Women: Rapid

RAPID
WOMEN
JULY 2012
Modeling the Impact of Family Planning and Other Women-Centered Interventions on Development
by Scott Moreland
A Technical Guide to the RAPID/Women Model
Photos (leftright) by: Frank Spangler, United States Agency for International Development
Suggested citation: Moreland, Scott. 2012. Modeling the Impact of Family Planning and Other WomenCentered Interventions on Development: A Technical Guide to the RAPID/Women Model. Washington, DC: Futures Group and the David and Lucile Packard Foundation.
CONTENTS
Acknowledgments .................................................................................................................. iii Abbreviations ...........................................................................................................................iv Introduction .............................................................................................................................. 1 Technical Approach of RAPID/Women.................................................................................. 3 Conceptual Framework ............................................................................................................................. 3 Model Structure ........................................................................................................................................ 4 Calculated Variables ................................................................................................................................. 7 Human Development Index (HDI) ........................................................................................................ 7 Female life expectancy at birth ............................................................................................................. 7 Contraceptive prevalence rate .............................................................................................................. 8 Percentage of women married/ in union ............................................................................................... 9 Total fertility rate ................................................................................................................................ 10 Under-five mortality, infant mortality, and maternal mortality .......................................................... 11 Gross domestic product per capita ..................................................................................................... 12 Labor force.......................................................................................................................................... 13 Investment (capital) ............................................................................................................................ 13 Sample Results from an Application in Mali ........................................................................ 14 Mali Baseline Situation ........................................................................................................................... 14 Scenarios ................................................................................................................................................. 14 Scenario Results ...................................................................................................................................... 15 Demographic indicators ..................................................................................................................... 15 Births at risk........................................................................................................................................ 16 Mortality ............................................................................................................................................. 16 Population effects................................................................................................................................ 18 Conclusions ............................................................................................................................................. 19 Using the RAPID/Women Computer Model ......................................................................... 20 Using the Model...................................................................................................................................... 21 Step 1: Establish initial values ............................................................................................................ 21 Step 2: Define policy scenarios........................................................................................................... 21 Step 3: Project the population ............................................................................................................ 22 Step 4: View the results ....................................................................................................................... 24 Defining Scenarios .................................................................................................................................. 25 Endnotes ................................................................................................................................. 28
ACKNOWLEDGMENTS
This activity benefited from the help and advice of a number of people and institutions. First, the USAID Office of Population and Reproductive Health funded related research and development of the RAPID/Women model through the USAID | Health Policy Initiative, Task Order 1 project. The David and Lucile Packard Foundation also contributed significant funding through a grant to help conduct a pilot test of the model in Mali. Further, Hannah Fortune-Greeley provided valuable assistance during model development by collecting data and running all of the statistical estimations. Mary Kincaid provided insightful guidance on the gender aspects of the model. Futures Groups project team in Mali, headed by Fofana Famory, was instrumental in facilitating the successful pilot test of the model. I would also like to thank many colleagues who participated in the expert group meetings along the way, especially Elizabeth Schoenecker of USAID and Sarah Clark of Futures Group. Finally, I would like to thank Cynthia Green and Elizabeth Leahy-Madsen for their assistance in helping with the technical editing and refinement of this guide.
iii
ABBREVIATIONS
AIDS CPR CS DHS FLE FP GDP GDPPC HDI HIV IMR Ln MDG MMR PPI RH TFR UN U5MR acquired immune deficiency syndrome contraceptive prevalence rate contraceptive security Demographic and Health Survey female life expectancy (at birth) family planning gross domestic product gross domestic product per capita Human Development Index human immunodeficiency virus infant mortality rate natural logarithm Millennium Development Goal maternal mortality ratio postpartum insusceptibility reproductive health total fertility rate United Nations under-five mortality rate
iv
INTRODUCTION
Currently, there is much interest in enhancing the roles and status of women in developing countries as well as revitalizing If we invest in the health of women programs for reproductive health and family planning. The U.S. and children, we ensure that families, communities, and nations thrive. government, as part of its Global Health Initiative, 1 is emphasizing women-centered programming 2 as a strategy for Melinda Gates, launch of the realizing development goals; the Global Fund to Fight AIDS, Reproductive Health Alliance Tuberculosis and Malaria has established a gender strategy for its September 22, 2010 3 programming; the World Bank launched a four-year gender 4 action plan in 2007; and other agencies such as the United Nations Women, United Nations Population Fund, Joint United Nations Program on HIV/AIDS, and others all emphasize the important role of improving the lives of women and girls, as both an objective in itself and as a strategy to achieve other development objectives such as the reduction of poverty or improvement in life expectancy. These donor efforts complement the work of the many international and national nongovernmental organizations that focus on supporting women-centered programs. Women-centered strategies concentrate on specific areas that improve the quality of life of women and girls. In the context it is used here, the focus is not on gender equity or equality outcomes per se, although these may be achieved as a result of the successful implementation of women-centered strategies. Women-centered strategies typically focus on the following: Improving womens and girls education Increasing their knowledge of health choices, especially in the area of reproductive health and family planning Increasing their decision-making autonomy and empowering women to have more control over resources at their disposal Increasing their economic power by increasing their incomes Reducing their exposure to various forms of violence, especially gender-based violence
While there is a growing body of research that shows the impact of women-centered strategies on specific development objectives like fertility, 5 child health, 6 or education, 7 the empirical evidence that links women-centered strategies to broader development outcomes such as economic growth is more limited and less well-known. The relative shortage of empirical evidence over time reflects both the difficulty of establishing direct causality between women-centered programming and development outcomes and the historical lack of high-quality, internationally comparable data on the relevant variables. 8 Over the last decade, however, research efforts 9 led by the World Bank and others have estimated the relationship between measures of gender equality (e.g., gender gaps in education and labor force participation rates) and measures of economic growth (e.g., gross domestic product, labor productivity, poverty reduction), building the case that increasing gender equality in developing countries can have a multiplier effect on overall development and growth. During the same timeframe, there has been an increase in the number of high-profile leaders and advocates that champion womens strategies as key to successful development or the eradication of poverty. Their statements often present the relationship as a given or point to a microlevel relationship (e.g., increasing womens education reduces child malnutrition) as evidence of the macro-level impact rather than directly referencing supporting data. One principal pathway through which women-centered strategies affect development outcomes, including economic growth, is by giving women agencythe ability to define goals and act on them, as defined in the 2007 World Development Report.10 Agency includes, importantly, agency over ones fertility. Family planning and reproductive health (FP/RH) programs provide women and their partners 1
Introduction
with the ability to choose if and when they want to get pregnant and the spacing of births. A large body of research has established the relationship between fertility and womens labor force participation, education, and health status: high fertility rates are associated with lower rates and quality of womens labor force participation, lower educational attainment, and lower health status of mothers and children. 11 FP/RH programs are an obvious and necessary intervention, with a multitude of benefits at the individual, household, community, and national levels. 12 They work in concert with other women-centered and general development strategies, including those focused on girls education, womens empowerment and political participation, livelihoods/income generation, and the legal status of women. A multidimensional program of interventions aimed at improving gender equality and raising the status of women is necessary to promote sustainable social and economic development. FP/RH programs are essential pieces of the package of interventions, as they give women the opportunity to engage in all the other dimensions. Therefore, in recognition of the critical importance of FP/RH programs, our modeling team set out to provide womens advocates with a user-friendly tool to more readily demonstrate the social and economic benefits of such programs in concert with other women-centered strategies. Our approach to exploring the relationships between women-centered strategies and family planning was to formulate and test a dynamic computer model, called RAPID/Women. Specifically, RAPID/Women focuses on the links between women-centered strategies, reproductive health, population, and economic development. It can help create an evidence base for womens rights advocates, governmental leaders, and womens organizations to use in advocating for increased investments in womencentered interventions and FP/RH programming in their countries.
TECHNICAL APPROACH OF RAPID/WOMEN

