Using a deductive search approach, peer-reviewed articles which assess the components that enable a favorable policy environment to harness the benefits of a demographic dividend were identified from PubMed and EconLit. Table 1 presents the search strategy employed for PubMed; a similar approach was followed for EconLit. The search was structured around three concepts, Economic Development, Fertility, and Sub-Saharan Africa. For the replicable search, we used the terms "economic growth" or "economic development" with the following combinations: "high fertility", "fertility reduction", "fertility decline", "demographic dividend", and "Sub-Saharan Africa".

The literature search returned 371 articles from PubMed and 351 articles from EconLit. Searches were conducted on 09/26/2018 on PubMed and on 10/01/2019 on EconLit. The search was then restricted to only include literature published in English in the last ten years, without any country restrictions. All articles were screened independently based on title and abstract examination. Disagreements in relevancy of content were solved jointly. Given that the search on EconLit was performed a year after the PubMed search, we conducted a second screening on EconLit articles based on the relevance for early and pre-dividend countries and on its "new" contribution of the articles not seen in PubMed. Table A-1 (see Extended data; Cardona et al., 2020) shows the list of articles that were found relevant in the first screening process, but later removed in the second screening process due to content overlap with PubMed search.

Following the identification of research articles from PubMed and EconLit, an inductive approach consisting of a "snowball search" was employed to expand the reference base further. Reference lists were reviewed to identify a set of landmark papers in the demographic dividend literature that were not listed in neither the PubMed nor EconLit results. This list is shown in Table A-2 (see Extended data). Additionally, reports and articles from gray literature sources such as the World Bank, the United Nations Population Fund (UNFPA), and the African Union, were downloaded and reviewed. The World Bank report on "Development Goals in an Era of Demographic Change" was the basis for all World Bank research included in this review ( Ahmed & Cruz, 2016; Ahmed et al., 2014; Ahmed et al., 2016; Cruz & Ahmed, 2018; Shekar et al., 2016).

Data were extracted using literature records following the checklist of items to include when reporting a systematic review, as suggested in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement ( Moher et al., 2009). The items included in the records were title, abstract, introduction (rationale and objectives), methods (information sources, data collection process, data items, analytical approach), results, remarks and future research, and relevance to policy.

Records were developed independently by each reviewer. Extracted data were color-coded into three categories—when possible. The three categories were: i) measurement; ii) policy interventions; and iii) key messages. Tables A-3 to A-6 (see Extended data; Cardona et al., 2020) synthesize the literature records, they report objectives of the study, location, methodology, and key findings. To address potential reviewer bias, the summaries presented in the Extended data ( Cardona et al., 2020) were conducted by a reviewer that did not participate in the data extraction process. These summaries were developed in parallel while the other reviewers were developing the framework.

Data from the literature records were used to develop the DD framework. We analyzed the color-coded categories to identify key concepts and trends between them to inform final conclusions on the main sectors and their components of the framework. For example, concepts such as fertility, contraception, contraceptive methods, family planning, pregnancy postponement, were classified into the family planning sector. The key sectors and interventions that we included for inclusion were based on its relevance to the demographic transition and its linkage to socio-economic growth and people’s welfare. The final sectors identified in the framework were jointly decided by both the data extraction team and the independent reviewer that developed the summaries presented in the Extended data ( Cardona et al., 2020).

The PRISMA Statement guided the reporting of results from the systematic review ( Moher et al., 2009). Figure 1 presents the flow of information through the different phases of the review. The initial search resulted in 371 articles from PubMed and 351 articles from EconLit. Language and time restrictions narrowed the search to 34 articles on PubMed and 216 articles on EconLit. In the eligibility phase, 23 papers from PubMed were found relevant to the association between fertility transition and socioeconomic growth and were included in the review. Of the 216 articles screened from EconLit, 46 papers were considered relevant to the demographic dividend literature in the first screening. In the second screening, 15 articles were removed, resulting in a sample of 30 articles relevant for countries with high fertility and positive projections of the working-age population. The inductive approach resulted in the identification of 25 articles. The final review consisted of 78 peer-reviewed articles, ten reports from the gray literature, and one book.

We developed a framework that identifies the main sectors needed to create a favorable policy environment to harness the benefits of a demographic dividend. The World Bank identifies four demographic dividend typologies based on demographic characteristics, similar to the demographic transition, and future development potential ( Ahmed et al., 2016). Low- and lower-middle-income countries with high fertility are characterized as pre- and early-dividend countries, while upper-middle and high-income countries make the late- and post-dividend typology. All sub-Saharan African countries are either in the pre-dividend or early-dividend typology ( WBG, 2016). A country is classified to be in a pre-dividend typology when the working-age population is projected to grow within the next 15 years, and when the total fertility rate is above four. Early-dividend countries follow a similar definition, except they have a total fertility rate below four.

