German Development Economics Conference 2025

12 Jun 2025, 14:00 → 13 Jun 2025, 18:30 Europe/Zurich

Goethe University Frankfurt

Andre Groeger (Autonomous University of Barcelona (UAB)), GDE 2025 - German Development Economics (University of Göttingen, Chair of International Economic Policy), Isabel Günther (ETH Zurich), Krisztina Kis-Katos (University of Göttingen, Chair of International Economic Policy), Martin Gassebner, Matthias Schündeln (GNOI), Michael Grimm (University of Passau), Renate Hartwig (Ruhr-Universität Bochum), Sarah Langlotz, Tobias Heidland (Kiel Institute for the World Economy and Kiel University)

The registration for GDEC 2025 is closed.

---

The annual conference brings together international scholars and researchers of development economics and neighboring fields. Plenary sessions with keynote speakers, parallel sessions with contributed papers, and poster sessions will reflect the current state of research in development economics and provide a forum for exchange for researchers and practitioners.

---

Special events:

• Keynotes: Abhijit Banerjee (MIT), Imran Rasul (University College London)
• Ceremony for the KfW Award for practice-oriented development research.

 

 

 

 

 

 

 

2025 Call for Papers GDE Frankfurt.pdf

2025 Program GDEC.pdf

Thursday 12 June

Registration

Keynote: Abhijit Banerjee (MIT)

16:00 Coffee Break

Parallel Session 1: Environment 1

1 How do reputation and gender differences affect contributions to environmental protection? Evidence from a lab-in-the-field in India

This study investigates whether women place more or less importance on their social image of being socially responsible than men in a controlled lab-in-the-field experiment conducted in rural India. Participants in the experiment perform a simple task to generate a monetary endowment, donate a contribution from this earned endowment in private to an environmental NGO, and finally elicit their preferences for public recognition of their contributions through an incentive-compatible mechanism. The baseline treatment is that participants are not able to choose the amount to donate but are obliged to contribute the entire monetary endowment earned to the NGO. They are also asked to state their WTP for public recognition. The findings indicate that, compared to men, women are more generous and are also willing to pay more to remain anonymous. However, we do not find any difference in WTP for public recognition between men and women. The mix of similarity and diversity in women's voluntary contributions to public goods and their concerns for social image shows that understanding women's behaviour is crucial for designing cost-effective norm-based policies.

Speaker: Souvik Datta (GNOI)

Reputation_and_gender-4.pdf

2 Top Down or Bottom Up? A Field Experiment on Public Goods Provision and Solid Waste Management

Bottom-up initiatives based on local capacity building have become a popular alternative to the traditional top-down provision of local public goods. This study compares the effectiveness of these two approaches. Based on a randomized controlled trial with 120 communities in rural El Salvador, we assess the impact of two interventions addressing solid waste contamination: (i) a top-down intervention where streets were cleaned by an external actor, and (ii) a bottom-up intervention where a facilitator raised awareness and mobilized for collective action. Using an objective measure of pollution based on geotagged photos and deep learning, we find large immediate effects for both interventions, with reductions in waste by 39 percent for the top-down intervention and 28 percent for the bottom-up intervention. Four months after the end of the project, these effects depreciated by 80 percent for the top-down and 60 percent for the bottom-up treatment. Our complementary data from 2,421 surveys and 883 activity records is consistent with a theoretical framework where many individuals are willing to contribute to public goods when others do, but fail to coordinate in the absence of a committed leader.

Speaker: Martina Jakob (University of Zurich)

Jakob-Coccia-Top-Down-or-Bottom-Up.pdf

Parallel Session 1: Gender 1

3 Spoils of colonization: The impact of British colonization on sexual prejudice

This paper tests the hypothesis that British colonial institutions promoted sexual prejudice—defined as negative attitudes toward sexual minorities—in postcolonial societies. We document five main findings. First, after accounting for differences in contemporary income per capita, OLS estimates from a cross-country sample of former European colonies reveal that former British colonies exhibit higher sexual prejudice than those of other European powers. Second, Geo-RDD estimates show that former British colonies have significantly greater sexual prejudice than former Portuguese colonies in Southeastern Africa, where local norms did not systematically condemn same-sex. Third, Geo-RDD estimates also indicate that former British and French colonies display similar levels of sexual prejudice in Western Africa, where religious norms systematically condemn same-sex acts. Fourth, additional evidence from areas in South America and Southeast Asia not characterized by homophobic social norms before colonization reinforces the external validity of our results on Southeastern Africa. Finally, mechanisms analysis suggests that the lasting presence of sodomy laws entirely accounts for the negative association between British colonial origin and contemporary sexual prejudice across countries. Overall, our results indicate that British colonial origin notably increased sexual prejudice in societies with social norms different from the penal codes imposed by colonizers.

Speaker: Alexsandros Cavgias (Ghent University)

4 Imperial Fault Lines: Colonial Legacy and Fertility in Sub-Saharan Africa

Colonization has profoundly impacted the economic and financial development trajectory in the colonized countries. However, previous work has not yet explored the impact of colonization on present-day fertility outcomes in Sub-Saharan Africa (SSA). This paper investigates the influence of colonial legacy on fertility through colonial legal institutions and the identity of the colonizer. Europeans differed not just in the legal institutions they transplanted in their colonies but also in their colonial administrative policies. I use the random splitting of ethnic homelands by the colonizers in 21 SSA countries as a natural experiment to identify the causal effect of colonial institutions and policies on current fertility in SSA. I find that the colonizer’s identity plays a major role in explaining current fertility outcomes. Countries colonized by Britain have nearly 19.3 percent lower fertility than those colonized by France. Lastly, I show that this differential persists through restrictive access to contraception in French colonies, a feature resulting from France’s colonial 1920 law, which sought to deter contraception use and promote fertility.

Speaker: Kritika Saxena (University of Groningen)

Colonial_Legacy_and_Fertility_in_SSA (1).pdf

Parallel Session 1: Labor 1

5 Formalizing employment in Africa’s small firms: Experimental evidence from Côte d‘Ivoire

Informal, low-quality employment in micro, small, and medium enterprises (MSMEs) remains a significant challenge in low- and middle-income countries. We present evidence from a consulting program in Côte d'Ivoire that advised firms on business management practices to show that a low-cost intervention can help improve job formality, without imposing undue financial burdens on firms. Using a randomized controlled trial with 448 MSMEs, we find that the intervention led to considerably higher overall employment formalization, driven in particular by greater reported minimum wage compliance and an increase in written contract provision. Drawing on a unique matched employer-employee dataset collected at three time points, we show suggestive evidence that these improvements were not driven by worker turnover or selection effects, but rather by employers' increased recognition of formalization's benefits. The intervention's financial implications were moderate, with evidence suggesting firms partially formalized previously informal payment streams, without a significant increase in total labor costs.

Speaker: Katharina Fietz (GIGA Institute | University of Göttingen)

Fietz_et_al_25_Formality_CIV.pdf

6 Supervision at Work: Evidence from a Field Experiment

How does supervision impact worker performance? This paper examines the effects of frontline supervision on the speed and quality of production in a Ugandan data-collection firm. We conducted a field experiment, varying the timing and intensity of supervision amongworkers, to study the effects of supervision onworker performance.We find that, following increased supervision, workers demonstrate improvements in the dimensions of performance where they were initially weakest. These effects persist, even on days, performance metrics, and tasks that are not directly supervised. Our findings suggest that supervision facilitates targeted on-thejob coaching of workers, which results in economically meaningful improvements in their performance. This underscores the significance of supervisors as conduits for the transmission of tacit production knowledge within organizations.

Speaker: Ritwika Sen (LMU)

Sen_Supervision_at_Work.pdf

Parallel Session 1: Measurement

7 Mapping the African Middle Class Across Space

This paper introduces a new method for estimating and mapping the size and spatial distribution of the middle class, leveraging geospatial data and machine learning techniques, and provides the first granular overview of the middle class in Africa. Existing middle-class definitions, based on income or expenditure thresholds and typically estimated using survey data, have never been available with geographical precision in developing contexts due to limited data. Using geolocated survey data as ground truth, we integrate different geospatial information—including infrastructure and nightlight data—mapped onto a hexagonal 96 km² grid (≈6km circumradius). Machine-learning algorithms enable extrapolation to regions without direct survey coverage.

Speaker: Mr Metiki Sekou (IFW)

Mapping_the_African_middle_class_across_space (2).pdf

8 Limitations of Satellite Data for Predicting Socioeconomic Outcomes and Policy Evaluation

Granular and repeated measurements of socio-economic outcomes are essential for understanding and improving economic livelihoods, but detailed survey data is scarce, especially across low-and-middle income countries. Recent machine learning approaches successfully leverage non-traditional data such as remote sensing data to predict consumption expenditures and asset wealth across locations but not their variation over time. This work systematically investigates the potential of public satellite imagery, combined with environmental and infrastructure data to predict levels and fluctuations in consumption expenditures and asset wealth and their use for policy evaluation. We present a framework that integrates spatial and temporal models to predict socioeconomic outcomes. Models trained on panel data from five African countries between 2007 and 2021 explain up to 73% of spatial variation but struggle to explain temporal variation (R$^2\approx$0.14). Utilizing ML-predicted outcomes, we are able to recover the sign, though not the magnitude of treatment estimates in an exemplary policy evaluation in Nigeria. These results underscore the potential of non-traditional data for policy evaluation while highlighting its current limitations.

Speaker: Mr Nathanael Schmidt-Ott (ETH Zürich)

Parallel Session 1: Microfinance

9 More loans, better skills, safer spaces: Evidence from clustering savings groups in Afghanistan

The Strengthening Women's Economic Empowerment Project (SWEEP) in Afghanistan was a US$2.74 million community-level intervention implemented by the Aga Khan Foundation (AKF) in cooperation with the World Bank in pilot areas of Afghanistan. SWEEP formed clusters of existing savings groups and provided capital to fund additional loans among cluster members. It further distributed business services, like business and soft skills training, mentoring, and access to markets and business networks. The project sought to increase and enhance the participation of women in income-generation and business activities, thereby fostering an improvement in overall well-being. SWEEP tackled various impediments hindering women in economic pursuits, such as restricted access to capital, skill training, and business networks. The project strategically build on AKF-established female-only Community Based Savings Groups (CBSGs) as its delivery platform. By strengthening female-only networks and deploying intervention components through these safe environments, the project addressed a spectrum of social and cultural barriers that had previously impeded the success of economic interventions targeting female employment and empowerment in societies characterized by extreme gender inequality, as exemplified by Afghanistan. This report presents the outcomes of the impact evaluation of SWEEP, employing a two-arm Cluster- Randomized Control Trial (C-RCT) design encompassing 2,164 individuals from 556 savings groups. To gauge the gender-specific efficacy of the project, the intervention was also implemented within male savings groups, encompassing approximately 20 percent of the studied savings groups and individuals.

Speaker: Juliane Zenker (Universität Göttingen)

10 Balancing the Burden: Shared Responsibility in Intergenerational Financial Support in Rural Ghana

Ensuring financial security for old age is a critical concern for rural smallholder farmers who mainly rely on intergenerational transfers. However, economic and demographic developments affect the intergenerational future transfers and expectations around them with potentially adverse consequences for private savings efforts. This study explores saving preferences of rural smallholder farmers in Ghana within the framework of an intergenerational contract, where children provide financial support for their parents’ old age. Using a lab-in-the-field experiment, we examine how parental savings and child transfers interact. Our findings reveal that high levels of parental savings persist regardless of anticipated child support. Similarly, child transfers remain consistent, uninfluenced by parental savings levels. We also find no evidence of strategic under-saving among parents or a crowding out of private savings by expected future transfers. Consistent with these results, we document shared responsibility of both parents and children for ensuring old-age financial security in an incentivized social norm elicitation.

Speaker: Kristina Czura (University of Groningen)

Shared Responsibility in Intergenerational Financial Support in Rural Ghana - with author details.pdf

Poster session

11 Are FDI restrictions inducing international migration? Evidence from Indonesia

Using disaggregated dyadic panel data on international migration flows
from Indonesian districts, this paper provides causal evidence for Foreign Direct Investment (FDI) and international migration being substitutes. Our empirical analysis exploits regulatory changes in the Negative Investment List, a product-specific FDI policy, that have been implemented by the Indonesian government from 2007 to 2014. Using a shift-share approach to measure the district-country-pair-level exposure to the sector-specific tightening of FDI regulation, we analyze the impact of changes in bilateral FDI inflows on international migratory movements of the population living in FDI receiving areas. We document that relative reductions in dyadic FDI inflows in the aftermath of the reform triggered an increase in emigration to the investor countries, especially among those with tertiary education.

Speaker: Anna Katharina Gasten (University of Göttingen)

12 Border crossings between 243 countries: The Global Transnational Mobility Dataset 2.0, 1995-2022

This paper introduces a new dataset that estimates the volume of human travel
across country borders worldwide between 1995 and 2022. It builds and expands
on pioneering work that presented estimates for 2011 to 2016 (Recchi et al., 2019).
The dataset enables the study of the volume, directions, and changes in global
human mobility. Our estimates reveal that total transnational mobility increased
from 4.87 billion trips in 1995 to 9.64 billion in 2019, largely outpacing global
population growth. Across the board, international migration constitutes a tiny
fraction of transnational travel (less than 1% worldwide and as low as .15% in
Europe). The rise of transnational mobility has been particularly sustained in East
and South-East Asia. This region was, however, also the hardest hit by Covid-19
travel restrictions and their aftermath, which brought its flows in 2022 back to
mid-1990s levels. Most border crossings are intra-regional, especially in Europe.
Despite the widespread growth in volume, the global network of cross-border
mobility has not significantly changed its overall configuration around nine major
clusters in more than a quarter of a century. Germany stands out as the main hub
in Europe and globally, followed by the US and China. However, some regional
mobility clusters have split and others have merged, with individual countries
shifting between clusters. The dataset may be used to study global-level
phenomena in fields such as migration and tourism studies, sustainability,
epidemiology, international economics, and international relations.