Conceptual Framework
The first challenge to make the links between family planning, women-centered strategies, and a measure of economic development was to create a conceptual framework to establish these empirical links. We developed a simple framework (Figure 1) that links intervention or program strategies to intermediate impacts on the quality of womens lives and then to national socioeconomic development.
Figure 1. Conceptual Framework of the RAPID/Women Model
Women-centered Strategies Elements of Womens Quality of Life Human Development Index (HDI)
Environmental and Natural Resource Microcredit/Income Generation Girls Education Programs Family Planning Programs Gender-based Violence Programs Political Participation and Leadership Other Health Interventions
Health
Gender-based Violence
Education
National Development
Decision Making
Income
On the left of the conceptual framework are a number of illustrative women-centered strategies. To keep the model simple and manageable, RAPID/Women only includes the three strategies (highlighted in orange and red). In the center of the conceptual framework are some key elements that affect the quality of womens lives. Lastly, on the right, we see the impact on national development as summarized by the UNs Human Development Index (HDI). In the following section, we provide more detail about how the conceptual framework is implemented in the model. Because of the models focus on women-centered strategies, the choice of variable to represent womens empowerment deserves special mention here. In RAPID/Women, womens empowerment is represented by the proxy variable, % of women reporting that wife beating is justified for certain reasons. This indicator is available in many Demographic and Health Surveys (DHS) and is one of the recommended indicators to measure community mobilization and individual behavior change. 13 A measure of womens attitudes toward intimate partner violence increasingly is included as a key domain in indices of womens autonomy, 14 along with decision-making power, mobility, and control over resources. 15 Further, women who report accepting attitudes toward intimate partner violence are more likely to have experienced or to expect such violence. 16 Empirically, we found that this indicator was negatively correlated with the 3
Technical Approach of RAPID/Women
percentage of women who report they can make their own health decisions (r=-0.38), meaning that those women who felt intimate partner violence was justified tended to not make their own health decisions. This result is in line with research showing a positive relationship between womens autonomy and health outcomes. 17 We also found below that this variable is highly correlated with reports of physical violence in countries where both indicators were collected (r= 0.7.). Intimate partner violence is challenging to measure accurately, due to stigma, underreporting, and differences in classification of types of abuse (variations in definition); and comparable data are not available in many countries.18 Even analyzing prevalence within one country over time can be difficult, as campaigns to raise awareness of gender-based violence can lead to a spike in prevalence rates as more incidents are reported, complicating analysis of trends. Given these various constraints, the prevalence of physical violence variable was not included in the model. Other measures of womens empowerment were considered 19 as well but were found to have various limitations for use in the modelmost often a lack of available data across a sufficient number of countries.
Model Structure
The model comprises two main sub-models: (1) the Demographic Sub-Model (Figure 2a), which includes variables that determine the growth and age structure of the population and (2) the Economic Development Sub-Model (Figure 2b), which calculates the HDI as the main development indicator. The Economic Development Sub-Model is linked to the Demographic Sub-Model through population variables, including the number of births (measured by the total fertility rate) and life expectancy. In Figure 2a, the yellow boxes represent the six policy variables that can be influenced by specific programs:
1. 2. 3. 4. 5. 6.
Contraceptive Security Access Index Womens empowerment/gender norms, measured by the percent of women saying wife beating can be justified Girls mean years of education completed Girls expected years of education Postpartum insusceptibility Sterility (infertility)
The user can change the values of these policy variables to simulate the effect that women-centered strategies would have. Changes in the policy variables affect other demographic and health variables in the model (shown in blue). For example, as girls obtain more education, they tend to marry later and use contraceptives, thus leading to fewer births. While the user inputs initial values for the policy variables, the future values are calculated directly by the model. The next section discusses each of the variables in Figures 2a and 2b, as well as the ways the variables are linked and the statistical relationships that were estimated.
Figure 2a. RAPID/Women Model: Demographic Sub-Model

Womens Empowerment/ Gender Norms Girls Mean Years of Education Completed Girls Expected Years of Education Completed
Contraceptive Security Access Index
Contraceptive Prevalence Rate
Marriage
Postpartum Insusceptibility
Sterility
Total Fertility Rate
Population
High-Risk Births
Under-5 Mortality
Life Expectancy
Figure 2b. RAPID/Women Model: Economic Development Sub-Model and HDI

Dependency Ratio GDP per Capita (t-1) Girls Mean Years of Education Complete Girls Expected
Population
Investment
Labor Force
Gross Domestic Product
Education
GDP per Capita
Life Expectancy
Human Development Index (HDI)
Calculated Variables
Nearly all the statistical relationships were estimated as log-linear relationships, meaning we regressed the log of the dependent variable against the logs of the independent variables. The double-log form provided the best fit between the variables and generated the smallest error term. Also, the coefficients can be interpreted as elasticities, meaning that the coefficient represents the percentage change in the dependent variable to a percentage change in the independent variable. The database used to create the behavioral equations contained data from 56 countries, but for some variables, data were only available for a smaller sub-section of these countries, as indicated by the n in the sample data tables.
Human Development Index (HDI)
We start with the HDI, since this is our ultimate measure of economic development and assesses quality of life. The HDI is based on three dimensions of human life:
1. 2.
A long and healthy life, as measured by life expectancy at birth Educational attainment, as measured by mean years of schooling for adults age 25 and older and expected number of years of basic education achieved 3. A decent standard of living, as measured by gross domestic product (GDP) per capita adjusted for purchasing power parity The HDI framework is shown in Figure 3. The specific equations to calculate the HDI, as well as countryspecific rankings, are found in the UN Human Development Reports. 20 Three indices (each accounting for one-third) consitute the index: life expectancy, education, and income.
Figure 3. Human Development Index
Female Life Expectancy at Birth
As indicated above, female life expectancy (FLE) at birth is seen as a high-level health impact indicator and figures into the calculation of the HDI. Life expectancy is also used in the dynamic component of the RAPID/Women model to calculate survival rates, which are in turn used in the population projections. Reductions in mortality rates in developing countries have, of course, contributed to their demographic transitions. These reductions have been due in large part to gains in childhood survival rates attributable to the introduction of vaccines and eradication of childhood communicable diseases, improved nutrition and access to water and sanitation, and improved birth spacing. 21 In RAPID/Women, we modeled FLE by establishing a statistical relationship with one of our child mortality indicators, the under-5 mortality rate. In developing countries, gains in life expectancy are first driven by gains in early childhood mortality. The estimated double-log regression equation for FLE, estimated with data from an international cross7
section of developing countries, is shown in Table 1. The results are statistically significant and show, as expected, that as under-5 mortality increases, FLE decreases, or conversely as under-5 mortality decreases, FLE increases.
Table 1. Female Life Expectancy
Ln (Female Life Expectancy) 22 Ln (Under-5 mortality rate) 23 Coefficient -0.19 t statistic 11.7 R-squared 0.71 n 55
In the model, male life expectancy at birth is assumed to differ from FLE by a fixed amount. In nearly every society, male life expectancy is lower than for females, usually between one and two years. The initial values for female and male life expectancy are drawn from the United Nations Population Divisions World Population Prospects. 24
Contraceptive Prevalence Rate
Fertility is calculated in the model by adapting the proximate determinants approach (see the section on the total fertility rate). The RAPID/Women model projects two of the proximate determinants using relationships uniquely devised for this model. The first modeled variable is the contraceptive prevalence rate (CPR), a key indicator for family planning that serves as a behavioral- or outcome-variable that contributes to the total fertility rate. There are multiple determinants of contraceptive use, but for RAPID/Women, we needed to construct a statistical relationship to calculate CPR based on specific, quantifiable variables. The extensive literature on the determinants of contraceptive use discusses both demand and supply variablesat multiple levels for each: individual, familial, and extra-familial for demand; and delivery system, managerial context, policy context, and client transactions for supply. 25 In RAPID/Women, we established a statistical relationship between demand-side and supply-side factors that one would expect to influence the uptake of family planning services. Contraceptive use has been shown to be heavily influenced by womens educational attainment, as well as social norms about family size, womens income-earning opportunities (the opportunity cost of womens time), household wealth, womens autonomy, and access to high-quality services, among other variables.26 After narrowing the list of possible variables to those for which data were available across a sufficiently large number of countries of interest and were statistically significant and of the expected sign in the regression model, we selected three variables to include as determinants of contraceptive use in the model. The first indicator is the average years of schooling for women, which is based on two variables: the mean years of schooling completed among adults age 25 and older 27 and the school life expectancy (expected number of years of school completed) among children now entering primary school. 28 It has been well established that female education is an important determinant of health behavior, including reproductive health behavior. 29 Women with higher levels of education tend to marry later, want fewer children or wish to space them apart, are more informed about their choices, and are better able to act on those choices. As Rosenzweig and Schultz explained almost three decades ago, education builds peoples skills to acquire and decode new information and thus effectively lowers the costs of using more beneficial child health and contraceptive technologies. 30 Evidence from developing countries since that time has supported the relationship between womens education and health behavior, while identifying multiple confounders that can affect the strength of the relationship.31 Second, we include an indicator of womens empowerment: womens report that wife beating is justified for specific reasons (e.g., burning the food, arguing with her husband, leaving the house without telling her husband, neglecting the children, and refusing sex). Women who agree that intimate partner violence is justified are likely to have experienced violence themselves (or live in communities where violence is 8
highly prevalent and strong, gendered power imbalances persist). We expect women who experience violence and/or live in high-prevalence communities to be less able to make autonomous decisions, including those regarding the use of contraceptives and other health behaviors. 32 Womens acceptance of wife beating (intimate partner violence) is an indicator of womens autonomy; 33 so we expect that as the percentage of women who justify intimate partner violence goes up, the CPR would go down. The third variable selected as a determinant of contraceptive use is an index called the Contraceptive Security Access Score 34 (CS Access Score). The CS Access Score is an established index that is based on key informant interviews about the family planning situation in a country, as well as information about financing. This indicator measures a countrys access to modern family planning methods, the proportion of a countrys contraceptives distributed through the public sector that go to the poor, and the spread of access to family planning methods. We expect that as this indicator increases, the CPR also will increase. Table 2 shows the double-log regression results for CPR as a function of the three variables discussed above. It can be seen that all independent variables have the expected sign and are significant.
Table 2. Contraceptive Prevalence Rate
Ln (CPR) 35 Ln (Female average school attendance, years) 36 Ln (Wife beating justified) 37 Ln (CS access score) 38 Coefficient 0.48 -0.16 1.81 R2 = 0.69 t statistic 5.1 -1.76 3.85 N = 36
The initial value for CPR (any method) can be drawn from the most recent DHS report, or, if none is available, from the United Nations Population Divisions World Contraceptive Use. It should be noted that in the model, the CPR is capped to be less than or equal to 75 percent.
Percentage of Women Married/ in Union
Another important proximate determinant of fertility, which is the second modeled in RAPID/Women, is the percentage of women of reproductive age who are married or in union, which serves as a proxy for sexual exposure. 39 This factor accounts for an important percentage of the decline in fertility during demographic transitions. Age at marriage and also the propensity to marry among women are closely related to their education status and level of educational attainment. Girls who remain in school and/or who go on to university will tend to marry at older ages. 40 More educated girls will also tend to enter the formal labor force and work, which may have an additional effect on age at marriage. 41 For the RAPID/Women model, we estimated a simple equation (see Table 3) that relates the expected number of years in school for girls to the percentage of women married/in union. Note that the coefficient on the education indicator has the expected negative sign and is significant.
Table 3. Percentage of Women Married/in Union

