Types of Research
Access to financial services can enhance low-income households’ ability to invest in their livelihoods, safeguard their assets, manage risks, smooth income, and escape poverty (Cull, Ehrbeck & Holle, 2014; Dupas & Robinson, 2013). Yet worldwide 35% of men and 42% of women remain unbanked (Demirgüç-Kunt et al., 2015). Approximately half of the financially excluded population has access to a mobile phone, which may facilitate access to new digital financial services (DFS) such as mobile money (GSMA, 2015). Increasing access to and adoption of DFS-based technologies such as mobile money may therefore represent a significant opportunity for increasing financial inclusion. This potential may be especially great among the poor, and among rural, low-income women in particular, who in the past have been under-served by conventional bank-based financial services.
In this report we analyze three waves of the Intermedia Financial Inclusion Insights (FII) Survey, a nationally-representative household survey conducted in 2013/14, 2014, and 2015 in Kenya, Uganda, Tanzania, Nigeria, Pakistan, Bangladesh, India, and Indonesia. We run multiple logistic regressions using pooled survey data and country and wave fixed-effects to explore sociodemographic and economic factors associated with mobile money adoption, awareness, and use across countries and over time.
We find that rural respondents and respondents below the Progress out of Poverty Index (PPI) score have statistically significant negative associations with awareness and adoption for each survey wave and for most countries, consistent with less reach of mobile money in these populations. Indicators of education, including level of education acquired and literacy, have a positive association with awareness and adoption across countries and across survey waves. Phone ownership and having a bank account are consistently strongly associated with awareness, adoption, and use of mobile money; and the effects of these factors are the largest in terms of magnitude for mobile money adoption and use.
Women have consistently lower levels of awareness, adoption, and use than men across countries. Even controlling for other sociodemographic and economic factors, we find that being female is associated with an overall lower likelihood of awareness of mobile money. Across all countries, women who are aware of mobile money are also less likely to adopt mobile money than are aware men, with the negative effect of gender increasing in each wave. Gender does not have a significant association with the use of mobile money among those that have adopted it, however, which suggests that barriers to first-time use may be the most important for women’s access to mobile money.
These findings indicate that to realize the potential of DFS to reach currently unbanked populations and increase financial inclusion, particular attention needs to be paid to barriers faced by women in accessing mobile money. While policies and interventions to promote education, employment, phone ownership, and having a bank account may broadly help to increase mobile money adoption and use, potentially bringing in currently unbanked populations, specific policies targeting women may be needed to close current gender gaps.
Cash transfer programs are interventions that directly provide cash to target specific populations with the aim of reducing poverty and supporting a variety of development outcomes. Low- and middle-income countries have increasingly adopted cash transfer programs as central elements of their poverty reduction and social protection strategies. Bastagli et al. (2016) report that around 130 low- and middle-income countries have at least one UCT program, and 63 countries have at least one CCT program (up from 27 countries in 2008). Through a comprehensive review of literature, this report primarily considers the evidence of the long-term impacts of cash transfer programs in low- and lower middle-income countries. A review of 54 reviews that aggregate and summarize findings from multiple studies of cash transfer programs reveals largely positive evidence on long-term outcomes related to general health, reproductive health, nutrition, labor markets, poverty, and gender and intra-household dynamics, though findings vary by context and in many cases overall conclusions on the long-term impacts of cash transfers are mixed. In addition, evidence on long-term impacts for many outcome measures is limited, and few studies explicitly aim to measure long-term impacts distinctly from immediate or short-term impacts of cash transfers.
Although the programs reported on in the reviews include 37 cash transfer programs from Sub-Saharan Africa, 36 from Latin America, 18 from South Asia, ten from East Asia and the Pacific, three from Europe and Central Asia, five from North America, and four from the Middle East and North Africa, much of the evidence comes from multiple studies evaluating the most prominent cash transfer programs, especially Latin American programs. More evidence on the long-term impacts of cash transfer programs and on how these impacts might differ by context and program design will emerge as more of the existing global cash transfer programs mature and are evaluated. The available evidence indicates that cash transfer programs can be cost-effective, depending on the context and program design. Evidence of the cost-effectiveness, scalability, and sustainability of these programs is limited, but a few studies present initial findings on factors that appear to support or hinder cost-effectiveness, scalability, and sustainability. Several reviews suggest that design characteristics of cash transfer programs, including conditionality, targeting, and payment size, timing, frequency, and duration may all affect the impacts of the programs. Evidence on whether CCTs or UCTs are more effective at improving particular outcomes is mixed and may depend on the outcome measure of interest, but the reviews indicate that programs generally have greater impacts when targeting poorer or more marginalized populations and when providing transfers of larger size. Few reviews compare outcomes between cash transfer programs and other types of programs or interventions, though there is some limited evidence indicates that cash transfer programs might be more effective than alternatives to improve outcome measures related to poverty under specific circumstances.