Speaker: Dr Tobias Grohmann (European University Institute (EUI))

104.0 GTMD2 Paper_20250206_GDE2025 submission.pdf

13 Burden your neighbor? Distributional effects of the EU CBAM on EU trading partners

This study examines the distributional impacts of the European Union’s Carbon Border Adjustment Mechanism (EU CBAM) on Turkish households. Using a global quantitative trade model linked to micro-level household data, we assess real income changes across various policy scenarios and sectoral labor mobility assumptions. While our findings indicate modest real income losses at the aggregate level, we uncover substantial heterogeneity across households driven by sector-specific income shocks and mediated by labor mobility. In the absence of labor mobility, 20% of Turkish households experience real income losses, while 80% benefit slightly due to price decreases. Removing labor frictions, all households face real income losses, albeit with reduced inter-household disparities. We further explore the effects of a Turkish policy response involving domestic carbon pricing and a Turkish CBAM. We show that public revenues are sufficiently large to offset negative impacts on households in the short run, but not in the long run. The study offers important insights into the feasibility and political constraints of potential policy responses by EU trading partners.

Speaker: Johannes Gallé (GNOI)

14 Do this or do that?

This paper aims to fill the methodological gap in development economics that until now there exists no quantitative tool that allows to prioritize reforms in a systematic nor optimal way. Following the recent debate on the issues Randomized Control Trials (RCTs) have with establishing external validity and general equilibrium effects, this paper proposes a micro-founded Growth Diagnostics framework to consider general equilibrium effects and prioritize policy prescriptions. Building conceptually on Hausmann et al. (2005), we set up two continous-time Overlapping Generations (OLG) models to account for the different net-marginal valuations of various economic activities rigorously. We solve the household and planner problem to respectively obtain the private and social net-marginal valuations of economic activities via the corresponding co-state variables. With these in hand, we define the wedges in the net-marginal private and social valuations to set up a new planner problem (we call super policy maker problem), where the planner minimizes the sum of wedges. We illustrate our framework with an application to the literature on structural change. The final wrapping optimization problem allows to prioritize optimally economic reforms in a second-best framework, thus, to put it in the words of Rodrik (2010), to first diagnose before one prescribes the remedy.

Speaker: Mr Hannes Tepper (Paris School of Economics)

15 Does Project Performance Matter? The Association Between World Bank Project Ratings and Local Development

The past decade saw an increase in publicly available evaluation ratings of official development assistance (ODA) projects. This study critically examines the prevailing assumption that these outcome ratings accurately reflect the aid project’s effectiveness. Drawing on the increasing criticism – referring to evaluator responsiveness to stakeholder pressure and trends towards easier achievable and measurable outputs – we argue that evaluation ratings are unsuitable proxies for real-life aid outcomes, even under strict geographic disaggregation. We test this argument by comparing the impact of differently rated growth-centered World Bank projects – successful and unsuccessful – on local economic development in Africa between 1995 and 2014. The research design combines information on economic aid projects, evaluation ratings, and sub-national gdp (the gross cell product) on the highly disaggregated 50 x 50 km grid cell – year level. The difference-in-difference estimations reveal that economic projects have a positive and significant impact on gcp. Projects evaluated successful, however, do not have a greater impact than projects evaluated unsuccessful. This implies that there is a disconnect between evaluation ratings and other micro-level outcome measures, which calls into question the utility of current evaluation practices.

Speaker: Sebastian Korb (Zeppelin Universität Friedrichshafen)

HoffmannKorb_DoesProjectRatingMatter_GDEC.pdf

16 Does subsidized adult labor reduce child labor in cacao: Evidence from Ghana

It is estimated that one in two children in cocoa-growing communities is involved in child labor, threatening their education, health and overall well-being. While previous interventions to reduce child labor have focused on improving school attendance or enforcing labor regulations, no study has directly addressed the underlying economic constraints that may drive child labor in cocoa-producing communities: financial constraints of households to pay for adult labor during the labor-intensive cocoa harvest season. This study aims to fill this gap by analyzing whether fully subsidized provision of labor to cocoa farms can serve as an effective strategy to reduce child labor. We conducted a cluster-randomized controlled trial (RCT) in 55 villages in the cocoa-growing regions of Ghana, comprising 1,171 households. Households in treatment villages were provided with adult labor to assist with cutting cocoa pods from trees, breaking pods and transporting cocoa beans to the farm. These are activities that typically involve child labor in these areas. The policy led to a 9.7 percentage points decrease in child labor in cacao from a baseline incidence of 37%. A broader measure of child labor, including all child labor activities, showed a reduction of 10.6 percentage points from an initial prevalence of 66%. These results suggest that financial constraints drive child labor in these communities, and that providing fully subsidized adult labor during the critical cocoa harvest period can address this root cause of child labor. Our findings also suggest that providing financial support to employ adult labor is more effective and efficient than most previous interventions to reduce child labor in cocoa regions.

Speaker: Jan Averkin (ETH Zurich)

Child_labor_Averkin_Gunther_Asiedu_GDE_draft.pdf

17 Experiential Learning to support community-based Climate Change Adaptation in the Zambezi Region of Namibia

Rural communities worldwide face rapidly evolving and complex challenges, such as climate change. The current generation is challenged to balance its own needs with those of future generations and must decide how much to invest in adapting to climate change to secure the future. This highlights the key intergenerational dilemma at the heart of efforts to adapt to climate change. Addressing these challenges requires collective governance and innovative tools to enhance the same. This study examines whether a game combining the mechanisms of public good and dictator games framed as an intergenerational dilemma, along with a visioning workshop, can influence rural community members' perceptions of climate change, future generations, and self-efficacy. In the game, participants faced decisions that involved balancing resource use between immediate needs and future generations, simulating the trade-offs inherent in climate adaptation. The visioning workshop then facilitated discussions about long-term adaptation strategies.

Our intervention took place in 26 villages in the Zambezi region of Namibia, where participants were interviewed before or after the intervention. We conducted a follow-up survey eight months later to evaluate the persistence of the intervention’s effects Outcomes were analyzed using ordinary least squares regression, employing a between-subject design to assess short-term learning effects and a within-subject design for follow-up outcomes. Results indicate that participation in the intervention significantly increased self-efficacy, awareness of climate risks, and demand for climate change-related information. While pre-intervention participants displayed strong concern for long-term issues and future generations, the intervention broadened their focus to include both short- and long-term considerations, as well as a greater sense of responsibility toward both current and future generations. Eight months after the intervention, the observed gains in self-efficacy were not sustained. However, while risk awareness had diminished, the demand for climate-related information had increased. These findings affirm the value of experiential learning in strengthening cognitive outcomes that drive collective action, while suggesting that long-term behavioral change requires sustained engagement.

Speaker: Franziska Auch (GNOI)

Experiential learning_FA.pdf

18 From Villain to Hero: Using Aspirational Narratives to Foster Technology Adoption Among Artisanal Gold Miners in the Peruvian Amazon

Narratives shape economic and political decision-making processes, yet understanding their causal impact remains a challenge. This paper investigates the influence of narratives on economic choices, in the context of environmentally sustainable mining in the Peruvian Amazon. We focus on how weaving information on clean technology into narratives affects artisanal miners’ decisions. The study compares miners receiving information embedded in an aspirational narrative against a control group that receives information in a factual manner. Our study aims to provide valuable insights into the effectiveness of narratives in policy-relevant settings.

Speaker: Arne Weiss (University of Alicante & University of Bern)

19 Multi-Unit Willingness to Pay Elicitation in the Field

We present and field-test a novel mechanism to elicit the willingness to pay for multiple instead of a single unit of a good. At each price of the canonical multiple price list (MPL) approach, we elicit varying multi-unit demand instead of limiting the choice to two options. We showcase our mechanism by applying it in the Lake Victoria fisheries of Tanzania to elicit valuations for fishing net panels, a production input generally purchased in bulk. Our application demonstrates the mechanism in a challenging field environment and gives best practice advice on minimizing common weaknesses of the MPL approach.

Speaker: Tillmann Eymess (IIES, Stockholm University)

muMPL (Diekert et al).pdf

20 My opinion, your opinion – Do group norms and perceptions influence farmers’ fertilizer practices?

Social norms and perceptions within farming networks can influence the adoption of new agricultural practices. In Indonesian rice farming communities, norms around the desired level of rice plant greenness are widespread, with some farmers valuing deep green plants. Since greenness levels depend on the content of chlorophyll in the plants, which in turn depends on nitrogen fertilizer inputs, these norms can lead to high usage of chemical fertilizer. This study uses a mixed-method approach to examine whether social norms, personal beliefs, and perceptions about peers’ opinions influence rice farmers’ fertilizer input decisions. We combine quantitative regression analyses with qualitative content analysis to explore these dynamics. Our findings show that farmers who are unaware of a saturation point for fertilizer application tend to use more chemical nitrogen and less organic fertilizer. These farmers are also less willing to experiment with new farming practices that might reduce plant greenness but improve soil health. However, second-order perceptions – beliefs about whether lower greenness levels lead to talking within the farming community – do not significantly affect fertilizer use or farmers’ willingness to try new methods. A survey experiment further confirms that increasing the salience of potential talking has little effect on farmers’ willingness to experiment with new practices. Dyadic regressions reveal that actual fertilizer adoption behaviors of neighboring farmers are more predictive of fertilizer input decisions than neighbors’ greenness norms. This suggests that while social norms around plant appearance exist, farmers’ decisions are more strongly influenced by their own knowledge and the observable actions of their peers.

Speaker: Udit Sawhney (University of Passau)

My_opinion_your_opinion_Sawhney.pdf

21 Pork Barrel Politics in Chinese Aid Allocation to Africa?

This study investigates the potential favoritism in Chinese development aid allocation towards the home countries of African Union (AU) Chairpersons during their tenure. The AU’s rotational presidency model provides a unique opportunity to test whether Chinese aid is disproportionately directed to the Chairperson’s home country during their one-year term. Estimating fixed effects Poisson Pseudo-Maximum Likelihood (PPML) models for 54 countries in the 2002-2017 period, we find a significant increase in Chinese aid to the Chairperson’s home country during their tenure, particularly in the economic sector. No similar pattern of favoritism is observed for aid from the World Bank and traditional Western bilateral donors. This study contributes to the literature on foreign aid by focusing on China’s strategic interactions with the AU, highlighting a distinct approach to aid allocation that favors influential positions within less Western-dominated international organizations.

Speaker: Lennart Kaplan (GAU)

22 Poverty, Inequality, and Working Hours: Evidence on Veblen Effects from Labour Markets in India

We study the labour market impact of the “Veblen effect” by estimating the relation between inequality and average work hours in society and how the effect varies between the poor and the nonpoor. Using data on actual hours worked by members of a household from India’s Periodic Labour Force Survey, we report three main findings. First, members belonging to absolutely and relatively poor households work for fewer hours than those in non-poor households. We observe that absolutely poor (relatively poor) households work for almost 45 (22) minutes less per day than the average working hours of non-poor households. Second, the impact of inequality (measured by a 90/50 percentile ratio) on average work hours is positive, large, and statistically significant. A one standard deviation increase in the regional 90/50 percentile ratio is associated with a 2.8 percent increase in the average work hours of a non-poor household. However, the impact of increased inequality on the increase in the work hours of absolutely poor households is smaller. Overall, higher inequalities induce poor households to work more by 1.68 percent. Third, we also find evidence that poor households do show more willingness to work for additional hours than non-poor households.

Speaker: Akansh Khandelwal (Indian Institute of Management, Indore)

Manuscript_Feb15.pdf

23 Regional South-South Trade and the Dutch Disease: The Case of Latin American Manufacturing Exports

This paper introduces two channels through which exports from commodity-dependent countries towards regional partners might be less affected by Dutch Disease effects than extra-regional exports. The first channel relates to a higher share of technologically more sophisticated products in intra-regional South-South trade, which are less sensitive to cost and price changes. The second channel is related to trade barriers and entry costs faced by extra-regional competitors in the regional market. The two channels are empirically tested through a panel data analysis of manufacturing exports from Latin American countries between 1996 and 2018. The evolution of exports to regional export partners is compared to extra-regional exports. Dutch Disease effects are most pronounced in exports to extra-regional partners, where a one-percent increase in commodity prices leads to a 0.48% decline in manufacturing exports, significantly larger than the 0.31% decline in regional trade. The effect is mainly driven by low-tech exports, which are more negatively affected than medium- and high-tech exports, with an elasticity of -0.95% in extra-regional trade compared to -0.58% in regional trade. The results support both channels, suggesting that technological upgrading and regional trade integration can mitigate the contraction of the manufacturing sector during commodity price booms.

Speaker: Martin Middelanis (Freie Universität Berlin)

DD_and_trade_partners_GDEC.pdf

24 Social pensions in Bangladesh: Do they improve well-being and trust in local governments?

Societies in South Asia are ageing quickly and only few elderly receive contributory pensions, highlighting the importance of social pension systems. However, evidence on the impact of social pension receipt on the beneficiaries' well-being and their trust in local governments is scarce. This paper uses data on eligibility criteria of several thousand beneficiaries and non-beneficiaries of Bangladesh's Old Age Allowance, which was collected before inclusion into the program, to estimate the impact of the social pension on consumption, mental and physical health, and trust in local government. We use a combination of coarsened exact matching and nearest neighbor matching and find that, on average, social pension receipt had no significant effect on consumption, physical or mental health, but a large positive impact on trust in local governments. The lack of significant impacts on welfare may be related to ineffective targeting approaches that benefit less poor individuals and households for whom the amount of the social pension makes a relatively small difference.

Speakers: Mr Heiner Salomon, Sebastian Fehrler

social_pensions_fehrler.pdf

25 Socially disadvantaged ethnic groups and distributive politics

This paper examines how initial income endowments influence the distribution of public goods through ethnic favouritism in Bolivia. We argue that economic advantages, such as income, sustain favouritism among co-ethnics. To guide our analysis towards this idea, we develop a theoretical framework of favour exchange, which predicts that higher-income groups are better positioned to sustain favouritism by using their economic advantages to enforce mutual cooperation—for example, rewarding loyalty with access to resources or punishing defection by exclusion from future benefits. Using individual-ethnic level data on public goods access before and after Evo Morales’s presidency (2005–2012)—Bolivia’s first Indigenous president—and implementing a difference-in-differences strategy that exploits variation in ethnicity, time, and political support, we document two key findings. First, Morales’s co-ethnics (Aymaras) experienced greater access to public goods in provinces where his party was politically dominant. Second, consistent with the proposed mechanism, we provide initial evidence that this favouritism was primarily driven by high-income Aymaras, who saw disproportionate gains in access to resources such as electricity, and piped water. Our results highlight how economic disparities amplify the benefits of political alignment, reinforcing existing economic hierarchies within favoured groups and perpetuating inequality even when marginalized groups gain political representation.