Ln (Percent of women married/in union) 42 Ln (Expected years in schoolgirls) 43 Coefficient -0.73 t statistic -3.0 R-squared 0.15 n 51
The initial value for the percentage of women of reproductive age married or in union can be drawn from the countrys most recent DHS report.
Total Fertility Rate
The total fertility rate (TFR) is used to calculate the number of births in the demographic component of the model. We use the Bongaarts proximate determinants model 44 to project the TFR. The Bongaarts framework is repeated here in full:
TFR = Cm Ci Cat Cs Cc TF where: TFR = total fertility rate Cm = marriage index Ci = insusceptibility index Ca = abortion index Cs = sterility index Cc = contraception index TF = total fecundity
Cm, the index of marriage, is simply the percentage of women in the reproductive age group who are married or in union. The baseline value is input from the DHS, and future values are calculated by the model as described above. Ci, the index of postpartum infecundability, is calculated as the ratio of the average birth interval in months with and without breastfeeding:
Ci = 20.0 / (18.5 + Period of postpartum insusceptibility)
In RAPID/Women, this is represented as postpartum insusceptibility (PPI) and is an exogenous variable that can increase the duration of a womans postpartum amenorrhea (temporary infecundability or inability to conceive). While many factors may influence this, prolonged breastfeeding has been shown to be one of them. The baseline value is input from the DHS, and future values are projected by the user, who can choose whether to vary from the baseline value. Ca is an index of induced abortion. Since we do not vary the level of this index in the RAPID/Women model, we have omitted it and adjusted the calculations accordingly. Cs, the index of sterility, is normally calculated from the percentage of women in union who remain childless at the end of their reproductive years. As with PPI, the baseline value is input from the DHS, and future values are projected by the user, who can choose whether to vary from the baseline value. However, it is unlikely that this value would vary much over time.
10
Cc, the index for contraception, is calculated as a function of the CPR, which we project using the equation described above. TF is an index of total fecundity and is the number of children a women would have if all the other proximate determinants were at their minimum levels. As with the abortion rate, we do not use this variable, as it is not likely to change and is not a policy variable.
Under-Five Mortality, Infant Mortality, and Maternal Mortality
Mortality rates are key indicators because not only do they reflect outcomes of health and other interventions, but they also influence demographic progress through the population projection component of the model. In the RAPID/Women model, we estimate under-five and infant mortality rates, as well as maternal mortality, by establishing links between fertility and high-risk births. There are two paths by which changes in fertility can lead to reductions in the number of under-five, infant, and maternal deaths. The most direct path is through a reduction in the absolute number of births and hence in the number of infants and young children at risk of dying and in the number of pregnant women giving birth. The second path is through a reduction in the proportion of births classified as high risk. This leads to a reduction in the infant and under-five mortality rates, as well as in the maternal mortality ratio. The DHS defines high-risk births as those that fall into one of the following risk categories: Mother is under age 18 (too young) Mother is over age 34 (too old) Birth is less than 24 months after previous birth (too close) Birth is to a mother who has had more than three births (too many) 45
The DHS program collects data on the risk factors facing women and on infant and child mortality. The data from a cross-section of countries with two or more DHS surveys were used to establish a statistical relationship between the variables concerned. We expect that as the percentage of high-risk births falls, so do under-five mortality rates (U5MR). A similar picture holds for the infant mortality rate (IMR) and the maternal mortality ratio (MMR). As described above, one of the high-risk categories is high parity. As the TFR declines, high parity naturally also declines. The other three risk categories also correlate with the TFR. Higher fertility is often associated with long periods of childbearing; thus, more births occur in the too young and too old categories. Moreover, births spaced too closely together often occur in high-fertility societies. Using data from a cross-section of several DHS, we followed Rosss analysis 46 and established the slope 47 of the relationship between changes in the percentage of at-risk births and changes in infant, child, and maternal mortality. We also estimated slopes for changes in TFR and changes in the percentage of at-risk births. We then used the relationships to model the effects of changes in contraceptive use (through changes in fertility) on infant, under-five, and maternal mortality. Thus, the pathway for the model is as follows:
TFR Percent of births at Risk IMR TFR Percent of births at Risk U5MR TFR Percent of births at Risk MMR
11
The results of the statistical analysis established the relationships that are used in RAPID/Women to link fertility to mortality (see Table 4).
Table 4. High-Risk Births, Fertility, and Under-5, Infant, and Maternal Mortality
Any Risk* TFR Infant Mortality Rate* Any Risk Under-5 Mortality Rate* Any Risk Maternal Mortality Ratio** Any Risk Coefficient 5.18 Coefficient 1.55 Coefficient 3.86 Coefficient 21.27 Rsquared 0.69 Rsquared 0.31 Rsquared 0.35 Rsquared 0.24 n 93 n 93 n 93 n 64
Sources: * Moreland and Talbird, 2006. 48 ** Ross and Stover, 2010 49
Because the IMR and U5MR are modeled independently, the model could return predicted values of the IMR that exceed that of the U5MR. Because it is not possible for deaths among children under age one to exceed that of all deaths to children under age five, the model imposes a condition that the IMR cannot be greater than the U5MR. The initial value for infant and child mortality rates and the maternal mortality ratio can be drawn from the countrys most recent DHS report. Alternately, infant and child mortality data are available from the United Nations Population Divisions World Population Prospects 50 and maternal mortality from World Health Organization et al., Trends in Maternal Mortality.
Gross Domestic Product Per Capita
As seen in the overall schema of the model, RAPID/Women estimates the impact of demographic variables on national development variables. We take gross domestic product per capita (GDPPC) as our measure of economic development. GDPPC is also an element in the calculation of the HDI, as discussed above. GDPPC is simply GDP divided by the population:
GDPPC = GDP/Population
The population is projected using the population projection model. GDP is calculated using a modified production function at the national level. In the production function, real output is determined by labor and investment. For the RAPID/Women model, we added an education term (female average years of school attendance) to the production function to take account of the impact of education on the productivity of the labor force. The estimated production function is shown in Table 5. All independent variables are significant, with the expected signs.
12
Table 5. GDP Production Function

Ln (GDP) 51 Ln (Investment) 52 Ln (Labor force) 53 Ln (Female average school attendance, years) 54 Coefficient 0.65 0.35 0.38 R2 =0.97 t statistic 24.7 10.8 5.37 N =89
Labor Force
We calculate the labor force from a parameter representing the proportion of the population in the labor force, using data from the International Labor Organization, and the total population projected by the model:
Labor Force = Population * [Labor Force Participation Rate] Investment (Capital)
The investmentor capitalterm is calculated using an estimated relationship between GDPPC, female education, and the dependency ratio (the share of the non-working age or dependent populationdefined as those who are under age 15 and age 65 and olderto those of working age). As discussed above, investment is included in the model, as it is an important factor in GDP and often is included in models as a proximate determinant of economic growth.55 Capital formation or investment is the part of income that is set aside or saved to invest in income-producing assets such as machines, tools, roads, and other fixed assets that enhance production. We expect that as income per capita increases and as education levels, wages, and disposable income rise, savings and investment also will increase. In addition, there is research to suggest a life-cycle effect on savings behavior: when households and societies have higher numbers of people above or below the working ages (old or young), we expect they will use more of their income to feed, clothe, educate, and provide health services to these non-working people, which will lower the total savings rate and hence reduce investment.56 So, we expect a negative relationship between the dependency ratio and investment: as declines in fertility over time reduce the dependency ratio, savings rates increase, and, subsequently, more capital is available for investment.57 Table 6 presents the results.
Table 6. Capital per Capita
Ln (Capital per capita) 58 Ln (GDP per capita) 59 Ln (Expected years in schoolgirls) 60 Ln (Dependency ratio) 61 Coefficient 0.87 0.43 -0.50 R2 = 0.93 t statistic 12.76 2.27 -2.24 N = 50
13
SAMPLE RESULTS FROM AN APPLICATION IN MALI

