V, De Brouwere V, Tonglet R, Van Lerberghe W. Strategies for 87 Labrique AB, Pereira S, Christian P, Murthy N, Bartlett L, Mehl G. Pregnancy reducing maternal mortality in developing countries: what can we learn from registration systems can enhance health systems, increase accountability and the history of the industrialized West? Tropical Medicine and International reduce mortality. Reproductive Health Matters 2012; 20: 113–7. Health 1998; 3: 771–82. 88 Currie J, Moretti E. Mother’s Education and the Intergenerational Transmission 108 Jayachandran S, Lleras-Muney A, Smith KV. Modern Medicine and the of Human Capital: Evidence from College Openings. The Quarterly Journal of Twentieth Century Decline in Mortality: Evidence on the Impact of Sulfa Drugs. Economics 2003; 118: 1495–532. American Economic Journal: Applied Economics 2010; 2: 118–46. 89 Breierova L, Duflo E. The impact of education on fertility and child mortality: 109 Witter S, Arhinful DK, Kusi A, Zakariah-Akoto S. The experience of Ghana in Do fathers really matter less than mothers? NBER Working Paper 2004. implementing a user fee exemption policy to provide free delivery care. Reprod Health Matters 2007; 15: 61–71. 90 Osili U. Does female schooling reduce fertility? Evidence from Nigeria. Journal of Development Economics 2008. 110 Bennis I, De Brouwere V. Fee exemption for caesarean section in Morocco. Arch Public Health 2012; 70: 3. 91 Baird S, Chirwa E, McIntosh C, Ozler B. The short-term impacts of a schooling conditional cash transfer program on the sexual behavior of young women. 111 Witter S, Dieng T, Mbengue D, Moreira I, De Brouwere V. The national free Health Economics 2010; 19 Suppl: 55–68. delivery and caesarean policy in Senegal: evaluating process and outcomes. Health Policy Plan 2010; 25: 384–92. 92 Duflo E. Women Empowerment and Economic Development. Journal of Economic Literature 2012; 50: 1051–79. 112 Ponsar F, Van Herp M, Zachariah R, Gerard S, Philips M, Jouquet G. Abolishing user fees for children and pregnant women trebled uptake of malaria-related 93 Hossain N. Exports, Equity, and Empowerment: the effects of readymade interventions in Kangaba, Mali. Health Policy Plan 2011; 26: ii72–ii83. garments manufacturing employment on gender equality in Bangladesh. Background paper for WDR2012. http://siteresources.worldbank.org/INTWDR2012/ 113 Wakabi W. Mothers and infants to get free health care in Sierra Leone. Resources/7778105-1299699968583/7786210-1322671773271/Hossain-Export- The Lancet. 2010; 375: 882. Equity-employment.pdf (accessed 2 Jan2013). 27 Investing in women’s reproductive health 114 Dzakpasu S, Powell-Jackson T, Campbell OMR. Impact of user fees on maternal 133 Douthwaite M, Ward P. Increasing contraceptive use in rural Pakistan: an health service utilisation and related health outcomes: a systematic review. evaluation of the Lady Health Worker Programme. Health Policy and Planning Health Policy Plan. 2013 Jan 30. 2005; 20: 117–23. 115 Ridde V, Morestin F. A scoping review of the literature on the abolition of user 134 Ezeh AC, Bongaarts J, Mberu B. Global population trends and policy options. fees in health care services in Africa. Health Policy Plan 2011; 26: 1–11. The Lancet 2012; 380: 142–8. 116 McPake B, Brikci N, Cometto G, Schmidt A, Araujo E. Removing user fees: 135 Ashraf N, Field E, Lee J. Household Bargaining and Excess Fertility: An learning from international experience to support the process. Health Policy Experimental Study in Zambia. econ.ucsb.edu. Plan 2011; 26: ii104–17. 136 Desai J, Tarozzi A. Microcredit, family planning programs, and contraceptive 117 Bellows NM, Bellows BW, Warren C. Systematic Review: the use of vouchers behavior: evidence from a field experiment in Ethiopia. Demography 2011; for reproductive health services in developing countries: systematic review. 48: 749–82. Trop Med Int Health 2011; 16: 84–96. 137 The World Bank. The Third Asset: Improving Development Outcomes through 118 Nguyen HTH, Hatt L, Islam M, et al. Encouraging maternal health service Agency. 2012. utilisation: An evaluation of the Bangladesh voucher program. Social Science and Medicine 2012; 74: 989–96. 138 Berlan D, Shiffman J. Holding health providers in developing countries accountable to consumers: a synthesis of relevant scholarship. Health Policy 119 Obare F, Warren C, Njuki R, et al. Community-level impact of the reproductive and Planning 2012; 27: 271–80. health vouchers programme on service utilisation in Kenya. Health Policy Plan 2012. doi:10.1093/heapol/czs033. 139 Björkman M, Svensson J. Power to the People: Evidence from a Randomised Field Experiment on Community-Based Monitoring in Uganda*. Quarterly Journal of 120 Bellows B, Kyobutungi C, Mutua MK, Warren C, Ezeh A. Increase in facility-based Economics 2009. deliveries associated with a maternal health voucher programme in informal settlements in Nairobi, Kenya. Health Policy Plan 2012. doi:10.1093/heapol/czs030. 