Mobile Phones Enable Voters to Hold Local Officials Accountable In Ugandan Elections (With Sarah Bush, Ryan Jablonski, Daniel Nielson, and Paula Pickering). Proceedings of the National Academy of Sciences, forthcoming.
Abstract — Many politicians manipulate information to prevent voters from holding them accountable; however, information communication technology (ICT) may make it easier for non-governmental organizations to credibly share information that is difficult for politicians to counter directly. We demonstrate the potential for ICT to improve democratic accountability by conducting a large (n=16,083) randomized controlled trial during the 2016 Ugandan district elections. In cooperation with a local partner, we compiled, simplified, and shared official information on irregularities in local government budgets via mobile text messages. Verified recipients of messages that described more irregularities than expected reported voting for incumbent councillors 6 percent less often; verified recipients of messages conveying fewer irregularities than expected reported voting for incumbent councillors 5 percent more often. The messages had no significant effect on votes for incumbent council chairs, potentially due to voters’ greater reliance on other sources of information in elections for higher-profile politicians. Our results point to the promise of ICT for empowering voters and holding politicians accountable, even in settings where elections are not free and fair.
Can Information Outreach Increase Participation in Community-Driven Development? A Field Experiment near Bwindi National Park, Uganda (With Brigham Daniels and Colleen Devlin). World Development, forthcoming. [ Final | PDF ]
Abstract — Decentralization and community-driven development intend to bring public decisions closer to the people, yet elites often capture local institutions. One way that local elites capture community-driven development is to limit information about opportunities for citizens to shape group decisions. We investigate whether sending citizens targeted and timely information about when and how they can participate in the planning of community-driven development projects increases knowledge, participation, and satisfaction with local institutions. We implemented a pre-registered randomized field experiment in partnership with the Uganda Wildlife Authority that involved sending residents in randomly selected villages near Bwindi National Park approximately 60 messages by mobile phone over eight months about how a park-sponsored revenue sharing program worked and how and when residents could participate. We do not find evidence that the information increased perceived knowledge, participation, perceived efficacy, or satisfaction with local institutions. Exploratory findings suggest that among women, who are often disenfranchised in Ugandan society, the information treatment backfired. More positively, we find that reaching more people in a community with information lead to promising results. We conclude that informational treatments are unlikely to empower participation on average, unless they are deployed broadly and in ways that promote collective action.
Preferences for Domestic Action Over International Transfers in Global Climate Policy. Journal of Experimental Political Science, forthcoming (With Lauren Prather) [ Preprint ]
Abstract — Cost-effective and equitable climate change mitigation requires the transfer of resources from developed to developing countries. In two behavioral experiments, we demonstrate that American subjects act according to a strong home preference, by making private donations and writing letters in support of public spending more often for mitigation programs located at home versus those overseas. We attempt to overcome the preference to act at home by randomly informing some subjects that foreign programs are more cost-effective than domestic programs. Only in the case of private donations is home preference ever mitigated. From a separate experimental treatment, we show that the preference against foreign programs is exacerbated when the co-benefits of mitigation programs are made salient. Importantly, home preference crosses party lines, indicating that it is a deep-seeded, affective preference. These findings highlight significant political obstacles to international cooperation on climate change that relies on transfers.
Aiming at the Wrong Targets: The Difficulty of Improving Domestic Institutions with International Aid. International Studies Quarterly, forthcoming (With Benjamin P. Buch and Bradley C. Parks) [ Final | Preprint | Replication ]
Abstract — We explain why international development organizations have had so little success building and reforming public sector institutions in developing countries, despite the extraordinary amounts of time, money, and effort that they have devoted to this enterprise and an apparently strong commitment to achieving measurable results. We demonstrate that, when access to financing is made conditional upon achievement of performance targets, countries tend to choose easy and shallow targets that measure how public sector institutions are organized. These targets often do not focus attention on solving public problems through institution building, but instead provide countries with low-cost opportunities to signal commitment to institution building to international development organizations. We demonstrate that this argument has both explanatory and predictive power for a sector of World Bank lending — environment and natural resource management — that focuses heavily on improving public sector institutions.
Titling Community Land to Prevent Deforestation: An Evaluation of a Best-Case Program in Morona-Santiago, Ecuador. 2015. Global Environmental Change 33: 32-43 (With Stuart E. Hamilton and Marco Millones). [ Final | PDF | Replication Files | Presentation ]
Assigning land title to collective landholders is one of the primary policies land management agencies use to avoid deforestation worldwide. Such programs are designed to improve the ability of landholders to legally exclude competing users and thereby strengthen incentives to manage forests for long-term benefits. Despite the prevalence of this hypothesis, findings about the impacts of land titling programs on deforestation are mixed. Evidence is often unreliable because programs are targeted according to factors that independently influence the conversion of forests. We evaluate a donor-funded land titling and land management program for indigenous communities implemented in Morona-Santiago, Ecuador. This program offers a close to best case scenario for a land titling program to reduce deforestation because of colonization pressure, availability of payments when titled communities maintain forests, and limited opportunities for commercial agriculture. We match plots in program areas with similar plots outside program areas on covariates that influence forest conversion. Based on matched comparisons, we do not find evidence that land titling or community management plans reduced forest loss in the five years following legal recognition. The results call into question land titling as a direct deforestation strategy and suggests land titling is better viewed a precursor to other programs.
