Final Report Summary - PODER (Policy Design and Evaluation Research in Developing Countries) Policy Design and Evaluation Research in Developing Countries (PODER) was a four-year projectrunning from 1st September 2013 to 31st August 2017 involving 6 European Universities and 1 African University, as well as 6 Associate Partners. PODER trained a new generation of European doctoral students in the use of new techniques to design and evaluate policies to fight poverty. The techniques relied on controlled experiments as well as analysis of natural policy experiments and aimed to allow us not only to evaluate whether a policy has been successful but also to test the theoretical channels through which the policy is assumed to operate. The training combined coursework and methodological tools taught at the academic nodes with on-site training in project management and data collection learnt from private sector partners.The research projects of the recruited researchers focused on three distinct Work Packages:1. Health and Education Policy (WP1)2. Private Sector Development (WP2)3. Governance, Political Economy and Institutions (WP3)And the key achievements are listed by partner below. Bocconi UniversityFellows at Bocconi have worked on gender and intra-household allocation (WP1) as well as institutions and social norms (WP3). In WP1, topics researched include: (i) household education choices, gender bias and vulnerability in the Philippines, with emphasis on extreme outcomes such as trafficking; (ii) intergenerational spillovers of education; (iii) child nutrition and mortality; and (iv) the effect of asset transfer programs on marriage markets. Within WP3, three separate lines of research have been advanced. First, on the political economy of development in Brazil, including: (i) media and elite capture during the dictatorship; and (ii) the effect of teachers’ party affiliation on electoral outcomes. The second line of research is on attitudes towards immigrants and in particular: (i) the role of information provision on people’s support for immigration; and (ii) the effects of refugee resettlement on local social cohesion. Finally, the third line of research has focused on how informal institutions and customary norms interact with the process of development, and on possible policy responses in the face of harmful social norms. Bocconi hired 5 ESRs and 1 ER over the course of the project. It also hosted one of the network conferences.Paris School of Economics (PSE)Work carried out by PSE resulted in a body of research that contributes key insights for the design and evaluation of public policies in developing countries. For WP1, research relying on a variety of quantitative methods (RCTs, quasi-experiments, descriptive approaches) focussed on important aspects of early childhood education, nutrition and health, including questions regarding measurement of early childhood outcomes, the role of information and parents as inputs for education and the quality of health service provision. This work had wide geographical coverage from Sub-Saharan Africa, to the Middle East, Eastern Europe, Latin America and Asia. For WP2, the research focussed on several aspects related to firm development and productivity using data from China, Cote D’Ivoire and Mali, while other work focused on agricultural productivity in East Africa. Finally, work was also done on the Industrial Organization of the Aid Sector. In total PSE hired 6 ESRs for a total of 61.5 person months and 1 ER for 10 months. PSE also organized a summer school on “Recent Advances in Impact Evaluation for Development”, with lectures of 3 of the most recognized world experts in the field.London School of Economics (LSE)LSE made important contributions to the project, specifically under WP2, Private Sector Development. One fellow was hired to work on the project ‘Tackling youth unemployment and risky behaviours in developing countries’ working on the Punjab Economic Opportunities Program (PEOP). Another fellow worked on ‘Strengthening state capacity in developing countries and understanding the impact on the private sector’, looking at the impact of transparency laws on FDI inflows. A third fellow was employed for the ‘Understanding competition and the constraints on retail development in developing countries’ project, specifically studying the second-hand tyre market in Accra, Ghana. The final two fellows worked on ‘Promoting basic entrepreneurship among the ultra poor’ looking at empowerment and livelihoods for adolescent girls in Sierra Leone and microfinance in Pakistan. Overall LSE hired 5 ESRs for 49 months. LSE also hosted two of the Network Conferences joint with the CEPR annual conference on Development Economics.Stockholm University (SU)PODER fellows from the node at Stockholm University have made important contributions in all three work packages. Within WP1, one fellow worked on the project entitled “Effect of a micro entrepreneur-based community health delivery programme on child mortality in Uganda: a cluster-randomized controlled trial” with local faculty. The resulting paper, now under revision for the American Economic Journal: Applied Economics, shows that incentivizing community health workers led to a decrease in under-five mortality. Another fellow focused on the importance of social interactions within education in her project “Stereotypes and Self-Stereotypes: Evidence from Teachers' Gender Bias”, showing how teachers' gender stereotypes affect student math performance and their track choices. Together with faculty from the Bocconi node, she also studied the educational choices made by immigrant students within the project “Goal and Gaps: Educational Choices of Immigrant Children”. Within WP2, one fellow investigated the benefits of training farmers in using dairy technologies together with faculty from the PSE node, showing how different non-monetary incentive schemes affect the take-up of the dairy technologies. She also studied how poor Ugandan farmers’ investment decisions in cattle rearing are affected by the returns to their existing herd in the project “Investing in agriculture when it is worth it. Evidence from rural Uganda”. Finally, in