Mali Baseline Situation
The model was tested and validated recently in Mali. Mali was chosen as a pilot country because it faces a lot of the issues that the approach was designed to address. Table 7 summarizes the baseline (initial year) data for Mali. Fertility is high, with a TFR of 6.6 and with a contraceptive prevalence rate of 8.2 percent (all methods).62 Economically, although GDP per capita is just over $1,000, Mali is in the bottom ranking on the HDI scale at 160 out of 169 countries. The IMR is 96 deaths per 1,000 births, and the U5MR is 191. Similarly, there are significant womens empowerment issues. More than 75 percent of women report finding it acceptable for their husbands to beat them and 85 percent have been subjected to some form of female genital mutilation. 63
Table 7. Mali Baseline
Indicator % married CPR TFR Births at risk IMR U5MR MMR Female Life Expectancy Investment GDP/pop Base Year Value 84.8 8.2 6.6 68% 96 191 464 48.8 252 1083
Scenarios
As described above, the RAPID/Women model has several indicators that can be manipulated to simulate the impact of various interventions (control indicators.) These include two education variables, one womens empowerment variable, and three family planning indicators.
Table 8 shows the baseline or initial values of these variables in 2010, HDI Ranking 160 as well as the assumed values of these for three scenarios in the end 64 year of the simulation, 2050. In the FP only scenario, we assumed that the CS Access Index increased from an initial level of 9.8 (out of 20) to 18, which represents almost a doubling. PPI is assumed to increase from 11.7 months at baseline to 17 months as a result of increased breast feeding and use of the lactational amenorrhea method (LAM). Natural sterility (infertility) is left unchanged. All other control indicators are assumed to remain unchanged.
Table 8. Baseline Values of Four Scenarios (Mali)
Scenario Value in Expected Years Female Education 8 8 10 10 Expected Years Male Education 9 9 10 10 Mean Years Female Education 0.99 0.99 4 4 Mean Years Male Education 1.76 1.76 4 4 Wife Beating Justified Access Index PPI
Base scenario FP only Womencentered Combined
2010 2050 2050 2050
75.2 75.2 40 40
9.8 18 9.8 18
11.7 17 11.7 17
The third scenario is titled women-centered because it only takes account of changes in womens education and the womens empowerment variables (acceptance of gender-based violence). Of course, family planning is also a strategy that focuses on women, but here we separate these for expository purposes. The indicators for mean years of female and male education are assumed to increase to 10 years and reach parity in 2050. These increases may not seem large, but the initial values compare favorably to 14
countries at much higher levels of development than Mali. The indicators for female and male expected years of completed education are also assumed to improve for both sexes and reach parity by 2050. The simulated improvements of some three years for girls and a little more than two years for boys would put Mali, in 2050, at a level similar to Malawi or Morocco. We looked at the relationship across 176 countries and determined that this level was consistent with the projected expected years of education. The womens empowerment indicatorpercentage of women who agree that wife beating is justifiedis assumed to fall from 75.2 percent to 40 percent by the end of the simulation period. Finally, the scenario titled combined puts the FP-only and women-centered scenarios together.
Scenario Results
Demographic Indicators
Table 9 shows the results for the demographic indicators of the three policy scenarios. We do not report the base scenario here. As we might expect, only the women-centered and combined scenarios have any effect on the percentage of women who are married because that indicator is influenced solely by the expected years of schooling among girls. In terms of contraceptive use, we see that under the FP-only scenario, the CPR increases from 8.2 percent to nearly 25 percent. While this may not seem like a lot over a 40-year period, the recent history of Mali has shown virtually no change in the CPR and this represents an increase of 0.4 CPR points a year, which is close to the sub-Saharan Africa average change in the past decade. Under the women-centered scenario, the CPR increases to 17.7 percent. This is because the CPR is a function of two women-centered indicators: acceptance of gender-based violence and the female mean years of education. Of interest is the combined scenario because the end year CPR is actually higher than the sum of the FP-only and womencentered strategies, so there are complementarities reflected from the fact that both strategies have a direct influence on the CPR. Regarding the impact on fertility (TFR), we see that the women-centered strategy has a slightly more pronounced effect on TFR than the FP-only strategy. The FP strategys effect on TFR is direct through the influence of the change in CPR, as described in the Bongaarts model. The women-centered strategy acts on TFR through the both the CPR and the percentage of women married. We see that the combined strategy results in a TFR of close to replacement level at 2.4 in 2050.
15
Sample Results from an Application in Mali
Table 9. Scenario Results of Demographic Indicators (Mali)

% Married FP Only WomenCentered Combined FP Only CPR WomenCentered Combined FP Only TFR WomenCentered Combined
2010 2015 2020 2025 2030 2035 2040 2045 2050
84.8 84.8 84.8 84.8 84.8 84.8 84.8 84.8 84.8
84.8 82.9 81.1 79.4 77.8 76.3 74.8 73.4 72.1
84.8 82.9 81.1 79.4 77.8 76.3 74.8 73.4 72.1
8.2 9.8 11.6 13.4 15.4 17.6 19.8 22.2 24.6
8.2 9.7 10.9 12.2 13.3 14.4 15.5 16.6 17.7
8.2 11.6 15.5 19.9 25.1 30.9 37.5 44.9 53.3
6.6 6.3 6.1 5.8 5.6 5.3 5.1 4.8 4.6
6.6 6.3 6.1 5.9 5.7 5.5 5.3 5.2 5.0
6.6 6.1 5.6 5.0 4.5 3.9 3.5 2.9 2.4
Births at Risk
In Table 10, we see the impacts of each scenario on the proportion of births at risk. Recall that this indicator is important in terms of determining infant, under-five, and maternal mortality rates. The TFR is the main influence here, so the effects of each scenario are indirect. We see that the impact on births at risk follows the impact on the TFR and that in the combined scenario, significantly fewer births are at risk, especially in the final years.
Table 10. Projected Births at Risk (Mali)
FP Only 2010 2015 2020 2025 2030 2035 2040 2045 2050 68% 67% 66% 64% 63% 62% 60% 59% 58% Women Strategy 68% 67% 66% 65% 64% 63% 62% 61% 60% Combined 68% 66% 63% 60% 57% 55% 52% 49% 46%
Mortality
The impact of births at risk on IMR, U5MR, and life expectancy at birth can be seen in Table 11. The impacts of each scenario mirror their impacts on births at risk. So we see that the women-centered 16
strategy has the smallest effect, reducing the IMR, for example to 83 in 2050 from 96 in 2010, while the FP-only strategy has a slightly greater effect by reducing the IMR to 80 in 2050. The combined strategy, however, decreases the IMR to 62, almost doubling the reduction from the base year. Similar results are seen in under-five mortality. The impacts of each strategy on life expectancy at birth have the same ranking but the impact is more muted.
Table 11. Projected IMR, U5MR, and Life Expectancy (Mali)
IMR FP Only 2010 2015 2020 2025 2030 2035 2040 2045 2050 96 94 92 90 88 86 84 82 80 Women Strategy 96 94 92 90 89 87 86 85 83 Combined 96 92 88 83 79 75 71 66 62 FP Only 191 186 181 176 171 165 160 156 151 U5MR Women Strategy 191 186 181 177 173 169 166 162 159 Combined 191 181 170 160 149 139 128 117 106 FP Only 48.8 49.1 49.3 49.6 49.9 50.1 50.4 50.7 51.1 Life Expectancy Women Strategy 48.8 49.0 49.3 49.5 49.7 49.9 50.1 50.3 50.5 Combined 48.8 49.3 49.9 50.5 51.1 51.9 52.6 53.5 54.6
The maternal mortality ratio is also affected by the births at risk. Table 12 shows the evolution of the MMR under each scenario. The differences between the FP only and women-centered strategies are not pronounced since the percentage of births at risk under these two scenarios are not that different, with the MMR falling from 464 to 442 and 443, respectively. The combined scenario, however, projects a lower ratio of 417 by 2050.
Table12. Maternal Mortality Ratio (Mali)
MMR FP Only 2010 2015 2020 2025 2030 2035 2040 2045 2050 464 461 458 456 453 450 447 444 442 Women Strategy 464 460 457 454 452 449 447 445 443 Combined 464 458 453 447 441 435 429 423 417
17
Sample Results from an Application in Mali
Population Effects Economic impacts. The main economic indicators are capital investment per capita, GDP per capita, and the HDI (see Table 13). For expository reasons, we report the HDI ranking of Mali under each scenario relative to other countries in the base year. Of course, during the 40-year simulation period, the HDI of other countries may change and with it their own ranking. Table 13. Projected Economic Indicators and HDI Ranking (Mali)
Capital per Capita FP Only Women Strategy Combined FP Only GDP/pop Women Strategy Combined FP Only HDI Ranking Women Strategy Combined
2010 2015 2020 2025 2030 2035 2040 2045 2050
$252 $251 $253 $258 $272 $287 $298 $307 $317
$252 $295 $371 $464 $584 $711 $838 $966 $1,117
$252 $292 $377 $483 $ 637 $810 $995 $1,212 $1,491
$1,171 $1,174 $1,184 $1,205 $1,253 $1,303 $ 1,340 $1,371 $1,406
$1,171 $1,458 $1,851 $2,303 $2,843 $3,405 $3,970 $4,543 $5,183
$ 1,171 $1,448 $1,871 $2,363 $3,004 $3,700 $4,427 $ 5,241 $ 6,219
161 160 160 159 158 158 158 158 157
161 157 155 154 146 143 133 132 129
161 157 155 153 144 135 132 128 121
Population impacts. The impact on population is shown in Table 14. Due mainly to effects on fertility of each scenario, we see the impacts on the projected total population of Mali up to 2050. Under the FPonly scenario, total population is projected to reach 47.4 million in 2050, while the women-centered scenario results in a lower projected population of 43.5 million. Combining the FP and women-centered scenarios results in an even lower 2050 population of approximately 36 millionsome 11 million fewer than the FP-only scenario.
It is interesting to compare these projections to the latest three UN variants; 65 the high-fertility, medium, and low variants predict a population of 46.7 million, 42.1 million, and 37.7 million, respectively. Thus, the FP-only scenario is similar to the high-fertility variant, the women-centered to the medium variant, and the combined to the low variant.
Table 14. Projected Population under Three Scenarios (Mali)
Total Population FP Only 2010 2015 2020 2025 2030 15,395,200 18,099,498 21,012,361 24,238,568 27,865,066 Womencentered 15,395,200 18,041,131 20,805,600 23,793,556 27,080,428 Combined 15,395,200 17,969,825 20,533,877 23,161,113 25,878,012
18
Modeling the Impact of Family Planning and Other Women-Centered Interventions on Development 2035 2040 2045 2050 31,983,113 36,633,161 41,790,587 47,381,115 30,726,776 34,732,597 39,034,144 43,536,658 28,664,646 31,422,249 33,988,047 36,188,668
Combining the population projections with the changes in the mortality rates allows us to calculate the impact on infant, under-five, and maternal deaths. In Table 15, we see how the scenarios affect cumulative deaths averted.
Table 15. Deaths Averted Under Different Scenarios (Mali)
Infant Deaths Averted FP Only
2010 2015 2020 2025 2030 2035 2040 2045 2050 307,472 350,022 427,511 549,943 733,768 1,001,359 1,378,881 1,893,076
U5 Deaths Averted FP Only