Speaker: Vangjel Bita (Leibniz Universität Hannover)

Socially_disadvantaged_ethni_groups_and_distributive_politics.pdf

26 Subsidizing Compliance: A field experiment on discounted sustainable fishing gear at Lake Victoria

Monitoring and punishment are common tools for increasing the cost of non-
compliance. In developing country contexts, however, these tools are often ineffective
or inefficient. We investigate an alternative approach: Subsidies that reduce the cost
of compliance. We conduct a randomized controlled trial with 799 boat owners at
Lake Victoria, Kenya and Tanzania, clustered at the level of landing sites. We
offer compliant fishing net panels for sale and investigate the effect of a susidy on
the purchase and use of non-compliant panels. We test the impact of the subsidy
under a regime with enforcement (Tanzania) and without enforcement (Kenya).
We find that a discount of 10 USD on compliant panels crowds out an average of
1.5 non-compliant panels in the absence of enforcement, which is higher than in the
presence of enforcement (half a panel). When offered a second time (3 months later)
the subsidy treatment significantly increases the likelihood of purchase, suggesting
that subsidies can lead to sustained reductions in non-compliance and this effect is
particularly pronounced under enforcement

Speaker: Sorell de Silva (Universität Heidelberg)

GDEC_subsidy_paper.pdf

27 The impacts of age of marriage laws on child marriage: Evidence from a cohort approach using 39 Sub-Saharan African countries

According to UNICEF, it will take 300 years to wipe out child marriage at the current speed of progress. Child marriage remains a huge problem despite global interventions and laws being in force in most countries, leading to profound adverse consequences. More and more countries are implementing age-of-marriage laws with a minimum marriage age of at least 18. Yet, previous literature has shown that the effectiveness of these laws can be questioned as existing evidence is inconclusive. This paper uses a novel retrospective cohort approach as an identification strategy to examine this issue. Household and context factors are simultaneously accounted for in a multilevel binominal regression for women aged 13-25 based on DHS, UNICEF, Girls not Brides, and World Bank data for 39 Sub-Saharan African countries from 1992 to 2019. After controlling for age, survey year effects, and the birth-year cohort trend, results show that a new child marriage law does not affect child marriage rates, although an update of existing law does have a reinforcing effect. The reinforcing effect of a legal update is most pronounced for child marriages between 15 and 18 years of age, it cannot be identified for the younger age group of 6- to 14-year-olds.

Speaker: Anne Lieke Ebbers (Radboud University)

The impacts of age of marriage laws on child marriage - draft paper - GDE - Ebbers.pdf

28 The Trickster’s Legacy: The Folklore Roots of Corruption

This study examines the role of oral traditions in shaping corruption. Using firm-level data on a large sample of 82,922 firms covering 285 cultural societies in 125 countries, we analyze how the representation of antisocial behavior in folklore influences corruption. Our findings reveal that societies with folklore portraying antisocial behaviors as successful tend to exhibit higher levels of corruption, while those with oral traditions emphasizing the negative outcomes of such behaviors experience lower corruption. By distinguishing demand- and supply-side corruption channels, we demonstrate folklore's pervasive influence on bribe-related decisions, affecting firms as both initiators and targets. Thus, our work contributes to the understanding of deep-rooted cultural determinants of corruption.

Speaker: Mr Jean-Baptiste Marigo (LaRGE (Research Center in Management and Economics) - University of Strasbourg)

Marigo & Weill 2025 The Trickster’s Legacy - The Folklore Roots of Corruption.pdf

29 Unseen Collateral: The Link Between Political and Intimate Partner Violence

This study examines whether conflictual events are linked to a short-term increase in intimate partner violence (IPV), the most prevalent form of violence against women in both conflict and non-conflict contexts. To this end, I use data from Domestic Violence module of 52 Demographic and Health Survey (DHS) surveys conducted in 22 countries and spatially link them to conflictual events retrieved from the Armed Conflict Location and Events Database (ACLED).
Using the discontinuities in time between the conflict event and the interview, the study estimates that a single battle occurrence increases physical or sexual IPV by 37.3 percentage points in the short-term. This effect is driven by an increase in the partner's alcohol consumption and the normalization of violence.

Speaker: Hanna BONY (Université Gustave Eiffel)

IPV_and_Conflict_BONY.pdf

30 “Rally Around the Flag” Effects: Terrorism and Political Trust in Africa

This paper utilises data from over 45,000 respondents in the Afrobarometer surveys, combined with the Global Terrorism Database, to investigate whether terrorist incidents trigger a localised “rally around the flag” effect. In the aftermath of such incidents, individuals
residing near terror attacks exhibit increased political trust, including higher trust in the president and the ruling party. Conversely, no significant changes in trust are observed for individuals living further away, with coefficient estimates indicating a near-zero effect.
The localised nature of this “rally around the flag” response is also driven by ethnic differences, highlighting the complex interplay between proximity, identity, and political trust in contexts affected by terrorism.

Speaker: Ms Josephine Baako-Amponsah (University of Bayreuth)

KfW-Prize Ceremony: Conference welcome (Krisztina Kis-Katos and Matthias Schündeln) and KfW Award for practice-oriented development research

19:30 Dinner

Friday 13 June

Registration for new arrivals

Parallel Session 2: Conflict 1

31 Does Affirmative Action Deepen Identity Divides? Exploring Inequality in Competitive Environments

Inequality in wealth, income, and opportunities based on natural identities, especially caste and gender, is widespread. One policy tool to reduce this inequality or "level the playing field" is Affirmative Action (AA) policies. Previous research has shown that AA policies effectively increase the willingness to compete with people who benefit from AA. However, most studies focus on resource differences between groups, neglecting resource differences within groups. But it could be precisely the rich person from the beneficiary group who benefits most, a concept coined ``creamy layer'' usually referring to wealthy non-upper caste members who are more likely to benefit from caste-based affirmative action. In our experiment, we introduce both within- and between-group inequality of endowment to study the effect of group- and income-based AA policies on the willingness to compete, policy choices, fairness perceptions, and inter-group relations. In line with previous literature, within-group homogeneity increases the willingness to compete in a Tullock contest for beneficiaries but does not decrease for non-beneficiaries; in contrast, within-group heterogeneity discourages non-beneficiaries, as predicted by theory. Regarding fairness, we find that participants find group-based AA the least fair, especially with heterogeneity. This is at odds with behavior as a social planner as everyone chooses the institution according to their benefit and prefers the group-based AA over the income-based AA. We further find that the creamy layer among the beneficiaries of affirmative action does suffer backlash when winning the contest, as many points are deducted from them.

Speaker: Ankush Asri (Chr. Michelsen Institute)

32 Trafficking as Choice? Evidence from Nigeria

The discourse surrounding human trafficking often portrays migrants as naive subjects of deceit and disinformation. Yet, some argue that migrants actively choose trafficking as a migration strategy. We shed light on this discussion by investigating original data from a two-wave repeated cross-section of approximately 2,700 individuals fielded in 2021 and 2023 in Edo State, Nigeria, a focal point for international migration and human trafficking. Our data shows that a significant share of respondents express a willingness to expose themselves to human trafficking to migrate even when this arrangement could lead to forced labor to pay back one’s debt to the trafficker. Results from a list experiment indicate that we likely underestimate the willingness to rely on human trafficking in direct assessments, especially among women. The downward-bias may result from female trafficking oftentimes leading to socially stigmatized and forced sex work. Yet on average, survey respondents are well-informed about the risks and earnings potential associated with trafficking. Investigating whether information is differently processed if aspirations are high, we find no evidence for self-deception or recall biases regarding the processing of information about trafficking. Instead, we suggest that individuals who would expose themselves to human trafficking exhibit high migration aspirations but limited migration capabilities, especially with respect to financial opportunities and formal education. While our study shows that choosing this high-risk migration mode does not stem from naivety, becoming receptive to trafficking due to low migration capabilities should not be confused with making the same decision under full freedom of choice.

Speaker: Tamara Bogatzki (WZB Berlin Social Science Center)

Bogatzki, Meierrieks_Trafficking_as_Choice_20250210.pdf

33 The grievances of a failed reform: Chilean land reform and conflict with indigenous communities

Land reforms are one of the most profound institutional changes in developing countries. Nevertheless, these reforms often fail, due to interruptions, reversals, and partial implementations. Unsuccessful land redistribution policies are likely to produce grievances and social unrest, further hindering economic development in areas with historically excluded populations such as peasants and indigenous groups. This paper analyzes the persistent effects of the Chilean land reform on a current indigenous self-determination conflict.
A substantial land redistribution process occurred in the study area but was entirely reverted due to a counter-reform following a military coup.
Using a detailed plot-level database, we find evidence of the persistent effects of this failed land reform on conflict.
Our results indicate that expropriated plots are at least four times more likely to be invaded and attacked between 1990 and 2021, an effect that extends beyond historically contested indigenous territories. The impact on these newly contested territories suggests that the land reform and counter-reform have a significant impact on the extensive margin of the current conflict.
These results are confirmed with instrumental variable estimates, based on our evidence that unexpected productivity shocks affected plot participation in the land reform. Potential channels explaining the results are the grievances of the failed reform, the implementation of current land restitution policies, and changes in the economic structure after the counter-reform.

Speaker: dany jaimovich (ZTUC)

mapuche_land_reform 29ago2024.pdf

Parallel Session 2: Environment 2

34 Supply-side Constraints to Technology Adoption: The Market for Energy-efficient Cookstoves in Rural Senegal

Increasing the diffusion of technologies with positive externalities is a global policy priority. We investigate the market for energy-efficient biomass cookstoves in rural Senegal, which have private and external benefits, but uptake is low. Our pre-specified study covers stove producers and intermediary vendors on the supply side, and we elicit demand from rural households using real purchase offers. We randomly relax supply-side constraints among vendors by providing demand information, marketing materials, and a transportation grant to bridge the last mile to villages where customers reside. We find that demand information and marketing materials alone do not increase vendor sales, but only in combination with the transportation grant. Yet, overall sales are still too low to deliver noteworthy take-up at the household level. We also show that market conditions are structurally challenging: markups are lower than for competing products and demand is highly elastic and volatile. This combination of high risks and low profits need to be addressed by policy interventions if large-scale diffusion is the goal.

Speaker: Maximiliane Sievert (RWI - Leibniz Institute for Economic Research)

MarketEEBC_EJ.pdf

35 Rocks and Regulations: Assessing the economic and environmental impacts of industrial mining

Minerals are crucial for the modern economy, including the transition to a green economy, but mining often causes serious environmental and social issues. Mining has both potential positive effects on development, such as economic growth and job creation, but also negative effects like the resource curse and significant environmental damage. This study looks at the impacts of opening and closing mines on a global scale, rigorously analyzing the causal effect of different types of mining in different circumstances. We focus as outcomes on changes in local economic activity as proxies for economic development, and on vegetation and air quality as proxies for environmental impact. Our empirical strategy uses state-of-the-art methods to compare areas affected by mines to similar areas that are not. Our preliminary findings show that mine openings lead to more economic activity and worse air quality, suggesting economic benefits come with environmental costs. Turning to mechanisms, we begin by assessing the mitigating effect of voluntary industry agreements, studying the role of the largest industry sustainability initiative. So far, we find mixed results: their impact varies greatly depending on the group of mines considered, with both positive and negative effects depending on the context. In addition, we will examine comprehensively on a global scale the mitigating effect of producer-country and country-of-origin regulations.

Speaker: Dylan Philippe (University of Bern)

Rocks and Regulations-- Assessing the econonmic and environmental impacts of industrial mining.pdf

36 10 Years After: Long-term Adoption of Electricity in Rural Rwanda

Power grid extension into hitherto unconnected areas is high on the policy agenda in Sub-Saharan Africa. Yet, connection rates and electricity consumption remain low in rural grid-covered areas, at least in the short and medium run. This paper provides a long-term follow-up on an evaluation of a large grid extension program in rural Rwanda. We study the adoption of grid electricity over time using a panel of 41 communities that were electrified up to ten years ago. We triangulate our own survey data with administrative consumption data. We find that in connected communities, almost half of the households remain unconnected. Electricity consumption and appliance use are low and did not grow over time. It is therefore difficult to justify investments into grid-based rural electrification by economic development impacts and cost-benefit considerations. Rights-based arguments rooted in equity and fairness considerations may provide a more compelling yet controversial justification for such investments.

Speaker: Jörg Ankel-Peters (RWI Leibniz Institute for Economic Research)

ssrn-5110634 (1).pdf

Parallel Session 2: Gender 2

37 Stuck in the Fields: How Rising Temperatures Deepen Gender Inequality in Rural India

This study highlights how rising temperatures due to climate change are amplifying gender inequality in developing countries. Using a rich district-level panel dataset from the Census of India and household surveys (1987–2018), we find that in regions with a high share of poor, landless farmers, as rising temperatures leave the agricultural sector less productive and more volatile, male agricultural laborers are increasingly transitioning into non-agricultural work. In contrast, female laborers remain trapped in agriculture and are unable to shift. We argue that entrenched gender norms impose significant costs of transition on women, which limit their ability to move sectors. We further explore additional mechanisms that exacerbate this pattern, which include declining rural-urban marriage migration of women, poor transport infrastructure, and lack of skills. Additionally, we find that as more women remain in agriculture and take on higher-valued tasks previously performed by men, their weak bargaining power leads to declining wages relative to men, which widens the gender wage gap.