An ongoing stream of EPAR research considers how public good characteristics of different types of research and development (R&D) and the motivations of different providers of R&D funding affect the relative advantages of alternative funding sources. For this project, we seek to summarize the key public good characteristics of R&D investment for agriculture in general and for different subsets of crops, and hypothesize how these characteristics might be expected to affect public, private, or philanthropic funders’ investment decisions.
Land tenure refers to a set of land rights and land governance institutions which can be informal (customary, traditional) or formal (legally recognized), that define relationships between people and land and natural resources (FAO, 2002). These land relationships may include, but are not limited to, rights to use land for cultivation and production, rights to control how land should be used including for cultivation, resource extraction, conservation, or construction, and rights to transfer – through sale, gift, or inheritance – those land use and control rights (FAO, 2002). Land tenure security – i.e., the level of confidence landholders have in their land rights – depends on the ability of informal and formal institutions to enforce those land rights and prevent others from challenging them (Feder & Feeny, 1991). In low and middle income countries land tenure security has been linked to improved land management including greater investments to improve land and agricultural productivity (Deininger & Jun, 2006; Deininger, Ali, & Alemu, 2011; Ali, Deininger, & Goldstein, 2014; Lawry et al., 2017). Having legal documentation in particular has been associated with a greater sense of ownership over land, increases in land productivity and capital investments associated with land, and in some cases additional financial opportunities such as access to credit for landholders with formal land titles (Deininger, Ali, & Alemu, 2011). But in spite of the widely recognized benefits of land tenure security more than 70 percent of the world’s population – and in particular many poor and vulnerable populations including ethnic minorities, smallholder farmers, and women – still lack access to formal systems to register their property and receive legally recognized land titles (Place, 2009; Enemark et al., 2014; Mitchell et al. 2016).
Previous EPAR research identified a high level of year-to-year change in crop portfolios by farmers, as well as large-magnitude changes in cultivated area, particularly for smallholders. This implies that farmers may be open to changes in crop mix influenced by development interventions targeting certain crops. Many agricultural development interventions focus on increasing land productivity, in particular relying on crop-focused strategies that emphasize raising yields of major cereal crops. Such interventions often assume that as farmers become more productive, they will specialize increasingly in their most productive crop, while relatively less productive farmers will shift to other crops, or shift out of farming into other rural employment, or migrate to look for urban employment. However, many smallholder farmers in Sub-Saharan Africa face constraints that also shape their livelihood decisions and may result in different farmer responses to land productivity changes than expected.
By examining how farmers respond to changes in crop yield, we provide evidence on how farmers are likely to respond to a yield-enhancing intervention that targets a single staple crop such as maize. Two alternate hypotheses we examine are: as yields increase, do farmers maintain output levels but change the output mix to switch into other crops or activities, or do they hold cultivated area constant to increase their total production quantity and therefore their own consumption or marketing of the crop? This exploratory data analysis is part of a long-term project examining the pathways between staple crop yield (a proxy for agricultural productivity) and poverty reduction in Sub-Saharan Africa.
We use panel data from three waves of the Tanzania National Panel Survey (TNPS) to investigate how Tanzanian farmers adjust their land allocation to maize following a change in their maize yield. We first evaluate whether households that experienced increasing maize yield from 2008-2010 (yield increasers) against those experiencing decreasing maize yield (yield decreasers) exhibit any shared baseline household, farm, or livelihood characteristics. We then use OLS and logistic regressions to analyse how much a change in maize yield within a given household is associated with changes in cropland allocation to maize versus other crops, controlling for household, farm, and geographic characteristics. Our results indicate that higher maize yields in 2010 are positively associated with increases in maize land allocation from 2010-2012 both as a land area value and when considering a household’s area share to maize, but that increases in maize yield from 2008-2010 are not significantly associated with maize planting decisions. A better understanding of the ways farmers respond to increases in maize yields will allow for better-informed and better-targeted investments in agriculture, and could inform ongoing structural transformation debates.