553,572 628,338 768,394 991,273 1,325,744 1,811,780 2,497,615 3,432,447
Maternal Deaths Averted FP Only

511 2,037 4,845 9,299 15,965 25,622 39,267 58,012
Women Strategy
310,401 370,846 476,033 635,789 868,307 1,198,701 1,655,591 2,266,266
Combined
312,827 413,764 590,704 860,512 1,251,461 1,799,090 2,563,022 3,527,483
Women Strategy
557,201 658,657 841,830 1,122,919 1,532,452 2,114,116 2,919,898 3,999,855
Combined
565,476 738,816 1,052,846 1,535,968 2,236,438 3,216,809 4,597,300 6,326,035
Women strategy
643 2,471 5,685 10,581 17,666 27,657 41,471 60,077
Combined
1,030 4,088 9,700 18,577 31,776 50,651 76,801 111,857
Conclusions
The application of the model to the data in Mali shows clearly that synergies can be realized when family planning and women-centered strategies are implemented together. The contraceptive prevalence rate is nearly two times greater under the combined scenario, compared with the FP-only scenario, with parallel impacts on fertility rates. This, in turn, has consequences for the percentage of births at risk and for early childhood and maternal mortality; the combined scenario results in significantly lower infant, under-five, and maternal mortality rates when compared with either the FP-only or women-centered scenarios alone. The model also shows that in the economic sphere, family planning has a modest impact on per capita income, but women-centered strategies, acting primarily through education, have significant effects that are amplified under a combined strategy. Consequently, Malis HDI ranking benefits more under a combined strategy when compared with the FP-only or women-centered strategies. Clearly, the strategies considered in the model and simulated in the case of Mali can be expected to benefit Mali. The model also shows that women-centered strategies, such as girls education or changes in gender norms, should be seen as complementary to, and not substitutes for, efforts to provide family planning. Therefore, to the extent possible, countries like Mali should simultaneously implement strategies that benefit women through family planning, girls education, and favorable gender norms in order to maximize impacts on the quality of life.
19
USING THE RAPID/WOMEN COMPUTER MODEL

The current version of RAPID/Women uses a hybrid software approach that combines an Excel model with the DemProj component of Spectrum. 66 Figure 4 shows how the two models are linked. To complete a projection, three steps are involved:
1.
Initially, users complete data entry in Excel for values of the policy and program inputs shown in Figure 2a, as well as GDP per capita and investment, during the baseline year. Users also input target values for the education, womens empowerment, and family planning indicators in the final projection year under four different policy scenarios. The Excel model then uses these baseline and projected values to calculate the demographic outputs of TFR and life expectancy. The user links the Excel model to DemProj, which uses TFR and life expectancy at birth to project the population over the time period specified in the model. Key population variables calculated by DemProj subsequently are used to calculate GDPPC and impacts on child and maternal mortality. The HDI is also calculated during this step.
Figure 4. RAPID/Women Computer Model Structure Excel Model Spectrum/Demproj
2. 3.
20
Using the Model

The Excel-based model has sheets set up for different functional areas. The model is currently programmed so that one can have four scenarios. One of these is the base scenario and is meant to be a scenario where everything stays the samethat is, all variables remain constant. The three other scenarios are user-defined. If the user prefers, the base scenario can also be used to represent a user-defined scenario that reflects changes in input or control variables. Each scenario is given a name. The Excel model is programmed so that there is a master control worksheet (called Control) where the user enters the initial base-year values of all the models variables and also defines the end values of the input or control variables. There are four worksheets that solve the model for each scenario: Base, Policy1, Policy-2, and Policy-3. There are two worksheets used as intermediary worksheets to handle the data flow between the Excel model and DemProj: ToDemproj and Population. The user does not have to enter data into these worksheets. There is also a worksheet that calculates child and infant deaths and a worksheet that calculates maternal deaths. Lastly, there are several worksheets that have graphs of the models main variables.
Step 1: Establish Initial Values
The first step a user needs to do is to enter the initial year (baseline) values for the inputs and main outputs of the model. This takes place in Row 4 of the Excel sheet called Control (see Figure 5).
Figure 5. Initializing the models inputs and main outputs
Inputs EDUCATION Mean Expected Yrs Yrs Male Female 9 4.46 9 4.46 Women's Empowerment Mean Yrs Male 5.76 5.76 Wife Beating Justified 53.5 53.5 Access Indx 10.6 10.6 Family Planning PPI 11.4 11.4 Sterility 1 1
Scenario Name Base
Value in: 2010 2050
Expected Yrs Female 9 9
Base Year
%Married 63.3
CPR 34.4
TFR 5.4
Initial Values of Calculated Variables Female Life Births @ Risk IMR U5MR MMR Expectancy 56% 51 81 454 56.2
Investment/POP 238
GDP/pop 1433
These are the only values for which the user needs to locate data. To assist the user in setting up the model, we have provided a database of the model that was used to establish the behavioral equations. (File: RW data 3-1-12.xlsx). Some of these data may have evolved or been revised, so it is recommended to verify the most current version of the sources cited in the database.
Step 2: Define Policy Scenarios
The second step is to establish the four policy scenarios. Figure 6 gives an example using data from Mali.
21
Using the RAPID/Women Computer Model
Figure 6. Policy Scenarios