Speaker: Manisha Mukherjee (Maastricht University and United Nations University MERIT)

38 Labor Market Penalty for Single Mothers in India

It is well established that there is a motherhood penalty in the labor market for child-bearing women. Theoretical models, as well as empirical estimates, suggest that unmarried or never married women without children have a relative advantage in terms of labor market opportunities. However, little is known about single mothers and their labor market outcomes. Aside from the fact that this is an expanding demographic worldwide, single mothers constitute an interesting case from a purely conceptual point of view. On the one hand, they might not have the typical social constraints of married women in traditional patriarchal societies, but on the other hand, they face the same constraints with respect to childcare and childbearing as other married mothers. While aggregate data suggests that single mothers’ labor market participation rates are usually higher than those of unmarried women, we argue that in contrast to married women without children and married mothers, this realized labor market equilibrium masks potential demand-side discrimination and likely reflects strong supply-side incentives. With the aim of uncovering potential demand-side discrimination effects, we conduct a correspondence study experiment that involves applying to real jobs using fictitious resumes. We show that equally qualified single mothers are much less likely to receive interview callbacks than unmarried women without children, married without children, and married mothers. For every interview callback a single mother has to apply to about 30 jobs, whereas an unmarried woman receives more than two callbacks for as many job applications. As a potential mechanism behind our findings, we find suggestive evidence of inaccurate statistical discrimination by employers.

Speaker: Ralitza Dimova (GNOI)

Chatterjee_Dimova_Shubham_final_draft_edit_1 (002).pdf

39 The impact of in-law mortality on women's labor supply in India

We examine how the death of co-residing in-laws affects married women’s labor force participation in India across three household surveys. Using simple and dynamic difference-indifferences we find that labor force participation increases upon the death of a co-residing father-in-law but not mother-in-law. Income effects are directionally consistent with priors but lack strong statistical support. Women with widowed fathers-in-law spend more time on domestic work, while those with widowed mothers-in-law spend more time in employment. Both increases come with reduced leisure and sleep. Agency shifts marginally, with increases in women’s decision-making power and financial autonomy following the death of in-laws.

Speaker: Rajshri Jayaraman (ESMT Berlin)

coresidence_jayaraman.pdf

Parallel Session 2: Health and education 1

40 Improving the Quality of Education in Developing Countries: An Experimental Evaluation of Teacher Training Programs in El Salvador

We conduct a large-scale evaluation study designed as an RCT in southeastern El
Salvador. Teachers are randomly assigned to either a control group or one of three
training programs focusing on (i) pedagogical knowledge, (ii) content knowledge, or
(iii) a combination of both inputs. To evaluate the programs we conduct teacher math
and pedagogy tests, student math tests and classroom observations. We find a lasting
effect on teacher content and pedagogical knowledge of up to 0.3 and 0.5 σ respectively
one year after program end. Yet, this only moderately changes teachers’ classroom
practices in the short-run and does not translate into significant student learnings.
The data most closely aligns with a setting where teachers face a dual challenge:
introducing new ideas in a rigid environment while navigating the significant learning
gaps present among students in later grades. Evidence therefore suggests focus should
be set on mitigating early learning deficits most effectively.

Speaker: Ms Carla Coccia (University of Bern)

41 Role of EdTech on numeracy and literacy for children with functional difficulties: Experimental evidence from Kenya

We conduct a field experiment to investigate the effect of an educational technology intervention on the numeracy and literacy skills of children with functional difficulties in a low-income setting with a high disability prevalence. Children with special needs in primary school are recruited through a government screening program at school. After randomization, at the school level, children in the intervention group are offered a computer-assisted-learning mobile app. Eight months post-intervention, we find that the intervention has positive and significant effects on literacy rates, with effect sizes exceeding the median effects reported in prior studies exploring literacy interventions. However, we find no significant effects on numeracy rates. Our disaggregated results shows that the intervention improves lower-order literacy skills and higher-order numeracy skills. On literacy skills, where aggregate effects are positive and significant, both males and females benefit from the intervention. Treatment effects are primarily driven by treatment compliance—whether students accept and use the tablet—rather than by changes in study effort, time allocation, perceptions about schooling, interactions with teachers and peers or in mental well-being. Our findings suggest that EdTech intervention may be effective in improving learning outcomes and that effectiveness is dependent on compliance with the treatment rather than ancillary behavioral or psychological mechanisms.

Speaker: Frank Odhiambo (ETH Zurich)

edtech_and_learning_GDEC (17).pdf

42 How English exposure in primary education shapes STEM choice: Lessons from India

English, the world's third most spoken language, is of global importance, especially in former British colonies, which are destinations for outsourced jobs. In India, with its many languages, English serves as the lingua franca, providing labor market advantages. Despite India's emphasis on vernacular languages in school education, English remains the primary medium of instruction in higher education, particularly for STEM subjects. Therefore, learning English in schools is expected to benefit students pursuing STEM. Through a language policy variation in West Bengal, India, we investigate the impact of English language exposure in primary school on subject choice in higher education. Our analysis shows that the exposed students have higher chances of enrolment in technical or STEM subjects in higher education by 5 to 7.6 percentage points. The chances of dropping out from higher education are reduced by 7.2 percentage points on average. The positive effects of English exposure on STEM selection are less pronounced among females. Improved English speaking skills seem to be one of the mechanisms, and we do not find any effect on the overall quality of education.

Speaker: Dr Gitanjali Sen (Shiv Nadar Institution of Eminence)

Parallel Session 2: History

43 The Persistent Legacy: Colonial Railroad and Agricultural Productivity in Africa

This paper examines the long-term impact of colonial-era railroad infrastructure on agricultural productivity across 24 sub-Saharan African countries. Leveraging multiple identification strategies, including comparisons with placebo lines, spatial first differences, and spatial discontinuity designs, we find robust evidence that railroad localities exhibit significantly higher crop yields in the present day. Our findings reveal that grid cells within 20 km of historical rail lines experience an agricultural productivity premium of up to 8% relative to unconnected cells. We argue the persistence arises from path dependence: early rail access enabled regions to shift into high-value commercial crops, fostering long-term specialization and agglomeration economies. Our results remain robust under a variety of specifications, including controls for geographic fundamentals, pre-colonial ethnic homeland characteristics, and post-colonial institutional differences. These findings shed light on how initial “big push” investments can lock regions into distinct spatial equilibrium path that persist for century. In turn, policymakers may either reinforce these legacy corridors or target under-invested areas to correct historical spatial inequality.

Speaker: Andu Berha (University of Alberta)

Latex_documents (16).pdf

44 The Legacy of Coercive Cotton Cultivation in Colonial Mozambique

We examine the long-term impact of forced labor on individual risk preferences and economic decisions. For that, we focus on a policy of coercive cotton cultivation enforced in colonial Mozambique between 1926 and 1961. We combine archival information about the boundaries of historical cotton concessions with survey data collected specifically for this study. By employing a regression discontinuity design to compare individuals living in areas inside and outside the historical cotton concessions, we document significant disparities in risk aversion and agricultural patterns between communities. Our findings indicate that individuals from regions unsuitable for cotton production, yet subjected to the colonial cotton regime, exhibit higher risk aversion, are more likely to engage in farming, sell their agricultural produce, adopt technology, and save. These findings are predominantly driven by individuals who recall the coercive cotton cultivation in their communities, particularly women who bore the brunt of the colonial cotton regime. This paper underscores the enduring impact of colonial agricultural policies on risk and economic behavior, providing insights into the challenges post-colonial societies face in overcoming historical legacies.

Speaker: Margherita Bove (Sapienza University of Rome)

Cotton_Risk_Draft_feb15_2025.pdf

45 Independence Movements and Ethnic Politics: The Mau Mau Origins of Ethnic Voting and Distrust in Kenya

This study examines the effects of the violent repression of independence movements on ethnic politics and social cohesion. We exploit local variation in the intensity of repression to analyze the long-run impacts of British detention camps in 1950s colonial Kenya. Using a rich body of census and survey data and a triple-difference design, we show that exposure to a detention camp increases ethnic voting in the contested 2007 presidential election and erodes contemporary trust. In addition, we show that affected individuals accumulate less wealth, are less literate, and have poorer labor market outcomes three to five decades after the event.

Speaker: Tobias Korn (Leibniz Universität Hannover)

Parallel Session 2: International 1

46 The (Other) China Shock and the Brazilian Soy Boom: Cui Bono?

Building upon the argument that factor endowments influence distributional outcomes, this paper examines the consequences of the China shock to global food markets for economic inequality in Brazilian municipalities from 1985 to 2020. I propose a new identification strategy that exploits plausibly exogenous variation in demand for soybeans based on fluctuations in the size of the pig stock in China and show that the proceeds of this China-driven agricultural bonanza have been rather unequally distributed. The soy boom has fueled land consolidation and economic inequality, especially in places dominated by large-scale mechanized agriculture. Income gains have been mostly limited to the top deciles of the distribution, while the poorest segments of the population have become worse off. Additionally, there is evidence that the more unequal a municipality, the more deforestation and rural conflict increase as soy expands.

Speaker: Samuel Siewers (Uni Göttingen)

Siewers_CSBSB_GDE.pdf

47 Global food prices and migration in Sub-Saharan Africa

In this paper, we study the effect of exogenous global food price changes on out-migration from agricultural and non-agricultural households in Sub-Saharan Africa due to economic reasons. We show that the effect of a locally relevant global food price increase on household out-migration depends on the initial household wealth. Higher international producer prices relax the budget constraint of poor agricultural households and facilitate migration. Unlike positive weather shocks, which mostly facilitate internal rural-urban migration, positive income shocks through rising producer prices only increase migration to neighboring African countries. We further find evidence that higher producer prices increase output conflict over the appropriation of surplus in agricultural districts, which serves as a parallel mechanism explaining the household decision to send a member as a migrant.

Speaker: Barbora Sedova (Potsdam Institute for Climate Impact Research (PIK))

48 Private Capital Allocation Across Developing Countries

As Chinese lending has grown to dominate global development finance, understanding its impact on other capital flows has become increasingly important. This paper provides the first causal evidence on how public investment through China’s Belt and Road Initiative (BRI) influences cross border private capital flows to emerging markets, and how geopolitical alignment moderates this relationship. To achieve this, I leverage a unique natural experiment setting: the staggered signing of a memorandum of understanding (MoU) to join the BRI. My analysis combines security level investment data before and after a country signs a BRI MoU coupled with a generalized difference-in-differences empirical strategy. I find that geopolitical alignment with China plays a critical role: for countries closely aligned with China, signing a BRI MoU crowds in private capital, i.e equity and bond investments. Conversely, for countries with less alignment, signing the MoU results in a crowding-out effect. These findings offer new insights into the strategic role of geopolitical alignment in shaping capital flows.

Speaker: Linda Maokomatanda (Ifw Kiel)

Parallel Session 2: Migration 1

49 Migration Aspirations and Knowledge About Legal Migration Opportunities

Despite vast economic disparities, international migration from developing to advanced economies remains low. This is because people do not know whether and how they can migrate, are not allowed to migrate, do not want to migrate irregularly, or cannot afford to migrate. This paper examines the effects of informing individuals about legal migration pathways on aspirations for mobility and qualifications. We conduct a randomized controlled trial (RCT) in rural Senegal, providing information and some basic assistance on the U.S. Diversity Visa Lottery, which offers medium- and high-skilled migrants access to permanent residence. The intervention significantly increases migration intentions and shifts preferences toward legal pathways. However, ineligible individuals, particularly those already contemplating irregular migration, show increased interest in irregular migration, which may be seen as an unintended consequence. The education aspirations only increase weakly at high baseline aspirations. These aspirations already surpass the requirements for the visa policy for most respondents, but our participants lack the capabilities to achieve them.

Speaker: Mr Yogam Tchokni (German Institute for Economic Research (DIW))

50 Does perceived labor market competition increase prejudice between refugees and their local hosts? Evidence from Uganda and Ethiopia

We study whether perceptions of labor market competition negatively influence out-group attitudes between refugees and their local hosts using a survey vignette experiment conducted in urban and rural Ethiopia and Uganda. Our vignette consists of a short story about a fictional job-seeker in which we randomize the citizenship (refugee/national) and occupation (same as/different from respondent). Our estimates suggest that host attitudes are significantly more negative when the vignette character is a refugee in the same occupation. Such prejudice against the out-group is not confirmed among refugees. Exploring the context-dependency of our results, evidence suggests that negative attitudes toward refugees that are tied to perceived labor market competition largely manifest in contexts of limited refugee worker presence. Hence, perceived labor market competition contributes to prejudicial attitudes, but only when the perception of job competition has little basis in reality. Additional heterogeneity analysis based on prior contact and ethnolinguistic proximity provides suggestive evidence that cross-group interactions may ameliorate concerns over out-group competition.

Speaker: Mr Mark Marvin Kadigo (Institute of Development Policy (IOB), University of Antwerp, Belgium)

Perceived Labour Market Competition and Prejudice_Experiment_JDE_SI.pdf

51 Migration Opportunities and Human Capital Investments

We examine how shocks to migration opportunities affect schooling outcomes in origin communities. We focus on the migration between Mexico and the United States, and exploit the expansion of the Secure Communities program in the US —a federal data-sharing program that substantially increased the risk of detainment and deportation for illegal migrants— as exogenous shock to the attractiveness of illegal
migration. Our results suggest that the Secure Communities program increased attendance,enrollment and educational attainment in municipalities that had stronger migration-network links with counties in the US that adopted the program early-on relative to municipalities that had ties with US counties that introduced the policy somewhat later. These results are consistent with the interpretation that the Secure Communities program implicitly raised the returns to education by making low-skill migration to the US less attractive.

Speaker: Dr Esther Gehrke (Wageningen University)

MigrShock_Educ (2).pdf

10:30 Coffee Break

Parallel Session 3: Behavior

52 Bad Money? Witchcraft Beliefs and Fairness Views in Kenya

Witchcraft beliefs are prevalent in many African societies and are generally seen as a source of distrust and an impediment to economic development. However, we have limited knowledge of individual perceptions on the role of occult forces in economic affairs. In this paper, we explore attitudes to income and inequality associated with witchcraft, using a large-scale lab experiment in Nairobi, Kenya. Combining evidence from spectator decisions with qualitative evidence from written comments and focus group discussions, we find that while some participants such income as undeserved and bad, and choose to redistribute or even destroy it, others see the use of witchdoctors as a legitimate business investment and prefer not to intervene in the allocation of income. Our findings can shed light on attitudes toward inequality and demands for income redistribution in societies where witchcraft beliefs are prominent.