This research considers how public good characteristics of different types of research and development (R&D) and the motivations of different providers of R&D funding affect the relative advantages of alternative funding sources. We summarize the public good characteristics of R&D for agriculture in general and for commodity and subsistence crops in particular, as well as R&D for health in general and for neglected diseases in particular, with a focus on Sub-Saharan Africa and South Asia. Finally, we present rationales for which funders are predicted to fund which R&D types based on these funder and R&D characteristics. We then compile available statistics on funding for agricultural and health R&D from private, public and philanthropic sources, and compare trends in funding from these sources against expectations. We find private agricultural R&D spending focuses on commodity crops (as expected). However contrary to expectations we find public and philanthropic spending also goes largely towards these same crops rather than staples not targeted by private funds. For health R&D private funders similarly concentrate on diseases with higher potential financial returns. However unlike in agricultural R&D, in health R&D we observe some specialization across funders – especially for neglected diseases R&D - consistent with funders’ expected relative advantages.
The concept of global public goods represents a framework for organizing and financing international cooperation in global health research and development (R&D). Advances in scientific and clinical knowledge produced by biomedical R&D can be considered public goods insofar as they can be used repeatedly (non-rival consumption) and it is difficult or costly to exclude non-payers from gaining access (non-excludable). This paper considers the public good characteristics of biomedical R&D in global health and describes the theoretical and observed factors in the allocation R&D funding by public, private, and philanthropic sources.
We first conducted a literature review on factors theoretically associated with funding for early-stage biomedical research, including the expected correlates of funding levels for basic research, pre-clinical studies, and Phase I – IV clinical trials. To explore possible relationships between theorized drivers of R&D funding and actual funding flows, we analyzed evidence on how public, private, and philanthropic investments are affected by the public good characteristics of four high-burden diseases that disproportionately affect low- and middle-income countries: malaria, tuberculosis, hepatitis C, and soil-transmitted helminthiases.
Multiple factors influence R&D investment by public, private and philanthropic funders, including disease pathology and epidemiology, the current intervention landscape, policy and regulatory environment, and current and projected market conditions. Private sector investments are theorized to be primarily determined by opportunities for positive financial returns, while public and philanthropic investments may be motivated by a variety of social returns. We examine the specific funding decision factors identified in the literature for each of the four selected diseases.
Factors influencing the allocation of funds for biomedical R&D vary by disease, resulting in uneven funding across diseases. Due to issues of transparency and a lack of systematically collected data regarding R&D investments for diseases in low- and middle-income countries, especially from the private sector, these factors can be difficult to observe and measure. Furthermore, persistent data gaps can affect both aggregate investment and cooperative agreements.
A “new wave” of digital credit products has entered the digital financial services (DFS) market in recent years. These products differ from traditional credit by offering loans to borrowers that can be applied for, approved, and disbursed remotely (often without any brick-and-mortar infrastructure), automatically (generally minimizing or eliminating person-to-person interaction), and instantly (often in less than 72 hours). Digital credit also increasingly considers creditworthiness by using alternative (nontraditional) data—ranging from mobile phone activity to utility payments and social media data—potentially allowing for loans to populations previously unable to access bank credit. Two EPAR reports review the characteristics of digital credit offerings in India, Kenya, Nigeria, Tanzania, and Uganda, and regulations specific to digital credit in Africa and Asia.
A growing body of evidence suggests that empowering women may lead to economic benefits (The World Bank, 2011; Duflo, 2012; Kabeer & Natali, 2013). Little work, however, focuses specifically on the potential impacts of women’s empowerment in agricultural settings. Through a comprehensive review of literature this report considers how prioritizing women’s empowerment in agriculture might lead to economic benefits. With an intentionally narrow focus on economic empowerment, we draw on the Women’s Empowerment in Agriculture Index (WEAI)’s indicators of women’s empowerment in agriculture to consider the potential economic rewards to increasing women’s control over agricultural productive resources (including their own time and labor), over agricultural production decisions, and over agricultural income. While we recognize that there may be quantifiable benefits of improving women’s empowerment in and of itself, we focus on potential longer-term economic benefits of improvements in these empowerment measures.