The framework is organized into two stages, based on the demographic dividend typology, see Figure 2. The horizontal axis of the framework represents the stages of how countries transition from high fertility to lower fertility levels. Governance and Economic Institutions is a core sector for all countries, regardless of their demographic dividend typology, which is why it is located at the far left and is the origin of the arrows that go into the other sectors. The first stage highlights key strategic sectors that pre-dividend countries should prioritize to harness the benefits of a demographic dividend. These sectors are Family Planning, Maternal and Child Health, Education, and Women's Empowerment. In addition to these four sectors, the second stage of the framework adds the Labor Market. During the early-dividend countries, the labor market is needed to capture the large ‘youth bulge’ in the population pyramid. The labor market creates the potential to increase productivity levels among youth entering the labor market. The wheel of prosperity, Figure 3, showcases a set of sector-specific programs, interventions, and areas for investment, to catalyze and harness a demographic dividend.

Regardless of whether a country is classified as pre-or early-dividend, much of the literature agrees that countries can harness greater gains from a demographic dividend if certain preconditions are in place ( African Union Commission, 2017b; Bloom et al., 2014; Canning et al., 2015; Shekar et al., 2016; UNFPA, 2019). Because these are preconditions, they are a central piece to harnessing a demographic dividend, and thus located as a cross cutting sector in the framework. Changes in the population age structure are not guaranteed to produce economic growth; a favorable policy environment is required. Bloom & Canning (2008) highlight the case of Latin America, a region that missed the opportunity to exploit its demographic window of opportunity due to high inflation, political instability, adversarial labor relations, and an inward orientation with respect to trade.

Good governance and strong institutions can encourage civil participation and enable policies that will make a favorable policy environment across all economic sectors to benefit from the demographic dividend ( Bloom et al., 2014; Shekar et al., 2016; UNFPA, 2019). Political stability and the rule of law are part of good governance practices which help create a climate to attract foreign direct investments that will be catalyzed to job creation and development. Institutions have the capacity to support (or block) the development of policies to realize the growth potential created by the transition ( Guengant & May, 2013), and institutional quality is correlated with growth ( Bloom et al., 1999).

Financial institutions can promote a culture of savings and implement contractual savings to finance the education of children and youth ( NASEM, 2016; Ssewamala, 2015). In turn, savings can be transformed into domestic investments that will fuel growth and development, or serve as a cushion to overcome the lack or limited availability of safety nets during periods of unemployment and old age ( Canning et al., 2015; Guengant & May, 2013). Economic stability can also increase the desire to postpone having children as the opportunity cost of raising children increases ( NASEM, 2016). In Nicaragua, a poverty-reduction program reduced the hazard of a birth by 15%, with a higher impact in high parity women ( Todd et al., 2012).

The multi-sectoral complexity of the demographic dividend poses additional challenges to close the gap between policy design and policy implementation. Political commitment and strong leadership are needed to ensure coordination across all sectors and levels. Strengthening and developing monitoring and evaluation plans for program implementation is also required to hold institutions, programs, policies, and governments accountable ( Shekar et al., 2016; UNFPA, 2019). The African Union highlights the importance of improving and strengthening the civil and political space to ensure people's participation ( African Union Commission, 2017b). AU's report titled “Harnessing the Demographic Dividend through Investments in Youth” (2017b) acknowledges that "Africa needs to invest in ensuring participatory, representative and inclusive political processes as well as responsive state institutions… ensuring protection and respect for fundamental civil, political and socioeconomic rights of young people including young women".

Bloom & colleagues (2014) also highlight the importance of a carefully constructed trade policy. Evidence from East Asian countries shows that relatively open and competitive markets contributed to East Asia's financial success, while in Latin America, a closed economic policy limited poverty reduction efforts ( Bloom et al., 2009; Canning et al., 2015).