Speaker: Arne Nasgowitz (Center for Applied Research at NHH (SNF))

Bad_Money.pdf

53 The Magic of the Game: Experimental Evidence on Sports Betting Behavior

As sports betting is surging worldwide, so are concerns about excessive gambling. To explore the drivers of this phenomenon, we conduct an experiment investigating how regular sports bettors in urban Tanzania value sports bets and form expectations about winning probabilities. We find that subjects assign higher certainty equivalents and winning probabilities to sports bets than to urn-and-balls lotteries with identical odds, even though, in fact, they are not more likely to win. We complement the experimental evidence with original survey data on sports betting frequency and motives. Overall, our results suggest that systematic misperceptions of the risks and returns associated with sports betting may contribute to its booming popularity.

Speaker: Dr Mattea Stein (University of Naples Federico II and CSEF)

Magic of the game_Chegere-Falco-Nieddu-Pandolfi-Stein.pdf

54 Discrimination by Teachers: Role of Attitudes, Beliefs, and Empathy

We investigate whether discrimination by teachers explains the large gap in educa- tional outcomes between students from marginalized and non-marginalized groups. Using the context of India, we start with a correspondence study to show that teachers assign 0.29 standard deviations lower grade to an exam of equal quality but with a lower caste surname. We then conduct incentivized surveys, behavioral experiments, and vignettes to highlight some of the invisible elements that are crit- ical to understanding discrimination. We  nd that teachers hold biased attitudes and beliefs about lower caste individuals, which are associated with poor grading outcomes. We conduct a mechanism intervention based on invoking empathy among teachers to mitigate discrimination. We  nd that discrimination disappears in the treatment group, and the e ect is largest for teachers with higher baseline empathy. These  ndings are not due to social desirability. Given the fundamental role that teachers play in shaping the future of students, our  ndings o er a proof-of-concept to understand mental processes that could be instrumental in designing policies to mitigate discrimination.

Speaker: Rajesh Ramachandran (Monash University Malaysia)

Discrimination_Teachers_RR_DL_ES.pdf

Parallel Session 3: Conflict 2

55 Less Bullets, More Bucks? Civil Conflict, Credit and Economic Activity: Evidence from Colombia's Peace Process

We examine how conflict and violence affect economic activity through their impact on the credit market, using Colombia’s peace process and agreement with the FARC guerrilla as a natural experiment. Our difference-in-differences analysis reveals a significant reduction in violence in municipalities formerly affected by the FARC, triggering a 12% increase in credit relative to deposits. Mechanisms include lower credit delinquency and a shift toward business financing. Additionally, we observe growth in bank branches and employment, indicating heightened competition among banks. Our results suggest that the peace process, which spurred an 8% growth in the manufacturing and a 6% growth in the service sector, positively affected economic activity in former FARC areas by increasing credit availability. More broadly, the findings emphasize the pivotal role of credit markets in facilitating economic recovery in post-conflict settings.

Speaker: Leonhard Spaeth-Rodas (GNOI)

56 The Lasting Effect of War on Female Labour Market Outcomes: Evidence from Vietnam

War results in sex imbalance, which in turn, has been shown to increase female labour force participation (FLFP). Existing evidence on the nexus between conflict and FLFP predominantly stems from developed nations which may not fully capture the dynamics in developing contexts. This paper examines the impact of the Vietnam War on Vietnamese women's labour market outcomes 14 to 44 years after its conclusion. To this end, I match comprehensive historical data on ordnance deployed by the United States in Vietnam to microdata. I find that war-induced demographic shocks contributed to increasing FLFP rates in South Vietnam, but not North Vietnam where Socialist ideology may have played a larger role in promoting FLFP. In terms of mechanism, I find that war widows increased their labour supply to compensate for the negative income shock caused by the loss of their husbands. I also find that daughters of widows are more likely to work than daughters of non-widows. However, I find a lack of support for demand-side mechanisms, namely, substitution towards female labour.

Speaker: Anri Sakakibara (TU Munich)

Sakakibara.pdf

57 The Unseen Pain of the Vietnam War: Long Term Effects of Agent Orange on Labor Market Outcomes

Using the nearest distances to Northern Vietnamese Army base as an instrumental
variable for exposure to Agent Orange, we investigate its long-term effects on education
and labor market outcomes of affected individuals in Vietnam. Our findings reveal
that a one-unit increase in the exposure score at the commune level is associated with
a significant reduction in education attainment, a decrease of 1.7 working days per
month, and a reduction of 0.4 working hours per day. These effects are particularly
pronounced among individuals who were born and continued to live in high-exposure
areas, with persistent adverse impacts observed among their children as well. Our results
remain robust after controlling for income and accounting for potential confounders.
Furthermore, we find that exposure to Agent Orange leads to a persistent reduction in
household and labor income by approximately 25%, highlighting the need for policies
to address the long term and intergenerational socioeconomic effects of Agent Orange
on the Vietnamese population.

Speaker: Po Yin Wong (Queen Mary University of London)

Agent_Orange_28Jan2025.pdf

Parallel Session 3: Gender 3

58 Mutual Knowledge of Social Norms and Political Behavior

Social norms are crucial drivers of human behavior. However, misperceptions of others' opinions may sustain norms and conforming behavior even if a majority opposes the norm. Privately shifting individuals' beliefs about true societal support may be insufficient to change behavior if others are perceived to continue to hold incorrect beliefs ("lack of mutual knowledge"). We conduct a field experiment with 5200 women in Kyrgyzstan to test whether creating mutual knowledge about social norms affects how perceived social norms influence behavior. We show that providing information about societal support for female political activism alone does not affect women's political engagement. However, when perceived mutual knowledge is created, the effect of information about social norms increases significantly. Using vignette experiments, we show that the effect of mutual knowledge on social punishment is a plausible mechanism behind the behavioral impact. These findings suggest that higher-order beliefs about social norms are an important force linking social norms and behavior.

Speaker: Lukas Hensel (Peking University)

Kyrgyzstan_Social_Norms-20.pdf

59 The Leaky Pipeline: When Career Expectations Meet Social Norms

Despite rising educational attainment for young women, traditional gender norms in developing countries limit women’s labor force participation. This study, in urban India, examines how career exploration during secondary school impacts students' expectations about women's labor force participation and views on whether women having higher education or earnings than their husbands causes marital problems. We find that the expectation to work in the future is almost universal for girls when marriage and childbearing are not mentioned but declines sharply when these life-changing events are mentioned for a similarly situated girl. Further, over one-third of students perceive that educational or income advantages for women relative to their husbands can lead to marital problems. The career exploration program causally increases labor force participation expectations post-childbearing and reduces concerns about marital problems potentially challenging traditional gender norms.

Speaker: Viola Asri (Chr. Michelsen Institute)

Asri_LeakyPipelines.pdf

60 Culture, agency and the willingness to compete: Evidence from a 1 randomized field experiment

Competition directly shapes the distribution of resources and rewards. Willingness 5 to compete, which often reflects dominant cultural norms, can therefore have outsized impacts 6 on individual outcomes. We introduced two distinct models of agency, highlighting either 7 independent or interdependent traits, into a multi-session psychosocial and personal finance 8 training among women in Vietnam. We show that women randomly assigned to the 9 independent traits are subsequently more likely to compete in an incentive-compatible lab 10 experiment than women assigned to the interdependent traits. A significant effect remains after 11 controlling for confidence and risk preference as potential mediators. We do not detect 12 heterogeneous treatment effects. (JEL C93, J16, O12)

Speaker: Robert Lensink (University of Groningen, The Netherlands)

Competition paperSubmission.pdf

Parallel Session 3: Growth 1

61 Accounting for Cross-Country Output Differences: A Sectoral CES Perspective

The standard macroeconomic literature suggests that output p.c. and productive technology are positively correlated. Removing technological disparities between countries would hence narrow the substantial variation in output p.c. across countries. But technology is factor-neutral in this standard development accounting framework, which is at odds with important macroeconomic frameworks and empirical results. We hence develop a CES production structure with factor-biased technologies and revisit the question: how much would the variation in output p.c. across countries narrow if barriers to technology adoption across countries are removed? Our results for the aggregate market sector suggest: probably not at all. But there are remarkable differences for this result across five subsectors. We use those to illustrate that our presumably surprising finding results from an interplay of complementary production factors, associated factor-bias in technologies, and the menu of available technologies.

Speaker: Konstantin Wacker (University of Groningen)

CESsectorlevel_obes_rr1.pdf

62 Disparate data, disparate results: Has the Sub-Saharan Africa region deindustrialized?

This paper demonstrates empirically that sustained and widespread manufacturing decline has not occurred in Sub-Saharan Africa, neither in the case of employment nor real value added. A review of key papers on African industrialization shows that past findings of deindustrialization are a product of small, non-representative country samples, and are relevant only for a small number of relatively rich African economies. This paper re-examines the debate using the most recent and largest available African sectoral data sources. It is demonstrated that, as the size of country samples increase and recent value added revisions are incorporated, evidence of deindustrialization disappears by all reasonable metrics. In a broad sample of 18 Sub-Saharan African countries, the average manufacturing shares of employment and real value added did not decline by a statistically significant magnitude in any decade since the 1960s when controlling for population and income, and the average manufacturing share of real value added increased in each decade in a near-comprehensive sample of 45 African countries.

Speaker: Calumn Hamilton (Groningen Growth and Development Centre)

MHindustrialization_WBER-1.pdf

63 Taking Kuznets Seriously

Numerous studies have tested Kuznets' hypothesis of a hump-shaped relationship between inequality and industrialization by regressing the Gini coefficient on the level and square of per capita income. Here, we examine Kuznets' original idea that inequality first increases and then declines when workers move from agriculture to manufacturing. We collect sectoral wage and employment data for 17 advanced countries over the period 1800-2020 and compute the `Kuznets Gini' as the Gini coefficient that results from changes in sectoral employment shares and the sectoral wage differential. We show the hump-shaped path of the Kuznets
Gini and establish, in panel regressions, that inequality arising from sectoral migration has been an important determinant of overall income inequality.

Speaker: Holger Strulik (University of Goettingen)

kuznets_revised.pdf

Parallel Session 3: Health and Education 2

64 The effect of mobile internet on child marriage and teen pregnancy in sub-Saharan Africa

Child marriage and teen pregnancy remain prevalent in sub-Saharan Africa. This paper examines the impact of mobile broadband internet (3G) expansion – a key driver of social media access – on these outcomes. We combine geo-referenced data on 3G coverage with individual-level marriage and fertility data from over 500,000 women-year observations across 20 sub-Saharan African countries between 2011 and 2022. Using two-way fixed effects, event study analysis, a stacked difference-in-differences approach, and an instrumental variables strategy, we find that 3G coverage significantly reduces the likelihood of child marriage and teen pregnancy. Adolescent girls exposed to 3G are between 1 and 6 percentage points less likely to marry and approximately 1 percentage point less likely to give birth, while no significant effects are observed for older women. The effects are strongest among wealthier and more educated adolescent girls, who are more likely to adopt digital technologies early. Exploratory analyses suggest that increased access to information about family planning, shifting gender attitudes, and behavioral changes – such as greater contraceptive use – serve as key mechanisms. Our findings imply that the expansion of mobile broadband may be more effective in reducing early marriage and fertility than many conventional family planning interventions.

Speaker: Jet Wildeman (Radboud University Nijmegen)

3G_coverage__child_marriage_and_teen_pregnancy_in_Africa_13_02_2025.pdf

65 Pedalling Progress? Assessing the Long-Term Impact of a Bicycle Intervention on Adolescent Maternal and Child Health in Zambia

This study evaluates the long-term impact of a gender-transformative intervention that provided bicycles to adolescent girls in rural Zambia to enhance school attendance and empowerment. Implemented in 2017 across 100 schools, the intervention aimed to reduce travel barriers and improve educational access. Using an experimental design with a final sample of 1,615 adolescent girls (676 treatment, 939 control), we estimate the causal impact of bicycle ownership on maternal and child health (MCH) and sexual and reproductive health (SRH) outcomes five years post-intervention.

Our findings indicate a significant increase in adolescent pregnancy among the treatment group, with an odds ratio (OR) of 1.47 (95% CI: 1.12, 1.94), suggesting a 47% higher likelihood of pregnancy compared to the control group. However, the intervention had no significant effects on antenatal care visits, iron-folic acid use, institutional deliveries, birth weight, or exclusive breastfeeding. For SRH outcomes, girls in the treatment group were significantly more likely to report ever having had sexual experience (OR: 1.54, 95% CI: 1.26, 1.88) and a higher desire for pregnancy (OR: 1.11, 95% CI: 1.04, 1.18). Causal mediation analysis suggests that increased empowerment partially mediated the observed rise in adolescent pregnancy.

These findings highlight the complexities of gender-transformative interventions in low-resource settings. While aiming to empower adolescent girls, increased mobility and decision-making power may intersect with unintended reproductive health consequences. This underscores the need for integrated program designs that complement empowerment initiatives with comprehensive SRH education and services to mitigate adverse outcomes.

Speaker: Shruti Shukla (GNOI)

66 The effect of temperature on congenital disorders: Evidence from Mexico

Congenital disorders are a principal cause of early mortality, long-term disabilities, impaired cognitive development and constitute a major challenge to families,communities, and health care systems alike. The origins of congenital disorders are not yet well understood with descriptive evidence from the fields of medicine and public health pointing towards the role of climate related factors. Leveraging detailed micro-data from Mexico, this study aims to provide the first causal evidence on the role of climatic shocks in affecting the onset of congenital disorders. For this purpose, we compiled a large dataset comprising about 19 million births from about 63,000 Mexican localities for the period 2008– 2021. Focusing on temperature shocks, our results
indicate that an increase in temperatures by 1°C, is related to an increase in congenital disorders by 10 percent (0.1 percentage points).

