Press "Enter" to skip to content

August 18, 2017 (links edition #22)

Of course, there is news on cash transfers. The flagship Indonesian CCT program, PKH, was estimated to increases vaccination rates among children by up to 30% — and even to 52% among children with less educated mothers. In Malawi, cash transfers have significant impacts on a range of schooling outcomes (courtesy of Dave Evans). Take up rates in the national CCT in Ecuador, Bono, are noteworthy (73% of eligible households in the poorest quintile enrolled to the program, and 65% of households in the bottom two), but a range of beneficiary high ccompliance costs, stigma, and distrust affect further progress. A cash transfer was among the interventions simulated to replace transport subsidies in urban Argentina, alongside an uncompensated subsidy removal and its substitution with two other subsidy schemes. Not easy task to tell the winner, with effects varying based on consumption (cash wins this battle), urban infrastructure, carbon dioxide emissions, and land and housing prices. Finally, here are the outcomes of a recent annual forum on humanitarian cash transfers.

If not in reading mood, a recent webstreamed discussion featured James Ferguson (Stanford) and Paul Niehaus (Give Directly) reflecting on the future of cash transfers in Africa (video is about 1h long). More to listen: you may recall earlier links on the ‘cash vs chickens’ debate (Gates vs Pritchett vs Blattman). The theme resurfaces in recorded interviews with two of the fighters, an EconTalk with Pritchett here and with Blattman here.

Facebook is increasingly finding its way into developmental issues: WFP introduced an artificial intelligence-based chatbot to monitor and communicate with safety net beneficiaries (see also here). In a new NBER paper, Bailey et al. measure ‘social connectedness’ based on the social network’s friendship patterns. They find that people with more geographically-dispersed friends are generally richer, more educated, more mobile, and have higher life expectancy. Talking of Zuckerberg, also Richard Branson has now joined the UBI bandwagon. Basic income was also the subject of a short blog of pros and cons, while there might be some relevant lessons for social assistance from the universalization of health.

Some fascinating fragility-related stuff: one includes a very insightful paper on state interventions in violent urban areas. There the authors (Blattman, Green, Ortega and Tobon) identified about 2,000 high-crime street segments in Bogota` and randomized them to eight months of increased security, municipal services, both, or neither. They find and discuss some interesting puzzles, like increasing public measures can reduce insecurity and other dimensions, but may have little impact on trust in the state. In another paper, a systematic review on aid and violence explores CCTs, public works and other interventions in 19 fragile countries point to the challenges from misappropriation of resources by violent actors, their sabotaging of projects, and the ensuing distrust between local population and governments. Bonus: we seldom talk of violence against elderly people (up to 83% of older people report being subject to some form of violence, according to a cross-country estimate). So here is an insightful commentary on the matter.

More on shocks and crises: an assessment of maternal health care (delivery services) in Liberia found that, during the Ebola crises, private clinics provided an important cushion to the overwhelmed government-supplied services. In line with previous research on the life-time effects of early shocks, Baez and colleagues document the socioeconomic outcomes of adult Mozambicans as influenced by weather shocks that occurred in childhood. Still on weather, a review summarizes the economic impacts of climate change.

A range of gender papers: a proposed framework sets out three critical dimensions of gender agency, namely goal-setting, perceived control and ability, and acting on goals, all three presented with an application to Sub-Saharan Africa; an article argues that low-pay jobs can be welfare-enhancing for women, but that possible negative intra-household effects are seldom assessed; in Nigeria, Croke et al show remarkable impacts of ICT training on women (especially among those ‘implicitly biased’ or who felt less need for such type of training); Erosa et al set out an elegant model looking at the substantial gender differences in the US labor market; finally, an interesting working paper traces the historical and multi-disciplinary origins of gender gaps.

More on history: a paper tracks the mediaeval roots of inclusive institutions, while a fascinating examination of the forces behind the industrial revolution points, as others have done in the past, to a watershed moment in explaining subsequent cross-country inequality and income divergence.