We consider the case for spending the marginal dollar on empowering female farmers as a means of increasing household productivity, either prioritizing women for new investments or re-allocating existing resources. The literature suggests at least two distinct avenues via which economic benefits from investing in women’s empowerment in agriculture might arise. The first is by equalizing access to productive resources (including access to and control over land, labor, and other inputs) between men and women, and the second by leveraging differences between men and women that might lead to improved household outcomes. For the first avenue, we consider two theorized pathways to economic benefits from women’s empowerment in agriculture that posit reducing female farmers’ constraints would allow them to be as productive as equivalent male farmers. Pathway 1 focuses on empowering women through increasing their access to and control over agricultural inputs, thereby increasing overall agricultural productivity by reducing gender productivity gaps. Pathway 2 focuses on women’s control over their own time and labor, hypothesizing that removing constraints to women’s mobility would increase overall household labor productivity.
In the second avenue, we consider three further theorized pathways from increasing women’s decision-making power over agricultural decisions to economically beneficial individual and household outcomes, given assumed male-female differences in decision-making under similar circumstances. Pathway 3 connects differences in men’s and women’s decisions of what crops to grow with household nutrition outcomes. Pathway 4 hypothesizes that differences in plot management between men and women, specifically women’s greater likelihood of intercropping, influence farm soil quality and long-term household agricultural productivity. Finally, Pathway 5 draws a connection between differences in how men and women spend income from agriculture to impacts on household nutrition and education outcomes. We note that any measured benefits from leveraging male-female differences in the resource choices they make may dissipate as women gain more access and control if the differences are not due to being a woman per se, but rather stem from being disempowered - since this would change the circumstances in which evidence of these differences in decision-making have been observed.
This review of the literature ultimately shows some - but not conclusive - support for portions of all five theorized causal pathways between women’s economic empowerment in agriculture and economic returns. The literature also provides some dissenting evidence surrounding women’s constraints and preferences, most notably highlighting that results surrounding returns to empowerment can be context specific. We also note some inconsistencies in published methods and findings, and several key data gaps. First, published estimates of economic returns to empowering women in agriculture are still relatively rare, are mostly non-experimental, and are often limited in terms of data quality. Second, while published estimates provide some indication that, in many contexts, economic returns to women’s empowerment might be substantial, differences in measurement and reporting impede readily comparing benefits across contexts. Third, key variables necessary for extrapolating study findings to broader estimates of the benefits of economic empowerment – including basic variables such as land area managed by women – are not readily available. Finally, data on the costs of interventions addressing (eliminating or leveraging) the male-female differences in the five pathways are limited, making calculations of potential returns per dollar of investment difficult.
Previous research has shown that men and women, on average, have different risk attitudes and social preferences and may therefore see different value propositions in response to new economic opportunities. We use data from smallholder farm households in Mali to test whether risk perceptions differ by gender in this setting and across risk domains. We model the association between gender and perception or expression of concern across six risks (work injury, extreme weather, community relationships, debt, lack of buyers at market, and conflict) while controlling for demographic (age, education, health, wealth, time poverty, number of children) and attitudinal (social orientation, access to information, worldview, optimism, and beliefs about self-efficacy) characteristics. Factor analysis highlights extreme weather and conflict as eliciting the most distinct patterns of participant response. Regression analysis reveals an association between gender and risk perception, with women expressing more concern across all risks studied except for extreme weather. Also, we find lower risk perception associated with an individualistic and/or fatalistic worldview, a risk-seeking outlook, and optimism, while education, better health, a social orientation, self-efficacy and access to information are generally associated with more frequent worry – with some inconsistency for extreme weather, debt, and conflict risk. Further, income, wealth, and time poverty exhibit important, and complex, association patterns. Understanding if, and how, men’s and women’s risk preferences differ could help development organizations to shape interventions targeting women, to increase the likelihood of adoption, and to avoid inadvertently making certain sub-populations worse off by increasing the potential for negative outcomes.