140 Murthy RK, Klugman B. Service accountability and community participation in the context of health sector reforms in Asia: implications for sexual and reproductive 121 Lagarde M, Haines A, Palmer N. Conditional cash transfers for improving uptake health services. Health Policy and Planning 2004; 19 Suppl 1: i78–i86. of health interventions in low- and middle-income countries: a systematic review. JAMA 2007; 298: 1900–10. 141 George A. Using Accountability to Improve Reproductive Health Care. Reproductive Health Matters 2003; 11: 161–70. 122 Gertler P. Do Conditional Cash Transfers Improve Child Health? Evidence from PROGRESA’s Control Randomised Experiment. The American Economic Review 142 Bjorkman-Nyqvist M, de Walque D, Svensson J. Information is power: 2004; 94: 336–41. experimental evidence of the long run impact of community based monitoring. Working Paper. 123 Barber SL, Gertler PJ. Empowering women to obtain high quality care: evidence from an evaluation of Mexico’s conditional cash transfer programme. 143 CARE Policy Brief. https://docs.google.com (accessed 7 Feb2013). Health Policy Plan 2009; 24: 18–25. 144 Papp SA, Gogoi A, Campbell C. Improving maternal health through social 124 Sosa-Rubí SG, Walker D, Serván E, Bautista-Arredondo S. Learning effect of a accountability: A case study from Orissa, India. Global Public Health 2012. doi:10.1 conditional cash transfer programme on poor rural women’s selection of 080/17441692.2012.748085. delivery care in Mexico. Health Policy Plan 2011; 26: 496–507. 145 Díaz M, Simmons R, Díaz J, et al. Action research to enhance reproductive choice 125 Lim SS, Dandona L, Hoisington JA, James SL, Hogan MC, Gakidou E. India’s Janani in a Brazilian municipality: the Santa Barbara Project. Responding to Cairo: Suraksha Yojana, a conditional cash transfer programme to increase births in case-studies of changing practice in reproductive health and family planning health facilities: an impact evaluation. The Lancet 2010; 375: 2009–23. New York, The Population Council 2002; : 355–75. 126 Das A, Rao D, Hagopian A. India’s Janani Suraksha Yojana: further review needed. 146 The World Bank. Peru - Recurso Programmatic AAA – Phase IV : Improving Health The Lancet 2011; 377: 295–6. Outcomes by Strengthening Users’ Entitlements and Reinforcing Public Sector Management. World Bank Report. 2011.https://openknowledge.worldbank.org/ 127 Lim SS, Dandona L, Hoisington JA, Hogan MC, Gakidou E. India‘s Janani Suraksha handle/10986/2739 (accessed 1 Feb2013). Yojana: further review needed – Authors’ reply. The Lancet 2011; 377: 296–7. 147 Gender Equality Data and Statistics. http://datatopics.worldbank.org/gender/ 128 Basinga P, Gertler PJ, Binagwaho A, Soucat AL, Sturdy J, Vermeersch CM. Effect on monitoring-progress (accessed 2 Jan2013). maternal and child health services in Rwanda of payment to primary health care providers for performance: an impact evaluation. The Lancet 2011; 377: 1421–8. n Rural Zimbabwe, No Fees, Better Care for Women and Children. worldbank.org. 129 I 2012.http://www.worldbank.org/en/news/2012/11/30/in-rural-zimbabwe-no-fees- better-care-for-women-children (accessed 2 Jan2013). 130 Koblinsky M, Tain F, Gaym A, Karim A, Carnell M, Tesfaye S. Responding to the maternal healthcare challenge: The Ethiopian Health Extension Program. Ethiopian Journal of Health Development 2010; 24. doi:10.4314/ejhd.v24i1.62951. 131 Medhanyie A, Spigt M, Kifle Y, et al. The role of health extension workers in improving utilisation of maternal health services in rural areas in Ethiopia: a cross sectional study. BMC Health Services Research 2012; 12: 352. 132 Admassie A, Abebaw D, Woldemichael AD. Impact evaluation of the Ethiopian Health Services Extension Programme. Journal of Development Effectiveness 2009; 1: 430–49. 28 Photo Credits Mark Tuschman Design Designlounge, Inc., NY Proofreader Inis Communication This publication is printed on the world’s most environmentally friendly coated paper: 9 Lives Silk. This paper is certified by the Forestry 55% Stewardship Council (FSC), an independent organisation which ensures all wood fiber is harvested using sustainable forestry management, in a socially responsible manner. 9 Lives Silk is made from 55% recycled materials (30% pre-consumer waste, 25% post-consumer waste), and 45% FSC certified virgin Elemental Chlorine-Free fiber. 9 Lives Silk is Green Label Singapore certified, and created at a paper mill meeting ISO 9001, ISO 14001 (EMAS) standards. The publication is manufactured at a printing plant meeting those same standards, using vegetable based inks, and energy efficient practices throughout the facility. Over 90% of the waste paper is captured for recycling during the manufacturing process. Women Deliver 588 Broadway, Suite 905 New York, NY 10012 Tel: +1.646.695.9100 Web: www.womendeliver.org For more information, please contact: info@womendeliver.org