Abstract — Both states and civil society groups face challenges in holding international organizations accountable for achieving their mandates. States often have difficulty monitoring the actions of international organizations. Civil society groups often have limited ability to directly influence decision-making at international organizations. States have created accountability mechanisms at international organizations to overcome these challenges. Accountability mechanisms allow civil society groups to submit complaints about the performance of international organizations. While ombudsmen offices, accountability panels, and complaint mechanisms have grown more common, it is unclear when and why these mechanisms can constrain behavior by international organizations that runs counter to the mutual interest of states and civil society groups. Using the World Bank Inspection Panel as a test case, I show that civil society monitors change the lending behavior of the World Bank when they enhance oversight by powerful states with similar interests. By combining their abilities in sanctioning and monitoring, states and civil society groups are able to more effectively govern international organizations.
Abstract — A large portion of foreign assistance for climate change mitigation in developing countries is directed to clean energy facilities. To support international mitigation goals, however, donors must make investments that have effects beyond individual facilities. They must reduce barriers to private-sector investment by generating information for developers, improving relevant infrastructure, or changing policies. We examine whether donor agencies target financing for commercial-scale wind and solar facilities to countries where private investment in clean energy is limited and whether donor investments lead to more private investments. On average, we find no positive evidence for these patterns of targeting and impact. Coupled with model results that show feed-in tariffs increase private investment, we argue that donor agencies should reallocate resources to improve policies that promote private investment in developing countries, rather than finance individual clean energy facilities. We suggest that international negotiations could usefully shift the focus of climate change finance towards adaptation in exchange for mitigation-improving policy reforms in developing countries.
When Do Environmentally-Focused Assistance Projects Achieve their Objectives? Evidence from World Bank Post-Project Evaluations. 2013. Global Environmental Politics 13(2): 65-88 (with Bradley C. Parks). [ Final | Preprint | Replication Files ]
Abstract — Scholars and practitioners alike have paid a great deal of attention to the factors that promote successful outcomes in environmentally-focused assistance projects. Previous studies have identified a number of potentially important predictors of successful outcomes, including the political commitment, institutional capacity, governance quality of the recipient country, the severity of environmental pressures in the recipient country, donor-recipient contracting dynamics, project characteristics, and civic participation in the recipient country environment sector. We compile post-project outcome ratings for all environmentally-focused aid projects approved since 1994 by the World Bank and use the resulting dataset to submit hypotheses about successful donor projects to a general empirical test. We find that strong public sector institutions in the borrowing country and proactive staff supervision contribute to satisfactory project outcomes. Our results also suggest that projects seeking to achieve global environmental objectives are less likely to succeed. We argue that future research will be most fruitful if it focuses on how the operational and management characteristics of individual projects lead to successful outcomes.
Does the Asian Development Bank Respond to Past Environmental Performance When Allocating Environmentally-Risky Financing? 2011. World Development 39(3): 336-350. [ Final | Preprint | Data | Source Data | Code ]
Summary — This paper provides an empirical test of whether the Asian Development Bank (ADB) adjusts allocation decisions about environmentally-risky projects to reflect borrower environmental performance in previous projects. This type of performance-based decision-making has been repeatedly highlighted as key to achieving favorable development assistance outcomes in a variety of programming areas. I collect recipient environmental performance information from all available post-project evaluations since 1990 and create an indicator of environmental reputation using a Bayesian updating model. I use this environmental reputation indicator to demonstrate that the ADB responds to previous borrower environmental performance when approving environmentally-risky projects, but that past environmental performance does not positively influence the allocation of projects with no environmental risks. These results demonstrate that performance-based allocation decisions are possible for development organizations within specific programmatic areas when low performance is a significant risk to the core functions of the organization, which in this case is the ability to approve and disburse lending projects.
Abstract — Alpine areas in northwestern Yunnan, China possess globally significant levels of biodiversity and are important locally for livelihood activities such as livestock grazing and medicinal plant collection. Because local land use has important impacts on alpine conditions and communities have significant capacity to manage alpine resources, we emphasized local collaboration during the initial stages of conservation planning. Our collaboration with local communities investigated how livelihood strategies affect the condition of alpine resources in northwestern Yunnan and how future conservation efforts can be compatible with local livelihoods. We sampled three livestock herding sites, each within a different alpine sub-region, using open-ended interviews and maximum variation sampling. According to interviewees, livestock grazing within the alpine zone currently does not appear to be negatively impacting the availability of forage. Medicinal plant collection, however, is showing unsustainable trends. Tourism is as yet a nascent industry, but is seen as having great potential by those interviewed. It is clear that with increases in population, access to regional markets, and tourism, northwestern Yunnan’s rich alpine resources will require careful management. In addition to the data collected, we found that the methodology used may be widely applicable to organizations with limited resources that wish to engage local communities during the formative stages of regional-level conservation planning.