Family planning programs are highly effective at managing fertility levels to allow couples to attain their desired family size. In addition, they have the potential to improve social conditions in countries with historically high fertility levels ( Bloom et al., 2010; Eastwood & Lipton, 2011; Li et al., 2018). Among these social improvements, women can be more empowered to make their own decisions in the household and take ownership of their reproductive health. Focusing on fertility management, particularly in pre-dividend countries (countries with TFR ≥ 4), can accelerate a shifting age structure and pave the path to harness greater gains from a demographic dividend ( Ahmed et al., 2016; Bloom et al., 1999). In East Asia, around one-third of the economic growth between 1965 and 1990 was attributable to a rapid demographic transition triggered by the implementation of family planning programs ( Bloom et al., 1999; Bloom et al., 2014; Bloom & Canning, 2008; Bloom & Finlay, 2009; Bloom & Williamson, 1998). Population simulations for Nigeria project that a change in fertility from the UN medium-variant to the low-variant level results in an increase in income per capita of 5.6% in 20 years, and 11.9% in 50 years ( Ashraf et al., 2013).

Higher socioeconomic status (SES) is positively associated with declining fertility. As families become smaller, household members have greater resource allocation per capita, which, for children, can be translated into higher educational investments and higher parental time per child ( Ahmed et al., 2016; Becker & Murphy, 1988). There is a vast literature base on the relationship between high socioeconomic status and small family size; evidence comes from both developed and developing economies ( Komura, 2013; Majumder & Ram, 2015; May & Turbat, 2017; NASEM, 2016). Fertility reductions in Matlab, Bangladesh were linked to positive effects on health, earnings, and household assets of women and children ( Joshi & Schultz, 2007). In Colombia, Profamilia was able to reduce fertility rates among young aged women by between 11–12%, which resulted in an increase of 0.08 years in education attainment ( Miller, 2009). In a cross-country analysis, Lee & Mason (2010) estimated that the relationship between fertility and educational investments per child relative to income have an elasticity of -1; this means that a 10% fertility reduction can increase schooling per child by 10%. This notion has been partly attributed to the concept of quantity-quality trade-off in children ( Becker & Lewis, 1973; Behrman, 2015; Bloom & Canning, 2008; Eltigani, 2009; Majumder & Ram, 2015; NASEM, 2016; Sinding, 2009). The quantity-quality trade-off reflects a change in the behavior of the parental dyad as a result of the modernization of society. As societies transition, the cost of raising a child becomes increasingly more expensive. Therefore, parents prefer to have fewer children because they want to allocate more resources per child ( Ashraf et al., 2013; Diebolt & Perrin, 2013; Galor, 2012; Herzer et al., 2012; Notestein, 1983; Prettner et al., 2013).

Women perceive the benefits of a smaller family size across the wealth distribution differently; these perceptions are associated with fertility trends. Evidence from Bangladesh, India, Indonesia, Nepal, Philippines, and Vietnam has found that women from lower SES had a time lag in attaining their preferred fertility compared to women from higher SES ( Majumder & Ram, 2015). This gap is partially addressed by contraceptive pricing schemes, which aim to reduce costs to benefit low-income groups, but the inequity in access and use remains high across the wealth distribution ( Campbell et al., 2013; Canning et al., 2015; Lewis, 1986). Government emphasis on the benefits of small family size may accelerate the time lag across income strata. This idea has already been integrated into the national plans of a few developing nations, such as Indonesia and Iran ( McKelvey et al., 2012; Shekar et al., 2016).

The pace of fertility decline is subject to contraceptive practices and women's demand for contraception; however, overall demand is greatly affected by the supply of quality services and methods ( Fabic et al., 2015; Lewis, 1986; Mwaikambo et al., 2011; NASEM, 2016). Research has found that inconsistent contraceptive availability impacts women’s contraceptive uptake ( Bongaarts & Watkins, 1996; Gertler & Molyneaux, 1994). Indonesia's demographic transition highlights the importance of modern contraceptives for achieving fertility decline. Between 1982 and 1987, contraceptive use was the dominant factor that influenced fertility decline and, thus, its demographic transition ( Gertler & Molyneaux, 1994). Therefore, ensuring the affordability and availability of contraceptives is paramount to stimulating fertility decline and attaining desired fertility.

Social norms and cultural practices are important determinants for women’s reproductive outcomes and are important considerations for successfully achieving a demographic transition. Despite declines in recent years, polygamy is still a common practice in SSA ( Shekar et al., 2016). Women in polygamous relationships often time bear more children than women in monogamous relationships due to the idea that they are competing for their husband's resources ( NASEM, 2016). In many African societies, men tend to dominate decisions related to women's reproduction and contraception. Studies have shown that women are likely to obtain contraceptive methods in discreet ways, particularly unmarried women, increasing the covert use of contraceptives ( Campbell et al., 2013; Shekar et al., 2016). Another barrier to reducing fertility is the tradition and encouragement of large families. ( Shekar et al., 2016). In communities in the Democratic Republic of Congo, it is common to have a large number of children for their own and kin's security ( Romaniuk, 2011).