Speaker: Kerstin Perlik (Bernhard Nocht Institut für Tropenmedizin)

Perlik_temperature_cong_disorder.pdf

Parallel Session 3: Migration 2

67 Marriage Matters: Migration and Marital Sorting in Indonesia

This paper examines how internal migration barriers affect marriage and migration decisions, shaping marital sorting and the spatial distribution of human capital. Using a structural matching model, I explore the trade-offs individuals and couples face when choosing spouses, marriage markets, and labor markets. The counterfactual analysis shows that removing migration barriers substantially affects both individual and joint migration decisions, which can be masked when only observing total population flows from rural to urban areas. Further, I document how this policy alters marital sorting patterns, and increases the prevalence of joint migration. Assortative mating by education persists, but highly educated couples become more evenly distributed across regions. These findings highlight the broader implications of migration policies on household dynamics and regional inequality.

Speaker: Ariane Gordan (University of Luxembourg)

Ariane-Gordan_Migration-and-Marriage.pdf

68 The Impact of Regional Growth on Internal Migration: A District-Level Analysis for Malawi

Research on the nexus between development and migration has mainly focused on cross-border flows. How income changes affect migration within developing countries is much less well researched even though addressing this topic might provide essential information about the process of structural transformation needed for economic development. In this paper, we provide new evidence on the link between income growth and internal migration for Malawi, one of the poorest countries worldwide where migration is predominantly internal. Employing a gravity approach and performing an instrumental variable regression where geo-coded future World Bank aid projects committed in the economic and production sector serve as instruments, we robustly find that, on average, rising incomes – proxied by changes in nightlight intensity – are associated with higher emigration rates. This effect is mainly driven by people emigrating from comparably richer urban areas. In the very poorest districts, by contrast, migration is shown to fall with increasing economic activity, which is in accordance with the notion that extreme poverty may force people to leave their home.

Speakers: Prof. Mauro Lanati (University of Florence), Rainer Thiele

Lanati and Thiele GDE.pdf

69 Externalities of U.S. Immigration Enforcement Policies in El Salvador

By 2020, about 25% of El Salvador’s population lived abroad, with 88% residing in the United States. Remittances contribute 25% of El Salvador’s GDP, underscoring the im- portance of migration to the country’s economy. U.S. immigration enforcement policies that increase deportations of unauthorized migrants are therefore likely to significantly impact El Salvador’s labor market. Using a shift-share approach, we estimate the ef- fects of these policies on employment outcomes. Our results show that enforcement policies reduce monthly income for waged workers, with the largest impacts observed among men with lower education, workers lacking social security or contracts, and those employed in agriculture. We also find a decrease in the probability of having a migrant within households and a reduction in both the likelihood and amount of re- mittances received. These findings suggest labor competition and reduced remittances as key mechanisms driving the observed effects.

Speaker: Christian Ambrosius (GNOI)

Ambrosius Quigua Velasquez Effect_of_Enforcement_Policies_in_El_Salvador.pdf

Parallel Session 3: Public Economics 1

70 Female community-driven development and empowerment: An RCT in Pakistan

We study the impact of a community-driven development (CDD) program targeting only women on social, political, and economic empowerment. Our intervention combines social mobilization and support packages for poor households. We randomized the treatment across 23 clusters of settlements and sampled 2290 households from 150 settlements. We find indication that the intervention might have increased information about local government for the whole sample and strong evidence for strengthened perceptions about political participation, as well as access to public goods for women who assume a leadership role. We can only identify such heterogeneous effects on self-selected female leaders because our control group also received treatment and selected leaders after the midline. We find no significant effects on intrahousehold decision-making, household’s economic well-being, and social cohesion.

Speaker: Andreas Landmann (Friedrich-Alexander-Universität Erlangen Nürnberg)

Success paper 230125.pdf

71 Place-based Policies and Household Wealth in Africa

This paper provides novel evidence on the impact of a prominent place-based policy - Special Economic Zones (SEZs) - on the economic well-being of African households. Exploiting time variation in SEZ establishment on a dataset of repeated cross-sections of households in 10 African countries during 1990-2020, we show that households living near SEZs become wealthier relative to the national average after SEZ establishment. The effect is not driven by residential sorting, accrues mostly within 10 km of SEZs, and is accompanied by increased access to household utilities, higher consumption of durable goods and a shift away from agricultural activities.

Speaker: Alina Mulyukova (Kiel Institute for the World Economy)

Mulyukova_SEZs.pdf

72 Vulnerability from welfare? Timing of benefits and consumption smoothing in India’s public works program

The stated objective of the world’s largest public works program, India’s NREGA, is to provide rural households with a safety net and to reduce their vulnerability to income and expenditure shocks through paid employment on demand. In practice, however, work supply is often rationed, its timing erratic and payments delayed. We investigate the consequences of these implementation frictions for households’ consumption paths and welfare. We combine administrative data from the workfare program with highfrequency consumption and transactions data for four villages in South India. We find that consumption is largely protected from public employment income ”shocks” in better-off villages, where beneficiary households command significant financial wealth relative to monthly consumption expenditures. In the poorest and remotest village, however, where financial wealth is minimal and 80 percent of beneficiary households are burdened by net debt, monthly consumption largely tracks NREGA income payments. In a structural calibration exercise we evaluate the welfare loss resulting from this systematic consumption variability at thirteen percent of average program benefits. We conclude that demand-driven social protection programs need to be more responsive to beneficiaries’ needs and deliver promptly in vulnerable settings to not only increase average incomes but also reduce intra-seasonal consumption volatility effectively.

Speaker: Stefan Klonner (Heidelberg University)

Oldiges_Klonner_March2024_2l.pdf

12:30 Lunch

Plenary Session with Keynote Imran Rasul (University College London)

Parallel Session 4: Agriculture 1

73 Weathering Labor Market Shocks: Droughts and gendered employment in India

Climate shocks can significantly reshape employment patterns in low- and middle-income countries. Given persistent structural, social, and economic gender disparities, women are likely to experience these effects more acutely, yet existing evidence remains inconclusive. This study examines how droughts influence female employment in India, where many women face systemic barriers to labor market participation. Using individual-level data from the Demographic and Health Survey (2005–2019), we exploit district-level variations to assess the impact of droughts on women’s employment. Our findings indicate that droughts increase female employment by six percentage points, but this rise is driven by economic distress rather than expanded opportunities. Women predominantly enter low-skilled and informal occupations, while simultaneously facing heightened social constraints. We document a reinforcement of gender-biased norms during droughts, including stronger son preference, lower female migration, and declining female educational enrollment. Further analysis reveals that employment gains are concentrated in informal, seasonal, and family-based labor, with a nine percentage-point increase in low-skilled occupations. These results suggest that climate shocks exacerbate gender inequalities in labor markets in India. Our findings underscore the urgent need for policies that not only enhance climate resilience but also promote gender-equitable labor opportunities to prevent the entrenchment of precarious employment patterns in similar contexts.

Speaker: Anna Minasyan (University of Groningen)

74 Droughts, Agriculture, and the Economy: Time and Place Matter

Climate change causes more frequent and more severe droughts in many regions of the world with potentially devastating effects on the local agricultural sector. In this paper, we study the transmission of such effects from agriculture to the wider economy. To this end, we construct a new spatially explicit drought index, which takes into account both climatic factors, and the local growing seasons of different crops. We validate our index using annual data from over 5,000 municipalities and 10 crops in Brazil from 1990 to 2020. We then aggregate our drought measure to the country-level to study macroeconomic effects. Across all countries in the world, one standard deviation in our drought index reduces total crop yields by five percent, about twice as much as we would have found without taking crop calendars into account. We also demonstrate that most, if not all, declines in GDP caused by drought originate in agriculture.

Speaker: Leonard Krapf (Uni-Göttingen (RTG 2654 - A1))

Parallel Session 4: Conflict 3

75 Roads to Peace or Discord: How does Infrastructure Shape Conflict Dynamics in Africa?

This paper examines the impact of infrastructure projects on conflicts in Africa using georeferenced data on infrastructure projects and conflict events at the spatial resolution of 0.1 degree× 0.1 degree for all of Africa between 2002 and 2019. Employing a spatial difference-in-differences design and instrumental variable strategy, we find that infrastructure projects, particularly roads, reduce high-intensity conflicts like battles. However, they also increase low-intensity conflicts such as riots and protests. The decrease in high-intensity conflicts is mainly attributed to reduced conflicts among stationary rebel groups, while the rise in low-intensity conflicts is concentrated in politically marginalized ethnic regions. We provide evidence that the opportunity cost channel is crucial in driving the relationship between infrastructure and high-intensity conflicts, while grievances play a significant role in low-intensity conflicts.

Speaker: Mingjie Xu (Department of Economics, University of Mannheim)

76 Insuring Peace: Index-Based Livestock Insurance, Droughts, and Conflict

We provide quasi-experimental evidence of how an innovative market-based solution using remote-sensing technology can mitigate drought-induced conflict. Droughts are a major driver of conflict in Africa, particularly between nomadic pastoralists and sedentary farmers. The Index-Based Livestock Insurance (IBLI) piloted in Kenya provides automated, preemptive payouts to pastoralists affected by droughts. Combining plausibly exogenous variation in rainfall and the
staggered rollout of IBLI in Kenya over the 2001-2020 period, we find that IBLI strongly reduces drought-induced conflict. Key mechanisms are an income smoothing effect and reduced migratory pressure for pastoralists, reducing the likelihood of miscoordination with other land
users. Our study suggests that market-based solutions are a scalable, cost-effective pathway to mitigate conflict, complementing political solutions such as institutional reforms.

Speaker: Kai Gehring (University of Bern)

Gehring Schaudt (2024 CESifo) Insuring Peace - Index-Based Livestock Insurance, Droughts, and Conflict.pdf

Parallel Session 4: Environment 3

77 Road to Green? The Effect of Highway Expansion on Industrial Emissions Intensity

Transportation infrastructure is widely recognized for enhancing economic efficiency and regional connectivity. However, the very projects that reduce trade costs and stimulate industrial growth may also contribute to environmental degradation, raising an important question: can this apparent trade-off be reconciled? This paper investigates how highway expansion, through reductions in trade costs, influences the aggregate emissions intensity of industrial firms via its effect on resource allocative efficiency. To examine this relationship, I extend the model from \cite{10.1257/aer.20120549} by incorporating emissions as an additional dimension, drawing on insights from \cite{copeland2005trade}. The extended model derives conditions under which highway expansion improves industrial emissions intensity by enhancing resource allocation across firms. These findings underscore a potential mechanism through which infrastructure projects can align economic growth with environmental objectives.

Empirically, I examine the causal effects of highway expansion on industrial emissions intensity using detailed firm-level data from China. First, leveraging the identification strategy in \cite{wu2023road}, I show that Chinese highway expansion reduces markup dispersion within narrowly defined industries and leads to a decline in emissions intensity at the province level, validating the allocative efficiency channel. Second, employing geospatial highway data and an additional instrumental variable strategy, I find that highway expansion also reduces emissions intensity at the county level. These findings highlight a previously underexplored environmental co-benefit of transportation infrastructure. By bridging the gap between economic development and environmental sustainability, this study offers a novel perspective on the dual role of transportation infrastructure.

Speaker: Siyu Zhang (Center for Doctoral Studies in Economics (CDSE), University of Mannheim)

Road to Green_ENTER Jamboree_Siyu Zhang.pdf

78 Road infrastructure, economic development and forest loss in the Republic of the Congo

Abstract
We analyze the impact of recently constructed, large-scale road infrastructure
on local forest loss and economic development in the Republic of the Congo
(RoC) between 2000 and 2020. Situated in Western Central Africa, the RoC
is home to a humid tropical forest that serves as the world’s only stable car-
bon sink, absorbing 1.5% of global annual carbon emissions. In recent years,
the government has sought to diversify an economy strongly dependent on
oil exports and defined by limited industrialization. A key component of
this strategy has been the construction of modern road infrastructure, which
has been historically inadequate. To answer our question, we employ Call-
away and Sant’Anna’s differences-in-differences estimator and combine data
on roads with remote sensing data of forest cover (capturing deforestation),
cropland (capturing agricultural expansion), nighttime lights and physical
built-up area (capturing market development and urbanization). We provide
evidence of trade-offs between local economic development and the environ-
mental costs of road infrastructure investments, observing notable differences
across various parts of the country. Overall, our data indicate that the con-
struction of new roads leads to significant forest cover loss in most areas
within a 10km corridor around these roads, with the effects becoming more
pronounced in later observation periods. In the most developed part of the
country, the south, road infrastructure is related to the expansion of markets,
agriculture, and urbanization. In the economically less developed center and
north, there is some evidence of agricultural development but little to no
evidence of urbanization. However, we do find evidence of market develop-
ment in the north and that this seems to be connected to the systematic
exploitation of forest resources.

Speaker: Dr Laura Metzger (Radboud University)

Parallel Session 4: Growth 2

79 Democracy and Capital Inflows

Democratic regime change has large positive effects on income per capita, yet the underlying transmission channels of this relationship are largely unexplored. We provide novel evidence studying the effect of democratic regime change on capital inflows in a large sample of economies from 1975 to 2015. Using heterogeneous difference-in differences estimators, we find that regime change causes an average 40-50% (60-95%) increase in gross capital (FDI) inflows within two decades. We document significant treatment effect heterogeneity by geography and colonial history (but not legal origin or culture), which we explain highlighting structural deficiencies in export concentration, trade costs, and complexity of production. We find little evidence that the geographic patterns arise from indirect effects of ‘nature’ shaping legal origin, culture, or colonial history.

Speaker: Markus Eberhardt (University of Nottingham (UK))

Democracy_and_Capital_Flows.pdf

80 Seeking Scientific Consensus – An Expert Survey on the Replication Debate between Acemoglu et al. (2001) and Albouy (2012)

The publication of contradictory replications often sparks fierce debates between replicators and original authors. This paper investigates whether impartial experts can reach a consensus on a famous yet unsettled replication debate about the seminal paper by Acemoglu, Johnson, and Robinson (AJR, 2001) and the replication by Albouy (2012). We recruited 352 experts from the pool of scholars citing one of the involved or similar articles. Through a structured online questionnaire, we assess the extent to which these experts align with AJR or Albouy. Our findings indicate no clear consensus on whether the original results hold, though experts slightly lean towards Albouy. Notably, experts with greater academic credentials are more likely to align with Albouy. Our study demonstrates a potential way to navigate replication debates, an issue that will become increasingly relevant as more replications generate further disputes.