Some interesting research from the US: the ‘cash-out puzzle’ has been a hot theme in US safety net debates since the 1980s, that is, why can equal transfers in-kind and cash yield different outcomes. New research from Brown University adds fuel to the fire showing that, in line with several other studies, SNAP food vouchers lead to higher food consumption than cash transfers. In another paper using quantitative data from over 30 million college students, Chetty et al found substantial variation in rates of upward mobility across colleges (and argue that increasing low-income students access to colleges could increase the contribution of higher education to upward mobility). Interestingly, an article revisits the US 1996 welfare reform and assess its (sobering) legacy.

A wave of research on India: Dreze and Khera have a new paper examining the state of the evidence and debate around core social protection interventions, namely school meals, child care services, employment guarantee, food subsidies, and social security pensions. Gibson and colleagues argue that, and India’s current stage of development, growth of secondary towns may do more to reduce rural poverty than does big city growth. Finally, a paper measures various aspects of Indian urban migrant networks such as size, diversity, efficiency and experience, and estimates their effect on probability of employment.

Will megacities continue to grow? Apparently so, with a paper identifying a ‘megacity relationship’, i.e., countries will soon have 1 megacity for every 100 million people. And yet, we often don’t have a consistent metric to define what an ‘urban’ area really is – a dilemma that is in full display in Kenya.

A subregional analysis in China found that provinces pursue distinct approaches to social welfare and proposes a typology to capture such divergence (developmental, social-autocratic, and minimalist welfare regimes); another intra-national analysis in Senegal examines the encouraging poverty dynamics in the Senegal River Delta due to, among others, investments in large and small-scale agriculture. Talking of agriculture, a paper assesses the performance of ag input e-vouchers in Nigeria, an intervention with significant impacts but with more modest cost-benefit ratios.

Several technical pieces on poverty measurement and methods: similarly to the classic ‘Voices of the Poor’ exercise, Oxfam and partners produced Survey of the Poor, an initiative that tracked the lives over 1000 poor Indian households. Among the findings, women were twice as likely to living in temporary shelters or outside (without walls) than men. In a new article comparing composite and individual poverty measures, Azeem et al show that 18 percent households experiencing multi-dimensional deprivations are not captured by the one-dimensional measures. A report discusses the challenges associated with collecting time-use data methods, particularly stylized questions and time diaries. Finally, a paper revisits the debates around the use of ICP PPPs for poverty calculations and shows that alternatives (using HH expenditure surveys and different consumption bundles) yield similar results.

Some jobs materials: MIT’s Acemoglu weighs it with a new paper on automation and jobs (one more robot per 1000 workers would reduce the employment to population ratio by about 0.18-0.34 percentage points and wages by 0.25-0.5 percent). But is unemployment duration a sorting criterion in hiring? An IZA paper argues that hiring chances of the long-term unemployed turn out to be dominantly driven by the perception of longer unemployment spells as a signal of lower motivation. Firm formalization is viewed through the lens of two countries: one paper on Vietnam found that formalization leads to an additional increase in profits and value added; in another paper, Levy and Rodrik merge their informality and structural transformation narratives into what they call the ‘Mexican Paradox’ – among others, they put the spotlight on social policies that, especially social insurance, inadvertently subsidized low-productivity jobs and taxed formal employment.

To conclude, some fiscal issues: Kose et al have set up a handy database on fiscal space for 200 countries and 28 indicators over the period 1990–2016, including disaggregating the analysis over the past quarter century, during financial crises, and during oil price plunges. Another paper presents results from four simulations of the impact of potential tax reforms in Pakistan (including finding that corporate income tax has to be raised by 10 percentage points to generate an equivalent yield as a 1 percentage point increase in the sales tax). In Ghana and Uganda, a report challenges the conventional wisdom that citizens demand more accountability from governments for taxes than for revenues from oil or aid: their results suggest no differences among taxes, oil, and aid in citizens’ perceptions of transparency, misappropriation risk, or public goods provision. More on Uganda, with a UNRISD paper synthesizing a multi-year research project on the political economy of domestic resource mobilization.

Be First to Comment

Leave a Reply