Understanding the social context of a community and the local perceptions of family planning can help guide the development of successful family planning programs to reduce national fertility rates. Lessons learned from other countries to accelerate the demographic transition highlight the need for effective programming. This can be achieved by working with and training local community health workers, midwives, and community change agents ( Bloom et al., 2014; Mwaikambo et al., 2011; Shekar et al., 2016). Evidence from high fertility regions has shown the power that social networks have in influencing fertility change through the exchange of information among individuals, or more recently, through social media. The Navrongo experiment in Ghana showed how integrating family planning at the community level was a key component to program success ( Bloom et al., 2014). In Kenya, Behrman et al. (2002) found that social networks can shape family planning attitudes and reproductive health behaviors. They found that the odds of using contraception increased by more than two if a woman had at least one family planning user in her network. In Ethiopia, Rwanda, Kenya, Nigeria, Tanzania, Senegal, India, Bangladesh, and Indonesia, community level task-shifting involving local community health workers, midwives, and community change agents have had positive effects on family planning ( Shekar et al., 2016).

Holistic family planning programs include both the provision of a range of contraceptive methods as well as comprehensive strategies to address cultural norms and beliefs related to reproductive health norms. Successful family planning programs combine initiatives in both the demand and supply side to ensure uninterrupted and equitable access to services ( May & Turbat, 2017; NASEM, 2016; Shekar et al., 2016). Governments play a key role in changing the general public's attitude towards family size and in ensuring the attainment of desired fertility by guaranteeing the availability of family planning information and services to all women of reproductive age.

Declines in infant and child mortality have a direct effect on fertility levels ( Eastwood & Lipton, 2011; Galor, 2012; Reher, 2011). Although the causal path can be argued, fertility decline could have preceded the decline in infant and child mortality ( Chowdhury, 1988; Galor, 2005). Some multi-country analyses suggest that causality runs from infant or child mortality to fertility, identifying that fertility decline ranges from 0.5 to 0.9 as a result of a fall of 50 in the infant and child mortality rate ( Conley et al., 2007; Eastwood & Lipton, 2011; Lorentzen et al., 2008). Reducing infant and child mortality, along with maternal mortality, is a global goal that is part of the Sustainable Development Goals and included in the national plans and policies of developing countries.

Although there is disagreement in the direction of the causal path, there is a consensus that the achievement of reduced child mortality, fertility decline, and lower maternal mortality are key to achieving fertility management and attaining a demographic dividend. Effective and equitable access to family planning affects desired fertility, children's survival, and maternal health. Access to contraception is estimated to have averted 32% of maternal deaths in 2008 in regions that have low contraceptive use ( Ahmed et al., 2012). Fertility reduction is also associated with a reduction in maternal deaths. Maternal mortality and parity have a J-shape relationship in developing countries ( Jain, 2011). The risk of mortality is higher during the first birth, mainly because the majority of births occur at younger ages. The risk decreases with preceding births and above the fourth birth, the risk increases again. The J-shape relationship is also associated with what is known as "too early, too late, or too frequent" ( Ahmed et al., 2012; Jain, 2011), in which women who give birth at short spacing intervals and those with multiple children are at increased risk of death. Empirical evidence has shown that eliminating births among high-risk groups could eliminate between 20% to 25% of all maternal deaths ( Winikoff & Sullivan, 1987).

Traditionally, families view children as essential labor resources that help with economic activities, such as agriculture, and as a form of security in old age ( Notestein, 1983). In countries with high child mortality, parents will desire to have more children to ensure protection against possible losses or to replace those losses ( Bloom & Canning, 2008; Conley et al., 2007; Majumder & Ram, 2015; Shekar et al., 2016). Fertility decline is one of the response mechanisms of declines in infant and child mortality. As societies transition into the modern world, child mortality is likely to decrease, and parents are likely to desire to have fewer children due to rising costs of raising a child ( Eastwood & Lipton, 2011).

There are multiple evidence-based interventions that can reduce infant, child, and maternal mortality. We summarize here some of the most compelling and relevant programming we found on our literature search.