Inputs EDUCATION Mean Expected Yrs Yrs Male Female 9 4.46 9 4.46 11 11 9 8.9 8.9 4.46 Women's Empowerment Mean Yrs Male 5.76 5.76 8.9 8.9 5.76 Wife Beating Justified 53.5 53.5 28.5 28.5 53.5 Family Planning Access Indx 10.6 10.6 18 10.6 18 PPI 11.4 11.4 11.4 11.4 11.4 Sterility 1 1 1 1 1
Scenario Name Base WomenStrat+FP WomenStrat FP
Value in: 2010 2050 2050 2050 2050
Expected Yrs Female 9 9 11 11 9
In the example shown in Figure 6, the base scenario simply holds all program variables constant over the time period of the projection, as described above. The outputs of this scenario inform the user what demographic, health, and development changes could be expected with no changes in the current policy environment. The scenario labeled WomenStrat+ FP corresponds to the worksheet labeled Policy-1, which shows the models calculations for this scenario. Similarly, the scenario labeled WomenStrat, is Policy-2, and FP is Policy 3 (FP-only). The names given to the scenarios here are used to label the results in the graphs. To set up a scenario, the user specifies an end value for each of the six policy indicators in the end year of the projection. In the Excel model, the final year default is set at 2050. 67 The model interpolates linearly between the initial year (here, given as 2010) and the final year (2050.) In a later section, we discuss strategies for defining target endline values. Users must also input values for the difference between male and female life expectancy at birth (this can be found in the United Nations World Population Prospects). The model assumes that this difference remains constant. It is normal that male life expectancy is lower than female life expectancy. Also the user must specify the labor force participation rate of the adult population (available from the International Labor Organization),which is also assumed to be constant (see Figure 7).
Figure 7. Labor Force Participation and Male and Female Life Expectancy
Labor Force Participation Rate Female-Male LE difference 0.49 1.6
Step 3: Project the Population
The third step is to run DemProj to perform population projections using the Excel models outputs for life expectancy and fertility as inputs. Several key demographic variables are then exported back to the Excel model from DemProj using the RAPID Transfer option in Spectrum. The first step is to open Spectrum (see Figure 8).
22
Figure 8
Unlike other Spectrum applications, the user should not open any Spectrum files or create any at this stage. Next, on the Tools page, select RAPID Transfer (see Figure 9):
Figure 9
23
Once RAPID Transfer is opened, select a country by clicking all four scenarios for the country chosen and select the RAPID/Women Excel file to be used (see Figure 10).
Figure 10
Once this step is completed, click on Calc. RAPID Transfer will run the four population projections by reading the TFRs and life expectancies from the Excel model for each scenario from the worksheet called To Demproj and will transfer the required demographic outputs for each scenario back to the Excel model in the worksheet called Population. There is no need to open DemProj. As noted above, when using the RAPID Transfer option in Spectrum, no Spectrum files can be open. However, if the user wishes to open the Spectrum projections after the calculations are made, check the box at the bottom of the page before the clicking on the Calc button (see Figure 10).
Step 4: View the Results
Users can view the results by looking at the projections in each of the calculated worksheets (Base, Policy-1, Policy 2, Policy-3, Child Survival, and Maternal Mortality). Or the user can view some of the major outputs with the provided graphs; Figure 11 shows these outputs.
24
Figure 11: Sample Output
Total Population
160 140 120 100 Millions 80 60 40 20 0 139 114 94
Base FP WomenStrat WomenStrat+ FP
81
2010
2015
2020
2025
2030
2035
2040
2045
2050
Defining Scenarios
An important step in using the model is defining the three policy scenarios. As discussed above, this consists of defining final year values for the four education indicators, the womens empowerment indicator, and the three family planning indicators. In this section, we address strategies for setting objectives or targets. In general, there are several approaches to establishing targets: Use established long-term goals (such as the Millennium Development Goals) Base the target on past performance (of a program, increasing by X percent) Emulate a high performer (a stellar program nearby) Determine what is feasible (financial, human resources, local population)
Ideally, as with any model, the choice of approach and the actual targets should be developed in consultation with local participants or partners. As an aid to users in considering reasonable and achievable targets for the education indicators and the gender-based violence indicators, we present some relationships between these and GDP per capita and between the two education indicators, based on the international cross-section data that we used to estimate the models equations. In Figure 12, we see the relationship between expected years of completed schooling for girls and GDP per capita. The black line is a curve showing the relationship.
25
Figure 12. Expected Years of SchoolingFemales