Speaker: Julian Rose (RWI - Leibniz Institute for Economic Research)

Parallel Session 4: International 2

81 Disaster-induced import dynamics: Evidence from South African floods

This paper examines the impact of floods on regional import dynamics in South Africa. Natural disasters like floods can disrupt firms’ production activities, hindering their participation in import markets. However, firms may increase imports to offset disruptions in their domestic supply networks. Our study explores this adjustment behavior using administrative firm-level and customs transactions data from South Africa. Analyzing a monthly panel of import aggregates and market entries across local municipalities from 2013 to 2021, we find that floods generally deter new firms from entering import markets, yet the disruption to domestic supply chains prompts firms to seek alternative suppliers abroad. Although overall average import values remain unchanged, firms adjust by increasing imports of capital goods. Our results also show stronger import responses from the European Union following supply chain disruptions due to floods. Additionally, firms in manufacturing or mining tend not to respond to flood shocks by adjusting their import levels. Instead, such adjustments are more likely to occur through the trade sector. Such adjustment dynamics would be overlooked when focusing solely on manufacturing firms’ data.

Speaker: Justyna Jantos (Universität Göttingen)

GDE_Floods_Trade_South_Africa.pdf

82 The Global Software Development Network

Can developing countries benefit from exporting opportunities in the growing sector of tradable services, given the near free information flow via the internet and wage differentials relative to developed countries? Focusing on the software development industry, we analyse data from 2.55 million software projects across 5,400 locations, and estimate an economic geography model in which locations trade tasks. The results reveal three factors limiting exports: (i) significant productivity differences within and between countries; (ii) a notable decline in trade volumes with distance; (iii) sorting patterns among software developers that are suggestive of brain drain.

Speaker: Carlo Birkholz

Birkholz_Gomtsyan_The_Global_Software_Production_Network.pdf

Parallel Session 4: Labor 2

83 Living Income measurement methods: A comparative study and application to cocoa farmers in Cameroon

How should the cost of a decent life be quantified? Are the available living income methods and indicators valid welfare measures? Additionally, are these suitable for the contexts where they are being leveraged for agrifood policies and interventions? This paper critically examines two prevailing methodologies for estimating living income indicators and their application in rural agricultural contexts, with a focus on cocoa producers in Cameroon. It compares the main approaches for estimating a living income benchmark (LIB), documenting and highlighting key differences in data sources and computational assumptions. The study finds that LIB estimates are highly sensitive to food expenditure assumptions and the valuation of non-food, non-housing (NFNH) elements of a decent life. Statistical and indicator property tests are then applied to assess the robustness of the living income gap (LIG). Stochastic dominance analysis demonstrates that LIG indicators consistently identify vulnerable groups and thus harness targeting potential. Simulations based on poverty axioms indicate the indicators are distribution sensitive, illustrating their potential for informing the design and monitoring of LIG-reducing policy instruments. As a result of these tests, a new censored LIG is proposed that further enhances the possibility of measuring and monitoring the LIG among more vulnerable strata. Ultimately, while the living income approach reframes the narrative on welfare analyses from a subsistence to decency framework, the potential of the indicators to support equitable outcomes in agrifood systems would be enhanced by integrating greater methodological rigor, replicability and harmonisation.

Speakers: Ajapnwa Akamin (Food and Agriculture Organization of the United Nations), Ana Paula de la O Campos (Food and Agriculture Organization of the United Nations), Katharina Krumbiegel (Joint Research Centre of the European Commission), Katia Covarrubias (Food and Agriculture Organization of the United Nations), Pascal Tillie (Joint Research Centre of the European Commission)

GDEC_LI methods paper_submit.pdf

84 A Multi-Modal Approach to Estimating Prevalence: The Case of Labor Trafficking

Policymakers, researchers, and other stakeholders often care about the prevalence of complex, sensitive outcomes or behaviors, such as human trafficking and forced labor, gender-based violence, child labor, or corruption. But the measurement of such issues is challenging: Comprehensive administrative data is usually unavailable, survey respondents may define these concepts very differently than existing legislation, and social stigma may lead to substantial underreporting. In this paper, we set up a conceptual framework that identifies four theoretically distinct types of prevalence rates. We then propose a survey design approach that is practically feasible in many contexts and that, with only one additional assumption beyond those commonly made in the literature, can recover all the conceptual information. We apply that approach to estimate the prevalence of labor trafficking in two countries, Malawi and Zambia.

Speaker: Laura Zimmermann (University of Georgia)

MalawiZambia_List_Experiment.pdf

Parallel Session 4: Microfinance 2

85 Reviving Group Lending: Asymmetric Group Loan Contracts and Moral Hazard

While joint-liability lending has been central to microfinance since the 1970s, recent evidence indicates that this mechanism has fallen short of its promise to reduce loan defaults and operational costs. Since borrowers of joint-liability loans commonly face symmetric contract terms, group members are subject to coordination and free-riding problems. Peers might enforce each other to repay their loans, but might also free-ride or jointly default. We propose that asymmetric joint-liability microfinance loan contracts are a suitable instrument to increase monitoring efforts and thereby repayment rates. With such asymmetric contracts, one borrower becomes an optimally incentivized lead-borrower (group-leader) through asymmetric interest rates, which is hypothesized to increase monitoring efforts and thereby reduce moral hazard of all group members. We rely on the theoretical model of Carli and Uras (2017) for a lending framework with ex-ante moral hazard and extend it to capture a lending scenario with ex-post moral hazard. We then test the impact of asymmetric loans in ex-ante and ex-post moral hazard contexts by means of a lab-in-the-field experiment with microfinance clients in urban Bolivia. The experimental results show that asymmetric loans (and incentivized group loan leaders) increase the incidence of monitoring in both moral-hazard scenarios by 9-13 percentage points. Overall, our findings suggest that assigning asymmetric roles within joint-liability groups and incentivizing a group leader can effectively address a broad range of moral hazard issues and enhance the stability of microfinance programs.

Speaker: Dr Manuela Fritz (Technische Universität München)

86 The Effects of Information Sharing on Moral Hazard in Credit Markets – Evidence from a Randomized Evaluation in the Philippines

Moral hazard in credit markets undermines loan repayment. To study whether information sharing among lenders strengthens repayment incentives for borrowers, we combine a theoretical framework with experimental evidence from information campaigns. The campaigns target 5,400 microcredit borrowers in the Philippines and randomly vary their knowledge of an existing credit registry. Campaigns increase borrower effort and reduce monitoring in borrowing groups. While neither investment size nor the use of alternative credit sources are affected, borrowers shift from high-risk, high-return investments to safer, lower-yield alternatives. Repayment rates remain high. An additional treatment arm reveals that these effects are partly driven by a reminder effect of the intervention rather than the informational content itself. This suggest that strategic communication by lenders, even without a registry, can influence borrowers in the short run.

Speaker: Lisa Spantig (RWTH Aachen)

16:00 Coffee break

Parallel Session 5: Agriculture 2

87 General vs. Tailored Information for Technology Adoption: Evidence from a Cluster Randomized Controlled Trial in India

Recent studies highlight information constraints as an important barrier to technology adoption, but there is little evidence that allows to distinguish the roles of different information frictions for adoption decisions across different types of technology. We conduct a cluster randomized controlled trial among 1,200 farmers in Haryana, India, to promote adoption of early sown wheat varieties and zero tillage technology—two agricultural innovations that can help farmers adapt to climate change by increasing resilience to heat and water stress. Two distinct information treatments are studied, a one-time group training and an individual-specific advisory service. The experiment is designed to identify distinct impacts of general versus tailored agricultural advice across different types of technology and test for spillovers on other farmers’ adoption decisions. This is guided by a theoretical framework that decomposes optimal usage choices of technology into a systematic component and an idiosyncratic component, generating testable predictions regarding the distinct roles of general and tailored information. The insights derived from this study have strong policy implications and help in understanding the role of information frictions for adoption decisions beyond the specific technologies considered here.

Speaker: Dominik Naeher (University of Göttingen)

88 Livelihood Adaptation, Transition, and Diversification: New Insights from High-Frequency Data

This paper examines the differential relationship between income diversification and intertemporal activity transitions on household well-being, addressing both intra-household and intertemporal aggregation problems.
Using high-frequency monthly data, I disentangle genuine diversification strategies from mere activity transitions that are often conflated in conventional analyses due to infrequent data collection and extended recall periods. This approach mitigates measurement errors and intertemporal aggregation bias. Also, the existence of two cyclone experiences in our time frame allows us to identify factors correlated with each strategy and their association with household well-being. I document that households engaging in diversification and transition can have different livelihood profiles, which are often linked to external shocks and forced adjustments. Furthermore, the findings reveal that genuine real-time diversification is more strongly linked to improved well-being, whereas transitions without diversification may indicate vulnerability, and relying solely on annual data can misrepresent these transitions as diversification. These insights highlight the importance of high-frequency data in accurately capturing household livelihood dynamics, which offers critical implications for targeted policy interventions.

Speaker: Farnaz Safari (Cornell University)

Livelihood Adaptation, New Insights from High-Frequency Data.pdf

89 Cultivating Change: Long-Term Effects of Repeated Training on Organic Farming Adoption in Indonesia

Most impact assessments of agricultural training evaluate one-time interventions over short time frames. However, farmers may show initial enthusiasm for a new technology but disadopt after a trial period. Others may adopt practices gradually over time. This study investigates the causal impact of repeated training on the adoption of organic farming practices among Indonesian smallholder farmers. We use four waves of data covering five years and rely on a randomized roll-out of training interventions. The training was hands-on and resembled traditional extension programs. Our findings show that repeated training significantly increased the adoption of organic farming practices including the use of lime and the leaf color chart. However, we observe only little substitution, i.e. only few farmers fully converted to organic farming. Although farmers substantially reduced the use of nitrogen, they still rely on chemical fertilizer and pesticide. Yet, the results show clear evidence for learning and increased awareness of the adverse effects associated with an overuse of chemical inputs. We contribute to the understanding of longer-term adoption dynamics following extension programs.

Speaker: Michael Grimm (University of Passau)

Parallel Session 5: Gender 4

90 Alcohol Prohibition and Women's Autonomy: Evidence of Policy Spillovers

Public policies often have the potential to generate substantial welfare effects beyond their intended objectives. One such public policy intervention that warrants closer examination of its second-order effects is alcohol prohibition. In this paper, I contribute to the debate on welfare implications of alcohol regulation policy by studying the impact of an alcohol ban implemented in the Indian state of Bihar in 2016 on women's autonomy outcomes. Using data from the two latest rounds of the National Family Health Survey (NFHS), and employing a difference-in-differences model, I find that the policy substantially increased women's autonomy outcomes - including contraceptive use, participation in household decision-making, and attitudes towards spousal beating. Investigating potential mechanisms, I find that these improvements in women's autonomy are primarily driven by reduced alcohol consumption and decreased incidences of Intimate Partner Violence (IPV). These findings contribute to our understanding of how substance abuse regulations in developing economies can have substantial spillover effects on gender-related development outcomes.

Speaker: Apurva Borar (Cornell University)

91 Can Media-Reported Sexual Violence Deter Women’s Educational Participation?

We study the effect of proximity to sexual assault events reported by media on women’s education in India. By combining novel geocoded data on media coverage of sexual crimes with nationally representative micro-data, we find that one standard deviation increase in the lagged average distance to sexual assaults increases the schooling of women by 0.17 years and the chances of middle school completion by 1.05 percentage points. Regions that have more access to television, radio, internet, and cellular devices drive our results. The effect is stronger in localities with higher son preferences, a proxy for gender norms. Our estimates are robust to empirical specifications that remove the effects of reported crimes. The results survive sensitivity and falsification checks. These findings highlight that while mass media can influence the stigma and threat of sexual violence, it can also exacerbate the problem of underinvestment in girls’ education by families in conservative societies.

Speaker: Subarna Banerjee (GNOI)

92 Conditional Cash Transfers, Labour Market and Unpaid DomesticWork: Gendered Trade-offs

Conditional cash transfer (CCT) programs are key social safety net tools in middleincome countries, designed to alleviate poverty and influence household behaviors in education and healthcare. While their effectiveness in achieving these goals is well-documented, research increasingly examines their broader consequences, including labour market outcomes. However, the way CCTs shape the trade-offs between labour market work, own-use production of services (OPS), such as unpaid care and chore work, and the potential time reallocation within households remains understudied. This paper addresses this gap using the case of the Brazilian Programa Bolsa Fam´ılia (PBF). We construct a unique panel from the Brazilian employment survey to examine labour market effects and leverage its crosssectional data to analyze OPS outcomes. Applying a difference-in-differences approach, we assess changes in labour market outcomes before and during PBF receipt. Additionally, we conduct a correlational analysis of OPS work, focusing on heterogeneous dynamics linked to household composition. Early findings indicate that while PBF does not reduce overall female labour market participation, it lowers the likelihood of securing formal employment and shifts women towards informal jobs, which offer greater flexibility for balancing work and domestic duties. These effects are uniform across women with and without children, though beneficiary mothers are less likely to hold formal jobs, suggesting an income effect that encourages specialisation in unpaid domestic work rather than a withdrawal induced by programme conditionalities. The correlational analysis reveals that PBF is linked to an increase in the weekly hours women devote to unpaid domestic work — in absolute and as a share of their total working time — a pattern not observed among men. Notably, regardless of PBF or family status, women consistently perform around 70% of unpaid domestic work of the family. These results highlight the gendered trade-offs induced by CCT programs, suggesting that while they provide crucial financial support and improve education and health outcomes, they may also maintain traditional gender roles by shifting women’s time allocation away from the labour market and toward unpaid domestic work.