Safe motherhood initiatives, vaccinations, child nutrition programs (including vitamin A), and case management of childhood illness in improving maternal and child health have been evidenced by several researchers and are considered as key interventions by governments ( African Union Commission, 2017a; Bongaarts & Casterline, 2013; Bongaarts & Watkins, 1996; Canning et al., 2015; Majumder & Ram, 2015; Shekar et al., 2016). In 1987, the Safe Motherhood Initiative was launched as a global campaign with the goal of reducing maternal mortality through interventions focused on making childbirth safer ( Ahmed et al., 2012; Starrs, 1997). Some of the most recognized interventions include deliveries with a skilled attendant at birth, improvements in the availability and quality of comprehensive emergency obstetric care, and expanded services for women during pregnancy, childbirth, and the immediate postnatal periods. In particular, those living with HIV to prevent mother-to-child transmission ( Canning et al., 2015; Wilson, 2015).

Compared to other regions, breastfeeding has been associated with longer birth intervals in Africa, and is associated with positive health gains for both mother and child ( Bongaarts & Casterline, 2013). Lactation increases the duration of postpartum amenorrhea, and long-term breastfeeding is consistently associated with long periods of postpartum amenorrhea, ovarian inactivity, and reduced fertility ( Majumder & Ram, 2015; Mwaikambo et al., 2011; NASEM, 2016). Postpartum infecundability is one of the close determinants of fertility and is a result of breastfeeding and postpartum abstinence ( Bongaarts, 2015). During the period of pre-transitional fertility, couples often do not consciously practice birth control to limit their children. Natural fertility is highest in populations with a low age at marriage and short durations of breastfeeding, which are largely determined by community customs ( Bongaarts & Watkins, 1996).

Vaccines are one of the most cost-effective health technologies to improve child survival ( Bloom et al., 2014; Bloom & Canning, 2008; Shekar et al., 2016). In SSA, vaccines have prevented a significant number of child deaths given that infectious diseases, such as malaria, are common ( Bloom et al., 2014; Eastwood & Lipton, 2011; Spence, 2009). Strengthening the community-based distribution of health technologies, such as vaccines, via existing health systems has been successful in the past and can accelerate the reduction of child deaths ( Shekar et al., 2016).

Under-five mortality reduction has a direct impact on increasing life expectancy at birth. The increase in life expectancy reflects the immense efforts in reducing mortality during the first five years of life ( Bloom & Canning, 2008; Reher, 2011). The overall increase in life expectancy at birth is a reflection of public health and medical interventions, such as access to clean water and sanitation and vaccines and antibiotics. Mortality decline has been faster on average across developing countries than in historical Europe; however, the world still has a set of 11 countries with a life expectancy at birth below the age of 51, six of which are located in Africa ( Cardona & Bishai, 2018).

Countries between the second and third stages of the demographic transition have the unique opportunity to boost their economies by taking advantage of the large "youth bulge" in their populations. Investing in education and tailoring education to the labor market can increase output per work, maximize productivity, and drive development

The African Union identifies education as one of the four pillars that underpin the roadmap to harnessing the demographic dividend in Africa ( African Union Commission, 2017a). Priorities identified by the union include improving access to quality education that matches labor market demands and embracing technical education to allow individuals to harness their natural abilities and create opportunities for self-employment ( African Union Commission, 2017b).

Forecasting simulations to 2030 found that in SSA, improvements in educational attainment can increase GDP per capita by 22.4% and reduce poverty by 38.9 percentage points, compared to the baseline scenario ( Ahmed et al., 2014). Cruz & Ahmed (2018) estimated that between 1950-2010, an increase of one percentage point in the working-age population boosted the GDP per capita growth rates by 1.1 to 2.0 percentage points, on average. Similar estimates were found by Bloom & Williamson (1998), with increases in GDP ranging between 1.4 and 2.0 percentage points. If the output per worker increases, the country's economy may grow remarkably ( Ahmed & Cruz, 2016; Eastwood & Lipton, 2011).

Individuals with higher levels of education are more efficient producers of health, which in turn is translated into a healthier population with higher productivity levels ( Grossman, 1972). Increased school enrollment in Malawi, Uganda, Ethiopia, and other African countries has had important implications for changing fertility rates, partly through changing desired fertility ( Behrman, 2015). Gertler & colleagues (1994) found that high levels of education in the 1970s in Indonesia, along with higher and equitable wages for both men and women, were significantly associated with reductions in the birth hazard. Women with an educational level beyond secondary school have access to higher wages, which increases the opportunity-cost between investing time in raising a child or spending time at work ( Behrman, 2015; Eastwood & Lipton, 2011; Galindev, 2011; Majumder & Ram, 2015).