16 14 12 10 Years 8 6 4 2 0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
GDP (US$)/Pop
Figure 13 likewise shows the relationship between the average years of schooling among females age 25 and older and GDP per capita.
Figure 13. Mean Number of Years of SchoolingFemales >25
16 14 12 10
Years
8 6 4 2 0 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000
GDP (US$)/Pop
26
In Figure 14, we see how these two education variables compare cross-sectionally. Of course, the number of years of schooling attained by women age 25 and older will be a function of their schooling experience when they were younger. Ideally, one would have a model that links these two together dynamically, but that is beyond the scope of RAPID/Women.
Figure 14. Relationship between Two Education Indicators
Average Number Years of Schooling females >25 12 10 8 6 4 2 0
10
12
14
16
Expected Years of SchoolingFemales
Next, in Figure 15, we show how GDP per capita and the gender-based violence indicator are related. The fitted curve indicates a tendency for acceptance of gender-based violence to decline with the level of development, as measured by GDP per capitaalthough the scattergram indicates quite a range.
27
Figure 15. Percent of Women Who Agree that Wife Beating Is Justified
100 90 Percent of Women 80 70 60 50 40 30 20 10 0 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000
GDP/Pop US$
Finally, in Figure 16, we show how the family planning access score varies with GDP per capita. There is a slight tendency for the score to increase with development, but the relationship is weak.
Figure 16. Family Planning Access Score
14 12 10 8 6 4 2 0
Score
2,000
4,000
6,000
8,000
10,000
12,000
14,000
GDP (US$)/Pop
28
ENDNOTES
1
U.S. Government Global Health Initiative (2011). Global Health Initiative Supplemental Guidance on Women, Girls, and Gender Equality Principle. Available at http://www.ghi.gov/resources/guidance/161891.htm. U.S. Government Global Health Initiative (2011). The United States Government Global Health Initiative Strategy. Available at http://www.ghi.gov/resources/strategies/159150.htm. Global Fund. (2008). Global Fund Gender Equality Strategy. Geneva: Global Fund to Fight AIDS, Tubercolosis and Malaria. World Bank. (2006). Gender Equality as Smart Economics. Washington, DC: World Bank. Abu-Ghaida, D., and Klasen, S. (2004). The costs of missing the Millennium Development Goal on gender equity. World Development 32(7): 10751107. Shroff, M., Griffiths, P., Adair, L., Suchindran, C., and Bentley, M. (2009). Maternal autonomy is inversely related to child stunting in Andhra Pradesh, India. Maternal and Child Nutrition 5: 6474. Behrman, J., Murphy, A., Quisumbing, A. and Yount, K. (2009). Are returns to mothers human capital realized in the next generation? The impact of mothers intellectual human capital and long-run nutritional status on childrens human capital in Guatemala. IFPRI Discussion Paper 00850. Washington, DC: International Food Policy Research Institute. Gill, K., Pande, R., and Malhotra, A. (2007). Women deliver for development. Lancet 370: 134757. Abu-Ghaida and Klasen, 2004. Gage, A., Sommerfelt, E., and Piani, A. (1997). Household structure and childhood immunization in Niger and Nigeria. Demography 34(2): 295309. Smith, L., Ramakrishnan, U., Ndiaye, A., Haddad, L. and Martorell, R. (2003). The importance of womens status for child nutrition in development countries. IFPRI Research Report. Washington, DC: International Food Policy Research Institute.
3 4 5
Behrman et al., 2009. Smith et al., 2003. Hallman, K. (2002). Mother-father resources, marriage payments and girl-boy health in rural Bangladesh. IFPRI Discussion paper briefs. Washington, DC: International Food Policy Research Institute. Quisumbing, A., and Maluccio, J. (2003). Intrahousehold allocation and gender relations: New empirical evidence from four developing countries. In Household decisions, gender and development: A synthesis of recent research, edited by Agnes R. Quisumbing. Washington, DC: International Food Policy Research Institute. Hausman, R., and Szkely, M. (2001). Inequality and the family in Latin America. In Population matters: Demographic change, economic growth and poverty in the developing world, edited by N. Birdsall, A.C. Kelley, and S. Sinding. New York: Oxford University Press. Udry, C., Hoddinott, J., Alderman, H., and Haddad, L. (1995). Gender differentials in farm productivity: Implications for household efficiency and agricultural policy. Food Policy 20: 407423.
Klasen, S. (2002). Low schooling for girls, slower growth for all? Cross-country evidence on the effect of gender inequality in education on economic development. The World Bank Economic Review 16(3): 345373. Waring, M. 1988. If Women Counted: A New Feminist Economics. New York: Harper & Row.
Abu-Ghaida and Klasen, 2004. Klasen, 2002. Knowles, S., Lorgelly, P., and Owen, D. (2002). Are educational gender gaps a brake on economic development? Some cross-country empirical evidence. Oxford Economic Papers 54(2002): 118149. World Bank (2001). Engendering developmentthrough gender equality in rights, resources and voice. A World Bank Policy Research Report. Washington, DC: The World Bank. World Bank, 2006. Wang, G. (2005). Exploring a Gender and Development (GAD) model of womens reproductive health. Paper presented at the annual meeting of the American Sociological Association, Marriott Hotel, Loews Philadelphia Hotel, Philadelphia, PA, August 12, 2005. Downloaded 9/1/2009 from http://www.allacademic.com/meta/p19881_index.html.
29
Endnotes
See also: Selvaggio, K., Mehra, R., Fox R., and Gupta, G. (2008). Value added: Women and U.S. Foreign Assistance for the 21st century. Washington, DC: International Center for Research on Women (ICRW) and Women Thrive Worldwide. Meinzen-Dick, R., Johnson, N., Quisumbing, A., Njuki, J., Behrman, J., Rubin, D., Peterman, A., and Waithanji, E. (2011). Gender, assets and agricultural development programs. CAPRi Working Paper No. 99. Washington, DC: International Food Policy Research Institute. Hausman, R., Tyson, L., and Zahidi, S. (2011). The Global Gender Gap Report 2011. Geneva: World Economic Forum.
10
World Bank. 2006. World Development Report 2007: Development and the Next Generation.Washington, DC: World Bank. See also: Sen, A. (1999). Development as freedom. Oxford: Oxford University Press. Malhotra, A., Schuler, S., and Boender, C. (2002). Measuring womens empowerment as a variable in international development. Background paper. Washington, DC: World Bank.
11
Bloom, D., D. Canning, G. Fink, and J.E. Finlay. (2009). Fertility, Female Labor Force Participation, and the Demographic Dividend. Journal of Economic Growth 14(2): 79-101. Grown, C., G. Rao Gupta, and A. Kes. (2005). Taking Action: Achieving Gender Equality and Empowering Women, UN Millennium Project Task Force on Education and Gender Equality. London and Sterling, VA: Earthscan. Gill et al., 2007. Al Riyami, A., M. Afifi, and R.M. Mabry. 2004. Womens Autonomy, Education and Employment in Oman and their Influence on Contraceptive Use. Reproductive Health Matters 12(23): 144154. Allen, R. (2007). The role of family planning on poverty reduction. Obstetrics & Gynecology 110(5): 9991002. Cleland, J., Bernstein, S., Ezeh, A., Faundes, A., Glasier, A., and Innis, J. (2006). Family planning: The unfinished agenda. Lancet 368(9549): 18101827.
12
Singh, S., Darroch, J., Ashford, L., and Vlassoff, M. (2009). Adding it up: The costs and benefits of investing in family planning and maternal and newborn health. New York: Guttmacher Institute and United Nations Population Fund. Bloom, S. (2008). Violence against women and girls: A compendium of monitoring and evaluation indicators. Chapel Hill, NC: Carolina Population Center, University of North Carolina at Chapel Hill, MEASURE Evaluation Project. Institute for Reproductive Health (2011). Summary report: Preliminary formative research findings. A report for GREAT: Gender Roles, Equality and Transformations Project. Washington, DC: Institute for Reproductive Health, Georgetown University. Available at http://www.irh.org/sites/default/files/Summary_GREAT_Preliminary_Research.pdf.
13
14
Shroff et al., 2009. Jejeebhoy, S. and Sathar, Z. (2001). Womens autonomy in India and Pakistan: The influence of religion and region. Population and Development Review 27(4):687-712. Jejeebhoy, S. (1998). Womens autonomy in rural India: its dimensions, determinants and the influence of context. In Womens empowerment and demographic processes: Moving beyond Cairo, edited by H. Presser and G. Sen. Oxford: Oxford University Press.
15
Woldemicael, G., and Tenkorang, E. (2010). Womens autonomy and maternal health-seeking behavior in Ethiopia. Maternal and Child Health Journal 14(6): 988998. Balk, D. (1994). Individual and community aspects of womens status and fertility in rural Bangladesh. Population Studies 48: 2145. Bloom, S., Wypij, D., and Das Gupta, M. (2001). Dimensions of womens autonomy and the influence on maternal health care utilization in a north Indian city. Demography 38(1): 6778. Vlassoff, C. (1992). Progress and stagnation: Changes in fertility and womens position in an Indian village. Population Studies 46(2): 195212.
16
Hindin, M. (2003). Understanding womens attitudes towards wife beating in Zimbabwe. Bulletin of the World Health Organization 81: 501508. Shroff et al., 2009.
17
30
Woldemicael and Tenkorang, 2010. Balk,, 1994. Bloom, Wypij and Das Gupta, 2001; Vlassoff, 1991.
18 19
Bloom, 2008. Malhotra et al., 2002 Luttrell, C., and Quiroz, S. (2009). Understanding and operationalisiing empowerment. Working Paper 308. London: Overseas Development Institute. Panda, P., and Agarwal, B. (2005). Marital violence, human development and womens property status in India. World Development 33(5): 823850.
20
United Nations Development Programme (2011). Human Development Report 2011: Sustainability and equity. New York: United Nations Development Programme. Nakamura, H., Ikeda, N., Stickley, A., Mori, R., and Shibuya, K (2011). Achieving MDG 4 in sub-Saharan Africa: What has contributed to the accelerated child mortality decline in Ghana? PLoS One 6(3): e17774. Muldoon, K., Galway, L., Nakajima, M., Kanters, S., Hogg R., Bendavid, E. and Mills, E. (2011). Health system determinants of infant, child and maternal mortality: A cross-sectional study of UN member countries. Globalization and Health 2011, 7:42, available at hppt://www.globalizationandhealth.com/content/7/1/42. Mturi, A., and Curtis, S. (1995). The determinants of infant and child mortality in Tanzania. Health Policy and Planning 10(4): 38494. Alam, N. (1995). Birth spacing and infant and early childhood mortality in a high fertility area of Bangladesh: age-dependent and interactive effects. Journal of Biosocial Science 27(4): 393404. Sathar, Z. (1991). Changes in mortality in Pakistan 1960-88. Pakistan Development Review 30(4 Pt 2): 66976. Young, T. (2005). Population health: Concepts and methods, second edition. New York: Oxford University Press. Lee, K., and Mills, A. (1985). Developing countries, health and health economics. In The economics of health in developing countries, edited by K. Lee and A. Mills. New York: Oxford University Press.
21
22
United Nations Development Programme. (2009). Human Development Report 2009. Overcoming Barriers: Human mobility and development. New York: United Nations Development Programme. United Nations Inter-agency Group for Child Mortality Estimation (IGME). Data downloaded from childmortality.org on Sept 30, 2010. United Nations Population Division. 2011. World Population Prospects: The 2009 Revision. New York: United Nations. Simmons, R., and Phillips, J. (1992). The proximate operational determinants of fertility regulation behavior. In Family Planning Programmes and Fertility, edited by J. Phillips and J. Ross. New York: Oxford University Press. Abu-Ghaida and Klasen, 2004. Angeles, G., Guilkey, D., and Mroz, T. (2003). The effects of education and family planning programs on fertility in Indonesia. In MEASURE Evaluation Working Papers, edited for United States Agency for International Development (USAID). Chapel Hill: Carolina Population Center, University of North Carolina. GarcsPalacio, I. C., M. Altarac, and I. C. Scarinci. 2008. Contraceptive knowledge and use among low-income Hispanic immigrant women and non-Hispanic women. Contraception 77: 2705. Kelaher, M., D. Dunt, and S. Dodson. (2007). Unemployment, contraceptive behavior and reproductive outcomes among young Australian women. Health Policy 82: 95101. Laskar, M., Mahbub, M., Yokoyama, K., Inoue, M., and Harada, N. (2006). Factors associated with contraceptive practices of married women in Bangladesh with respect to their employment status. The European Journal of Contraception and Reproductive Health Care 11: 2207. Bongaarts, J., and Bruce, J. (1995). The causes of unmet need for contraception and the social content of services. Studies in Family Planning 26(2): 5775. Mekonnen, W., and Worku, A. (2011). Determinants of low family planning use and high unmet need in Butajira District, South Central Ethiopia. Reproductive Health 8:37. Available at http://www.reproductive-health-journal.com/content/8/1/37.
23
24 25
26
31
Endnotes
Cleland, J., and Wilson, C. (1987). Demand theories of the fertility transition: An iconoclastic view. Population Studies 41(1): 530. Bulatao, R., and Bos, E. (1992). The spread of fertility regulation as collective behavior. In Family Planning Programmes and Fertility, edited by J. Phillips and J. Ross. New York: Oxford University Press. Xu, Bentley & Kavanagh, 2011. DeRose, L., and Ezeh, A. (2009). Decision-making patterns and contraceptive use: Evidence from Uganda. Population Research and Policy Review 29: 42339. McNay, K., Arokiasamy, P., and Cassen, R. (2001). Why are uneducated women in India using contraception? A multilevel analysis. Population Studies 57: 2140. Moursund, A., and Kravdal, O. (2003). Individual and community effects of womens education and autonomy on contraceptive use in India. Population Studies 57: 285301. Becker, G., and Lewis, G. (1973). On the interaction between the quantity and quality of children. Journal of Political Economy 8(2): S279S288.
27
Barro, R., and Lee, J. (2010). A New Data Set of Educational Attainment in the World, 1950-2010. NBER Working Paper No. 15902. United Nations Development Programme, 2011. Kirk, D. and Pillet, B. (1998). Fertility levels, trends and differentials in sub-Saharan Africa in the 1980s and 1990s. Studies in Family Planning 29(1): 122. Cochrane, S. (1980). The effects of education on health. World Bank Working Paper No. 405. Washington, DC: World Bank. Colclough, C. (1982). The impact of primary schooling on economic development: A review of the evidence. World Development 10: 67195. See also FN 23.
28 29
30
Rosenzweig, M., and Schultz, T. (1982). Child mortality and fertility in Colombia: Individual and community effects. Health Policy and Education 2(3-4): 305348. McNay et al., 2001. Dreze, J., and Murthi, M. (2001). Fertility, education and development: Evidence from India. Population and Development Review 27(1): 3363. King, E. (1990). Educating girls and women: Investing in development. Washington, DC: World Bank. King, E., and Hill, M. (1991). Womens education in developing countries.Washington, DC: World Bank. Xu, Y., Bentley, R., and Kavanagh, A. (2011). Gender Equity and Contraceptive Use in China: An Ecological Analysis. Women & Health 51(8): 739758. For a discussion of behavioral economics and educational outcomes in developing countries, see also: Glennerster, R., and Kremer, M. (2011). Small changes, big results: Behavioral economics at work in poor countries. Boston Review 36(2): 1229.
31
32
Institute for Reproductive Health (2010). Standard Days Method: Building gender equity and engaging men in family planning. FAMProject Brief. Washington, DC: Institute for Reproductive Health, Georgetown University. Available at http://www.irh.org/?q=content/fam-promoting-gender-equity-family-planning. Shroff et al., 2009. Jejeebhoy and Sathar, 2001. Jejeebhoy, 1998. Hindin, 2003.
33
34
USAID | DELIVER PROJECT, Task Order 1. (2009). Contraceptive Security Index 2009: A Tool for Priority Setting and Planning. Arlington, VA: USAID | DELIVER PROJECT, Task Order 1. Selected Demographic and Health Surveys 20002009.
35
32
36 37 38 39
Barro and Lee, 2010. Selected Demographic and Health Surveys 20002009. USAID | DELIVER Project, 2009. Bongaarts, J. (1978). A framework for analyzing the proximate determinants of fertility. Population and Development Review 4(1): 105132. For a discussion of the degree of exposure to coitus among married women, its variation in sub-Saharan Africa, and implications for fertility modeling, see: Brown, M. (2000). Coitus, the proximate determinant of conception: Inter-country variance in sub-Saharan Africa. Journal of Biosocial Science 32(2): 145159. Also discussing the important role of young, unmarried women in fertility decline in sub-Saharan Africa: Kirk and Pillet, 1998.
40
Klasen, 2002. Magadi and Agwanda, 2008. Aryal, T. (2007). Age at first marriage in Nepal: Differentials and determinants. Journal of Biosocial Science 39(5): 693 706. Nguyen, L, Nguyen, M., Swenson, I., and Pham, B. (1993). Selected determinants of fertility in Vietnam: Age at marriage, marriage to firsh birth interval and age at first birth. Journal of Biosocial Science 25(3): 303310. Islam, M., and Ahmed A. (1998). Age at first marriage and its determinants in Bangladesh. Asia Pacific Population Journal 13(2): 7392. Sathar, Z., and Kiani, F. (1986). Delayed marriages in Pakistan. Pakistan Development Review 25(4): 535552.
41
Aryal, 2007. Martn, T. (1992). Delayed childbearing in contemporary Spain: Trends and differentials. European Journal of Population 8(3): 217246.
42 43 44 45
Selected Demographic and Health Surveys 20002009. UNESCO Institute for Statistics. Available at http://www.uis.unesco.org. Bongaarts, 1978. Govindasamy, P., M.K. Stewart, S.O. Rutstein, J. Ties Boerma, and A.E. Sommerfelt. (1993). High-Risk Births and Maternity Care. DHS Comparative Studies No. 8. Columbia, Maryland: Macro International Inc. Email communication from John Ross, July 16, 2005. The slope is the change in infant or child mortality rates due to a change in the percent of high-risk births. Moreland, S., and Talbird, S. (2006). Achieving the Millennium Development Goals: The contribution of fulfilling the unmet need for family planning. Washington, DC: Futures Group, USAID | Health Policy Project, Task Order 1.
46 47 48
Ross, J., and Stover, J. (2010). How Does Contraceptive Use Affect Infant and Child Mortality? PowerPoint Presentation, April 26, 2010.
50 51 52 53 54 55
49
United Nations Population Division, 2011. World Bank. (2008). Data downloads available at data.worldbank.org. World Bank, 2008. World Bank, 2008. Barro and Lee, 2010. Klasen, 2002. Taylor, A. (1998). On the costs of inward-looking development. Journal of Economic History 58(1): 128.
56
Bloom, D., and Williamson, J. (1998). Demographic transition and economic miracles in emerging Asia. World Bank Economic Review 12(3): 41955.
33
Endnotes
Deaton, A., and Paxson, C. (1997). The effects of economic and population growth on national savings and inequality. Demography 34(1): 97114. Raut, L. (1992). Demographic links to savings in life cycle models: Identification of issues for LDCs. Indian Economic Journal 40(1): 11638.
57 58 59 60 61
Klasen, 2002. World Bank, 2008. United Nations Development Programme, 2009. UNESCO Institute for Statistics. Available at http://www.uis.unesco.org. World Population Prospects, 2008.
62 Cellule de Planification et de Statistique du Ministre de la Sant (CPS/MS), Direction Nationale de la Statistique et de lInformatique du Ministre de lconomie, de lIndustrie et du Commerce (DNSI/MEIC), and Macro International Inc. (2007). Enqute Dmographique et de Sant du Mali 2006. Calverton, Maryland: CPS, DNSI, and Macro International Inc. 63 64 65 66 67
Female genital mutilation is the term used in Mali; it is the same as female genital cutting. We did not change sterility in this report, since this is rarely a policy variable. United Nations Population Division, 2011. Available for free download from http://www.futuresgroup.com/resources/software_models/spectrum. If the user wishes to change this to a longer or shorter period, the Excel model would need to be re-programmed.
34
Futures Group One Thomas Circle NW Suite 200 Washington, DC 20005