WP3, one fellow has worked on a project that examines the re-election of women politicians in local governments in India. In the resulting paper, “The impact of gender quotas: Has India’s political reservation system improved the representation of women in local government?” she finds that mandated exposure to a woman mayor in a given election increases the probability that a woman is elected in the subsequent election, an effect only partly driven by the re-election of incumbent women. Another fellow has made two additional contributions under this work package. One project examines the effect of the 1994 genocide in Rwanda on schooling attainment. Another investigates the impact of resource inequality on ethnic conflicts across Africa. Overall SU hired 3 ESRs for 29 months and one ER for 12 months. University of Namur (FUNDP)Work carried out at the University of Namur (FUNDP) contributed essentially in training doctoral students to collect and process original data sets obtained through household surveys, and to develop, on the basis of these original data, research questions that fitted the work packages that were defined, as well as produce relevant research outcomes to guide policy. In particular, research done on WP2 showed the following: (a) In Sub Saharan Africa, even moderate processes of land titling can have a number of perverse effects, in particular with respect to woman access to land within the household; and (b) Traditional microfinance institutions, such as roscas, may serve as a substitute for insurance, and may help poor households to cope with sudden health emergencies. Within WP3, research showed that (a) in India, public policies aimed at increasing the basic income of the poor have an ambiguous impact on the possibility of civil conflicts, as indicated by the strong correlation between the Rural employment guarantee Scheme and the Maoist insurgency; (b) the Belgian colonial institution of “paysannat” in Congo had a persistent impact on local land institutions through its impact on the privatization of collective land; and (c) historical inequality has long lasting effects, even between ethnic groups located in the same country. This led to a negative and stable relationship between ethnic inequality and economic growth over the time period 1950-2010 for most Sub Saharan African countries. Namur hired 4 ESRs for 52 months and 1 ER for 12 months. The University of Namur also organized a summer school on ‘Big Data for Development’. Universitat Pompeu Fabra (UPF)UPF has contributed to the project primarily by training five fellows, for a total of 63.5 months, who had the opportunity to grow intellectually and to develop their research agenda. One fellow has now earned his PhD and is working on the impact of post-conflict reconstruction on insurgency. Another fellow has completed her PhD after spending about 1.5 years at UPF as a post-doctoral student is pursuing a research agenda with a focus on the recent immigration waves in Europe. A third fellow has been working on a number of public health projects in Africa. Much of her research explores sources and difficulties of measurement of key outcomes in public health research. She will start soon with a career in data science. A fourth fellow has been working on the importance of beliefs and expectations in the adoption of agricultural technologies and on the consequences for productivity of the adoption of different contractual forms in developing countries. The last fellow has been studying the importance of different mechanisms to monitor the performance of politicians and how this affects the quality of candidates running for office. In addition, UPF has also hosted a network conference on Development Economics in October 2015. University of Cape Town (UCT)Three fellows at the University of Cape Town worked on the project relating to “the transition from school to work” under WP1. One fellow wrote two papers with his local supervisor. The first investigated the reliability of responses from young South Africans to survey questions regarding their reservation wages. The second looked at the effect of neighbourhood on wage and school outcomes. Subsequent to his fellowship, this fellow was appointed as a lecturer at the University of Mannheim. Another fellow investigated the role of male relatives’ characteristics on male wages in South Africa. She found that father-in-law's schooling has a higher effect on male workers' labour market earnings than own father's schooling. In particular, the father's background plays a more important role in explaining labour market income of black South African males compared to white males, while this does not hold for the father-in-law's background. Another fellow focused his research on a large cash transfer programme reform in South Africa, and its effects on transitions in and out of employment for both direct and indirect recipients. This study analysed the impact of a decrease in the age threshold of the public pension system on the employment outcomes of the elderly and found large disincentive effects on employment, but concentrated exclusively on those jobs of the lowest quality. The inter-generational component of these transfers has also received extensive consideration, with a particular focus on whether these transfers could facilitate entry into self-employment for young unemployed individuals. A third fellow worked on "Asset-based Poverty traps”, in particular looking at ways of improving the utilisation of community assets and infrastructure as a means of poverty reduction. This fellow produced three working papers during her time at UCT and, after completing her PhD, accepted an Economist position with the World Bank's behavioural science unit. Two more fellows worked on the theme of ‘access to credit’ as part of WP2. One investigated the role of access to capital in allowing the unemployed to transition into informal sector employment. The other fellow investigated the role of the availability of credit, deriving from a closer social network, as a channel to explain the relation between ethnic diversity and labour market outcomes in South Africa. Overall UCT hired 5 ESRs and 1 ER.