Countries must work to ensure girls have access to education and are encouraged to stay in school. Novel approaches have been implemented across the region. The Kenyan Government changed the requirements of primary schooling from seven to eight years, which resulted in greater educational attainment among girls and reduced marital education gaps ( Shekar et al., 2016). The implementation of universal primary education in Nigeria resulted in a reduction of 0.26 births resulting from every additional year of schooling ( Osili & Long, 2008). Tanzania is now offering free primary and secondary education to all students in the country and is committed to investing more in the education sector ( African Union Commission, 2017b).

Furthermore, the quality of education has been shown to increase economic gains ( African Union Commission, 2017b; Shekar et al., 2016). According to the World Bank, one of the preconditions to catalyze a demographic dividend in Africa is to invest in the quality of education from primary, secondary, and tertiary level institutions ( Shekar et al., 2016). However, one of the challenges in tracking progress on education quality is the limited availability of information, not only in Africa, but amongst all developing economies.

Creating and capitalizing on the benefits of a demographic dividend requires empowering women and girls through improvements in health, education, and decision-making power ( Canning et al., 2015). Women who are empowered have the potential to transform countries' economies from a state of high fertility, low education, and sluggish economic growth towards a state of high education and fast economic growth ( Bloom et al., 2009; Canning et al., 2015; Prettner & Strulik, 2017). For example, one-third of the economic growth experienced by the Asian Tigers between the 1960s and 1990s has been attributed to improvements in women's empowerment and participation in the labor force as a result of a rapid fertility transition ( Bloom et al., 1999; Bloom et al., 2014; Bloom & Canning, 2008; Bloom & Finlay, 2009; Bloom & Williamson, 1998).

In traditional societies, married women are expected to bear children to fulfill their caregiver role and to ensure the reproduction of kin ( Folbre, 2001). Patriarchal societies tend to have a high fertility rate, while societies in which women are empowered tend to have a lower fertility rate ( Komura, 2013). Women with fewer childbearing responsibilities have more time to invest in the labor market and increase labor supply, which in turn increases household resources and welfare. This relationship is also inverse. Women who hold stable jobs and earn a high income are more likely to have fewer children because income losses during childbearing are higher for them than for women with low wages ( Ahmed & Cruz, 2016; Bloom et al., 2009; Canning et al., 2015; Eastwood & Lipton, 2011; Feyrer et al., 2008; Herzer et al., 2012).

Child marriage impedes on the rights, health, and wellbeing of young girls ( Chari et al., 2017). Child marriage results in higher rates of school drop-out, less decision-making power in sexual relationships, and higher fertility rates as a result of higher exposure to marriage ( Campbell et al., 2013; Canning et al., 2015; Shekar et al., 2016; UNFPA, 2017). Underemployment and unemployment is high among girls who marry early, resulting in an inability to earn wages and a cycle of poverty in future generations ( UNFPA, 2017). Increasing the median age at first marriage does not only reduce the exposure to pregnancy but also has the potential to increase the proportion of children who spend time in a two-parent family while growing up, which is beneficial for the long-term health of the child ( Cherlin, 1999). In Bangladesh, India, Indonesia, Nepal, Philippines, and Vietnam, the use of contraceptive methods and a higher median age of first marriage were the main drivers of fertility decline ( Majumder & Ram, 2015).

Family planning and education programs implemented at the community level have been shown to successfully delay the age of marriage in young girls; findings come from Ethiopia, Burkina Faso, and Yemen ( Engebretsen & Kaboré, 2011; Erulkar & Muthengi, 2009; Freij, 2010). These programs have shown that laws alone are not enough to delay marriage and prevent child marriage. For example, Indonesia experienced no change in the age of marriage after the National Marriage Act was passed in 1974 ( Cammack et al., 1996). Programs that tackle complex cultural and economic norms within the community are important for long-term, sustainable change.

Empowered women can be characterized by various attributes such as autonomy to make their own reproductive health decisions, economic self-reliance, and household decision making ( African Union Commission, 2017b). In addition to policies and initiatives to empower women, empowering women to the full extent requires a societal shift in gender norms. To protect women's autonomy, the African Union (2017b) recommends that countries should "adopt and enforce target[s] to end all forms of discrimination and violence against all women and girls, including domestic and sexual violence as well as harmful practices such as child and forced marriage and Female Genital Mutilation"; "strengthen domestic criminal accountability, responsiveness to victims and judicial capacity"; and "affirm the primacy of international and regional human rights law and constitutional laws over religious, customary and indigenous laws as a means to ensuring women's emancipation and autonomy."