Speaker: Mathilde Bouvier (Dauphine-PSL University)

Bouvier.pdf

Parallel Session 5: Health and Education 3

93 Menstrual Stigma and Girls’ Outcomes: An RCT Intervention among Schoolgirls in Nigeria

Menstruation poses a challenge to girls in low and middle-income countries. However, little is
known about psychosocial constraints facing girls’ menstrual experience and the resulting
impact. In this study, we propose a Randomised Control Trial (RCT) targeting 880 schoolgirls
in Nigeria. Through the RCT intervention, we intend to unpack the economic, knowledge, and
psychosocial constraints facing schoolgirls of menstruation age. The intervention proposes
three treatments varying exposure to free pad distribution & information, anti-stigma training,
and belief correction. Results will reveal whether treatment improves girls' well-being,
menstrual-related behaviour, and academic outcomes

Speaker: Terhemen Agabo (KU Ingolstadt-Eichstätt)

PAP_Menstrual Stigma and Girls Outcomes_final.pdf

94 Unconditional Cash Transfers and Child Labor and Schooling in Turkey: Short-Term Evidence from the Family Support Program

We examine how a modest unconditional cash transfer policy affects child labor and schooling during periods of economic crisis by studying Turkey's Family Support Program, launched in 2022. Using a regression discontinuity design based on the program's per capita income eligibility threshold, we analyze the program's short-term effects within six months of implementation. Despite the program's relatively modest transfer amounts—approximately one-third of the monthly minimum wage—we find significant reductions in children's participation in family businesses and agricultural work. Notably, these labor reductions occurred without corresponding increases in school enrollment or time spent on educational activities, which were already high at baseline. We also find improvements in children's emotional well-being and daily protein consumption, suggesting that even a modest transfer policy can enhance child welfare through multiple channels.

Speaker: Dr Aysun Hiziroglu Aygun (Istanbul Technical University)

textFeb13_2025.pdf

95 Medical and Cost Uncertainty in Health Care Seeking

Health care decisions are sometimes a matter of life and death, but also their financial consequences can be disastrous for many households around the world. While it is easy to imagine how the fear of making an expensive mistake may deter particularly poor households from making otherwise sensible health care choices, there is almost no scientific evidence on the importance and joint role of medical and cost uncertainty. In this paper, we provide theoretical insights based on simulation exercises, introduce a novel measurement instrument for medical and cost uncertainty related to health investment choices, present data collected using this instrument from low-income households in Pakistan, and test the theoretical predictions using this data. Our analysis reveals that both medical as well as cost uncertainty in health care decisions exist to a substantial degree. The empirical results regarding the influence of beliefs are largely in line with theoretical predictions and suggest that uncertainty in the financial dimension deters health investments. These results indicated potentially large dividends to providing accurate information. Yet, our survey experiment, in which we provided such information on hypothetical health scenarios shows little effects on beliefs and decisions highlighting the difficulty to resolve these uncertainties.

Speakers: Alina Imping, Lisa Rogge

Parallel Session 5: International 3

96 The (Un)intended Consequences of Export Restrictions: Evidence from Indonesia

Do voluntary export restrictions promote domestic economic development? While an export ban foregoes export revenue, and may result in a permanent loss of market share, it may also boost local processing industries, thus stimulating local employment and manufacturing growth more generally. We address this question in the context of Indonesia, a major producer of raw materials such as nickel and bauxite, which banned the export of unprocessed ores in 2014. We find that the ban had a positive and statistically significant impact on local employment in nickel-producing districts - not only in sectors like manufacturing and construction but also through spillover effects in the public sector and services. This is rationalized by large investments in processing capacity for nickel ore. In contrast, bauxite extraction collapsed, leading to declines in employment and wages in bauxite-producing regions. Without additional refining capacity, the export ban halted mining activity and had unintended negative consequences for economic growth in these areas.

Speaker: Else-Marie van den Herik (Vrije Universiteit Amsterdam)

ExportRestrictions.pdf

97 Falling Trade Barriers and Rising Crime? Evidence from Colombia

Greater openness to trade can create benefits through lower prices, greater product diversity, and higher export earnings. But additional exposure to import competition may reduce income and employment in affected sectors. These negative consequences can increase crime by altering the net benefits of engaging in illicit activities. We analyze the impact of the U.S.-Colombia Free Trade Agreement on the growth of crime in Colombia from 2012 to 2017. Using panel data from 887 Colombian municipalities and a shift-share instrumental variables approach, we find that increased exposure to competition in staple crops led to more home burglaries and robberies of individuals, but fewer business robberies. Crime impacts were con-centrated in the Andes and non-coca-producing regions. Our results suggest that trade liberalization should be accompanied by policies that mitigate negative effects on exposed sectors.

Speaker: Pilar Useche (University of Florida)

FTA_and_crime_summer2024_submitted.pdf

98 Stronger Together? Female Export Consortia

Exporting offers high private and public returns, yet in most countries, only a few large, male-managed firms engage in exporting. We incentivize small, female-managed firms to form a consortium, a corporate group, to test whether they can collectively overcome the fixed export costs. We randomize 176 firms into four sectoral consortia or a control group and study the consortium’s effect on firm performance. After two years, treated firms are 16pp (+63%) more likely to export and double their sales. The results are driven by consortia members expanding and using their business networks more, gaining entrepreneurial confidence, and improving management practices. Yet, cooperation is also costly. Only half of the invited firms ultimately join the consortium, and conflicts arise over mutual effort and joint decision-making. Incentivizing horizontal integration of small firms offers a costeffective, scalable way to share investment costs for export or technology adoption, with the potential to foster broader economic development.

Speaker: Munch Florian (London School of Economics)

JMP Florian Munch.pdf

Parallel Session 5: Political Economy

99 Populism and institutional decay: empirical evidence from the Philippines

This study examines the presidency of Rodrigo Duterte (2016-2022) in the Philippines who became widely known due to his populist agenda and especially his support for the extrajudicial killing of drug users and criminals. To evaluate the impact of his presidency on institutional quality, this study uses data from 2006 to 2022 and the synthetic control method. The results suggest that the presidency of Rodrigo Duterte had a significant and decreasing effect on the rule of law in the Philippines. With the approach used, this study provides counterfactual evidence on how the Duterte presidency has deteriorated institutional quality. Additionally, the study discusses his major policies that led to institutional decay, while also looking at other dimensions of institutional quality. As a large part of his policies aimed at improving economic performance, this study also discusses the short-term economic outcomes of his efforts.

Speaker: Dr Sven Fischer (Philipps-Universität Marburg)

SF_Philippines_4Feb2025.pdf

100 Building Fiscal Capacity with Traditional Political Institutions: Experimental and Qualitative Evidence from Sierra Leone

I argue that governments in weak states can build fiscal capacity by collaborating with non-state, traditional political institutions (TPIs). To study the impact of collaboration, I partnered with the local government in Kono District, Sierra Leone (the KDC), and embedded an experiment within their awareness campaign for a new rural property tax. Property owners in 118 villages were shown videos with varying content. Those in the treatment group viewed an additional segment where their paramount chief discussed the collaboration between the chiefdom government and the KDC in the tax effort. Priming collaboration significantly increased tax compliance and strengthened property owners' belief in their obligation to pay taxes. To assess mechanisms, I developed additional video segments where paramount chiefs emphasized either their coercive capacity or their accountability to constituents. The experimental findings, reinforced by qualitative evidence from 300 interviews, demonstrate that both coercion and accountability are crucial sources of TPIs' authority.

Speaker: Kevin Grieco (International Centre for Tax and Development)

grieco_tpi_March25.pdf

101 Favoritism and Firms: Micro Evidence and Macro Implications

We study the economic implications of regional favoritism, a form of distributive politics that redistributes resources geographically within countries. Using enterprise surveys from low- and middle-income countries, we document that firms located close to leaders’ birthplaces grow substantially in sales and employment after leaders assume office.
Firms in favored areas also experience increases in sales per worker, wages, and measured total factor productivity. These effects are short-lived, and operate through rising government demand in the non-tradable sector. We calibrate a simple structural model of resource misallocation in a two-sector and two-region economy on our estimates. This exercise implies that, despite large firm-level effects, output losses caused by favoritism are small because leaders do not tend to redistribute funds towards less productive regions.

Speaker: Zareh Asatryan

Favoritism_and_firms_20250102.pdf

Parallel Session 5: Public Economics

102 Tax Refund Delays and Firm Performance: Evidence from Zambia

This paper explores and quantifies the inefficiencies induced by delays in tax refund payouts in developing countries. We construct a novel dataset on Value Added Tax (VAT) refunds in Zambia and document extreme delays of more than 700 days on average implying that refund payments take 2300 % longer than they should. Drawing on the universe of firms' tax filings we show that these delays have significant real effects. Firms that receive refunds increase their sales by 13 %. Exploiting random variation induced by a reform of refund administration, we further show that firms whose refund payments amounting to a third of their monthly sales double their investments in the two years following the payment. As refund claims build up, profits decline but do not recover immediately upon refund payout. Our findings highlight how a malfunctioning of the tax system can impede firm growth and investment. It further questions the one-fits-all recommendation of adopting VAT in developing countries with limited state capacity.

Speaker: Daniel Overbeck (University of Mannheim)

103 Beyond the tax bill: Measuring tax compliance costs for Ugandan firms

For low-income countries looking to enhance revenue mobilisation without harming firm growth, understanding the full burden of taxation, beyond just tax liabilities, is crucial. This paper documents the substantial and often regressive tax compliance costs faced by small and medium-sized firms in Uganda. Using original survey data from nearly 2,000 taxpaying firms across Uganda, matched to administrative tax returns data, I show that compliance costs are significant, equivalent to two percent of turnover for the median firm, with smaller firms bearing a disproportionate burden. Moreover, total compliance costs often exceed firms' tax liabilities. Breaking down cost components, I find that labour time spent on tax compliance activities is the largest component, with tax compliance consuming a median of 34 hours of labour time per month, and approximately 20 percent of firm owners' working hours. I show that the majority of firms outsource at least part of their tax obligations to a tax agent, often to compensate for limited tax knowledge. These agents are relatively expensive, costing a median of USD 54 per month. Adopting compliance technologies does not significantly reduce reported compliance costs or time, although firms perceive that technology makes compliance easier. Finally, I use a survey experiment to test how sensitive compliance costs measures are to the measurement strategy, finding significant divergence between estimates obtained through a detailed set of questions and a more aggregate question. Discrepancies between estimates are reduced when respondents are primed with the detailed set of questions first, suggesting that simple aggregate questions might not capture compliance costs in a consistent manner.

Speaker: Adrienne Lees (Institute of Development Studies)

250127 - Tax compliance costs in Uganda - Lees.pdf

104 Fraud Detection Under Limited State Capacity: Experimental Evidence From Senegal

Tax administrations in low-income countries face widespread tax evasion and high enforcement costs. They thus need information to detect where tax evasion is most severe, and allocate scarce resources accordingly. This paper shows that leveraging large firms’ trading network to collect information about their suppliers is a cost-efficient way to detect tax evasion and increase future audit returns. We collaborate with the Senegalese tax administration on a vast data collection effort to digitise lists of payments submitted by the largest firms and show that 88.6% of these firms provide incomplete information about their suppliers. This prevents any cross-checking against income declared by the suppliers themselves. We then randomise a low-cost communication campaign across all 3,487 misreporting firms, to discourage future misreporting. The intervention increases the prevalence of suppliers’ identification information by 52%. In aggregate, this allows to uncover $145.5 million in unreported revenue (i.e. 0.5 % of GDP). Most of it accrues to a few tax-registered suppliers, as opposed to informal ones. A simulation exercise shows that exploiting the newly available information to target the largest under-reporting suppliers would increase audit returns by at least 100%.

Speaker: Dr Mattea Stein (University of Naples Federico II and CSEF)

Czajka-Sarr-Stein CSEF conf version.pdf

Parallel Session 5: Technology

105 Entrepreneurial Presence on Social Media Platforms in Sub-Saharan Africa: Why Language Matters

Seeking evidence on how the presence of a social media platform affects the performance of platform-based entrepreneurs in a developing country context, we conducted a field experiment and a replication study among Zambian entrepreneurs to whom we provided a Facebook business page. To our surprise (and contra our hypotheses), we discovered that the existence of a social media platform reduces sales even though the platform increases market opportunities through wider market reach. Post hoc analyses revealed that the sales reduction was evident only in the case of language dissimilarity. This outcome, we argue, demonstrates that – in the context of a developing country – the highly adaptive nature of transaction processes requires high levels of comprehension so that the transaction conditions will be clearly communicated and understood, and adjusted accordingly. This paper contributes to a deeper understanding of how and why expanding information and communication technologies may disadvantage entrepreneurs looking to flourish in the context of a developing country.

Speaker: Dr Egbert Amoncio (Goethe Universität Frankfurt)

106 Virtual Business Networks and Knowledge Diffusion: Experimental Evidence from Liberia

This paper examines how virtual platforms can extend the benefits of business networks to micro-entrepreneurs in developing economies. Through a randomized control trial in Liberia involving 1,131 entrepreneurs, we study the impact of structured virtual business discussion groups on business outcomes. Entrepreneurs in the treatment group participated in weekly phone-based discussions focused on business challenges, sales strategies, and financial management. Through detailed moderator summaries of each discussion session, we can directly map observed effects to specific topics and business strategies discussed during these meetings. We find that the intervention had a significant effect on entrepreneurs' business practices and strategies. Treatment effects were concentrated in three key areas. First, participants adopted more innovative business practices, experimenting with new marketing approaches and strategically diversifying their sales locations. Second, they were more likely to adopt digital technologies: treatment group members were 46\% more likely to own cellphones, tripled their use of phones for business purposes, and showed significantly higher adoption of mobile money services. Finally, the intervention fundamentally reshaped business networks: treated entrepreneurs shifted away from seeking advice from friends and family, instead building professional relationships with peer business owners--a change that persisted months after the program ended. While we do not find significant effects on short-term profits or revenues, our results suggest that low-cost virtual platforms can effectively facilitate knowledge transfer and business practice adoption among micro-entrepreneurs in developing economies.

Speaker: Golvine de Rochambeau (IFC & CEPR)

BDG_Agarwaletal_feb2025.pdf

107 Social Signaling and the Adoption of New Technologies: Experimental Evidence from Rural Uganda

Speaker: Lotte van der Haar

18:00 Conference Closing