Women: Rapid

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Women: Rapid

Uploaded by

Copyright:

Available Formats

RAPID

A Technical Guide to the RAPID/Women Model

TECHNICAL APPROACH OF RAPID/WOMEN

Technical Approach of RAPID/Women

Figure 2a. RAPID/Women Model: Demographic Sub-Model

Contraceptive Security Access Index

Contraceptive Prevalence Rate

Total Fertility Rate

Technical Approach of RAPID/Women

Figure 2b. RAPID/Women Model: Economic Development Sub-Model and HDI

Gross Domestic Product

GDP per Capita

Human Development Index (HDI)

Female Life Expectancy at Birth

Technical Approach of RAPID/Women

Technical Approach of RAPID/Women

Table 3. Percentage of Women Married/in Union

Technical Approach of RAPID/Women

Sources: * Moreland and Talbird, 2006. 48 ** Ross and Stover, 2010 49

Table 5. GDP Production Function

SAMPLE RESULTS FROM AN APPLICATION IN MALI

Base scenario FP only Womencentered Combined

2010 2050 2050 2050

Sample Results from an Application in Mali

Table 9. Scenario Results of Demographic Indicators (Mali)

2010 2015 2020 2025 2030 2035 2040 2045 2050

84.8 84.8 84.8 84.8 84.8 84.8 84.8 84.8 84.8

84.8 82.9 81.1 79.4 77.8 76.3 74.8 73.4 72.1

84.8 82.9 81.1 79.4 77.8 76.3 74.8 73.4 72.1

8.2 9.8 11.6 13.4 15.4 17.6 19.8 22.2 24.6

8.2 9.7 10.9 12.2 13.3 14.4 15.5 16.6 17.7

8.2 11.6 15.5 19.9 25.1 30.9 37.5 44.9 53.3

6.6 6.3 6.1 5.8 5.6 5.3 5.1 4.8 4.6

6.6 6.3 6.1 5.9 5.7 5.5 5.3 5.2 5.0

6.6 6.1 5.6 5.0 4.5 3.9 3.5 2.9 2.4

Sample Results from an Application in Mali

2010 2015 2020 2025 2030 2035 2040 2045 2050

$252 $251 $253 $258 $272 $287 $298 $307 $317

$252 $295 $371 $464 $584 $711 $838 $966 $1,117

$252 $292 $377 $483 $ 637 $810 $995 $1,212 $1,491

$1,171 $1,174 $1,184 $1,205 $1,253 $1,303 $ 1,340 $1,371 $1,406

$1,171 $1,458 $1,851 $2,303 $2,843 $3,405 $3,970 $4,543 $5,183

$ 1,171 $1,448 $1,871 $2,363 $3,004 $3,700 $4,427 $ 5,241 $ 6,219

161 160 160 159 158 158 158 158 157

161 157 155 154 146 143 133 132 129

161 157 155 153 144 135 132 128 121

U5 Deaths Averted FP Only

Maternal Deaths Averted FP Only

USING THE RAPID/WOMEN COMPUTER MODEL

Using the Model

Scenario Name Base

Value in: 2010 2050

Expected Yrs Female 9 9

Using the RAPID/Women Computer Model

Figure 6. Policy Scenarios

Scenario Name Base WomenStrat+FP WomenStrat FP

Value in: 2010 2050 2050 2050 2050

Expected Yrs Female 9 9 11 11 9

Step 3: Project the Population

Using the RAPID/Women Computer Model

Figure 11: Sample Output

Base FP WomenStrat WomenStrat+ FP