Finally, women's and girls' education are a key component of women's empowerment and an important factor in the ongoing process of fertility transition ( Shapiro, 2012). Girls’ education does not only have implications on fertility outcomes - it also directly affects employability and age at marriage ( Majumder & Ram, 2015; Shapiro, 2012). Most theories predict that better-educated women desire fewer children than less-educated women. Other arguments for the impact of women's education on fertility outcomes is the role that income plays on families' economic prospects and the improved social status that comes with higher education. Promoting women's education also improves the welfare of both parents and children by increasing the transfer of knowledge between generations and further increasing the value of the next generation's human capital ( Kim, 2016).

Africa has a young population; according to UN projections, more than half of the population on the continent will be below the age of 25 by 2050 ( WPP, 2020). Thus, countries need productive and supportive labor markets to optimize their production capacity to absorb the growing young population ( Shekar et al., 2016). In addition, with declining fertility stimulating the demographic dividend, a "youth bulge" will be produced as a result of a large proportion of working youth compared to the number of dependent children ( Canning et al., 2015; Reher, 2011; Ssewamala, 2015). Particularly, for early dividend countries (countries with TFR < 4), the youth bulge can maximize the benefits from a demographic dividend through employment ( WBG, 2016). However, when the labor market is not able to satisfy the needs of this working-age population, the youth bulge can become a burden on society, and young adults are often forced into performing work within the informal sector ( Shekar et al., 2016; Turner, 2009). Similarly, the youth bulge can be pushed into the informal sector when they are undereducated and low-skilled, thereby entering the labor market unprepared ( Ssewamala, 2015).

Countries that are in the early-dividend typology, those with a total fertility rate above four children per woman and the share of the working-age population expected to grow, should focus their efforts on implementing policies and programs that benefit the labor market. However, this does not mean they have to stop investing in the five sectors prioritized in the pre-dividend era.

Projections show that in order to keep employment rates constant, the region needs to create 1.6 million jobs every month by 2030, and more than 2.0 million per month by 2050 ( NASEM, 2016). In 2015, the International Labor Organization estimated that 6 out of 100 people living in SSA were without work but available for and seeking employment. Countries should be able to create jobs at a rate that exceeds the rate of population growth in order to achieve full employment ( Ahmed et al., 2016).

There is a dearth of jobs within formal labor markets in low-resource settings, and jobs for youth and women are scarce ( Earl et al., 2017). Formal labor markets have to support the population through laws that protect both employees and employers ( Guengant & May, 2013). However, countries need to consider the risks and benefits of their labor policies. Bloom & colleagues (2003) found that rules governing the hiring and firing of workers can limit employers from taking risks; minimum wage above the market price can discourage hiring and training; and labor-management bargaining can trigger labor market inertia.

The benefits from a demographic dividend can be greater if young adults are able to accumulate human capital for when they enter the labor force and if the labor market is able to offer productive jobs ( African Union Commission, 2017b; Ahmed et al., 2016; Bloom et al., 2014; UNFPA, 2017). Nigeria is an interesting case study that experienced a mismatch between skill sets and labor demand. Unemployment rates were higher among young adults with secondary education than among young adults with primary education ( Reed & Mberu, 2014). However, Nigeria's economy has been slowly shifting from concentrating its labor force in agricultural activities to the services sector, which is associated with a more developed economy ( WBG, 2010). This shift has also occurred throughout SSA but at a slower pace. In the early 1990s, workers in the agricultural sector accounted for 63% of the labor force, which decreased to 55% by 2015, while workers in the services sector increased by six percentage points. Countries in SSA need to focus on building skill sets that match the demands of the labor market, which includes technical and vocational education and training ( African Union Commission, 2017a; Ahmed et al., 2016; Shekar et al., 2016; UNFPA, 2017). For example, Tanzania is currently implementing "The Education and Skills for Productive Jobs (ESPJ)" program to close the mismatch between labor demands, labor supply, and skills shortages in the job market ( African Union Commission, 2017b).

As mentioned in the previous section, women's participation in the labor market is crucial for greater gains in the economy, and it has a multiplier effect on human capital and population health. Countries must strive to create a favorable policy environment and provide incentives for female workers to find employment in the formal sector ( Bloom et al., 2010). Throughout the 1990s and 2000s, the participation of women in the labor market has been a great contributor in maximizing the benefits from a demographic dividend across different developing countries ( Ssewamala, 2015).

Participation in the formal labor market provides benefits beyond economic growth and development. Workers that are not part of the formal labor market are unlikely to have contributed to a retirement fund. When this population of informal workers reaches an old age and are no longer in conditions to work, they need to rely on their personal savings to live ( Dramani & Oga, 2017; Shekar et al., 2016). However, if they were unable to save money during working-age, they have to rely on family members or on the state– if the state has a welfare system in place.

When fertility levels are low, a low youth-dependency ratio can be harnessed through higher savings rates that have the capacity to boost a faster physical accumulation ( Ashraf et al., 2013; Eastwood & Lipton, 2011; Guengant & May, 2013; Turner, 2009). Not being able to save or have a pension becomes an incentive to have more children in which parents can rely during old age ( Jiang et al., 2013). Most countries in SSA present an unsustainable mix of poor saving rates and low capital productivity ( Eastwood & Lipton, 2011). The Asian Tigers' experience has shown that changes in the labor force per capita, in the savings rate, and in the investment rate are three mechanisms through which a changing age structure may affect the rate of economic growth ( Bloom & Williamson, 1998). It is estimated that in African countries with high fertility, a one percentage point increase in the working-age population share has the capacity to increase the savings share of GDP by 0.8 percentage points while reducing poverty by 0.76 percentage points ( Ahmed et al., 2016).

Financial markets that provide financial education and incentives to encourage household savings are important ( Ssewamala, 2015; Turner, 2009). Novel approaches to encourage savings with a special focus on the youth have promising results. A randomized control trial in Uganda found that adolescents that were given the opportunity to have a savings account were more likely to save for their education than those that did not receive a youth savings account ( Curley et al., 2010). In addition, participants became encouraged about their futures. Other examples from developing countries are Equity Bank in Kenya, Barclays Bank in Ghana, Khan Bank and Zoos Bank in Mongolia, and Banco Nacional de Bolivia in Bolivia ( Ssewamala, 2015). Most of these programs restrict withdrawal until reaching age 18 and provide incentives to save. Some of the incentives include double interest if no withdrawal in a quarter, free banker's checks to pay school fees, and matched savings with varying match rates.

The fourth stage of the demographic transition is characterized by low birth and death rates in which population growth is virtually zero or even negative ( Reher, 2011). During this period, there is no longer a youth bulge that can push the economy. In the third transitional stage, the living standards of the population tends to increase, and individuals are more likely to live longer. In return, people are more likely to increase their savings to ensure monetary security in the future ( Bloom & Canning, 2008). Both in India and China, longer lifespans have been associated with increased savings for retirement that were accompanied by favorable trade policies and favorable investment rates ( Bloom et al., 2010). The economic mechanisms by which people in different age groups produce, consume and save resources are of crucial importance for the demographic dividend to occur and for societies to accumulate generational wealth. Central to demographic dividend and generational economy are work, consumption, sharing and saving ( Lee & Mason, 2011).

Our review has unique strengths. First, to our knowledge, this the first systematic literature review in the demographic dividend literature that stratifies findings by demographic transition stage and demographic dividend typology to establish a multi-sectoral framework. Our review also identifies sector-specific programs and interventions. Second, the combination of literature from public health, economics, and gray literature provides holistic evidence for demographic dividends benefits beyond just economic benefits. Third, a hybrid between a deductive and inductive approach for identifying literature allowed for the inclusion of key landmark papers in the demographic dividend literature from leading institutions such as the World Bank, which are not necessarily published as peer-reviewed articles.

Our study has several limitations. First, we limited our public health search to PubMed and the economic search to EconLit. These databases are the largest for public health literature and economic literature, respectively, although the content of these databases may not be exhaustive for the existing literature about demographic dividend. Second, we cannot account for studies that are not available online that might be generated by NGOs or local country offices or governments. Thirdly, the review only included literature published in English, which may leave behind local documents. Finally, the rapidly changing drivers of global economies and challenges that did not exist when the reviewed literature was produced may limit the applicability of our findings in some contexts.

All data underlying the results are available as part of the article and no additional source data are required.

Harvard Dataverse, Gates Institute Demographic Dividend, Supplementary files - Generating and capitalizing on the demographic dividend potential in sub-Saharan Africa: a conceptual framework from a systematic literature review, https://doi.org/10.7910/DVN/F3WL99 ( Cardona et al., 2020).

This project contains the following extended data:

Harvard Dataverse: PRISMA checklist for “Generating and capitalizing on the demographic dividend potential in sub-Saharan Africa: a conceptual framework from a systematic literature review”. https://doi.org/10.7910/DVN/F3WL99 ( Cardona et al., 2020).

Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).