Research

Job Market Paper


Local Journalism and Government Responsiveness: Evidence from Tanzania pre-analysis plan | paper

Does local journalism improve government responsiveness? Independent journalism is often considered a cornerstone of accountable governance, but its impact is unproven outside of developed democracies. I argue local journalism improves service delivery in developing countries by helping central governments monitor district-level bureaucrats. I test the argument through a national-scale randomized controlled trial of local investigative journalism in Tanzania. In collaboration with 15 regional radio stations, I identified 206 communities suffering from service delivery problems like flooded roads, broken water points, absent teachers, and missing medical supplies. I then randomly assigned 103 communities to the treatment group and 103 communities to a pure control condition. Journalists from local radio stations visited treatment communities and investigated the source of the service delivery problem, broadcast 20 minute news stories about their findings on regional radio, and conducted follow up investigations several months later. Independent auditors visited all communities 7 months after the investigations were broadcast and scored changes in the targeted service delivery problem. I find that treated communities received higher audit scores on average (coefficient = 0.25 standard deviations, randomization inference p-value = 0.026), amounting to one road or water point repair in every four treated communities. Investigative reports by local journalists spurred responses by un-elected government ministries but not citizens, local bureaucrats, or members of parliament. This study offers the first experimental evidence that local journalism improves government performance outside of a developed democracy.

Dylan W. Groves


Published and Forthcoming Papers


The Effects of Independent Local Radio on Tanzanian Public Opinion: Evidence from a Planned Natural Experiment (Forthcoming, Journal of Politics) project website

We describe a natural experiment occasioned by an abrupt increase in the transmission range of an independent Tanzanian radio station whose broadcasts emphasize current affairs and gender equality. Some villages that formerly lay outside the catchment area of this radio station could now receive it, while nearby villages remained outside of reception range. Prior to the change in transmitter range in 2018, we conducted a baseline survey in both treated and untreated villages and found them to be similar in terms of prevailing social attitudes and political interest. An endline survey conducted in 2020 shows that respondents in areas that received the new radio signal were substantially more likely to listen to the station, and their levels of political interest and knowledge about domestic politics were significantly higher than their counterparts in villages where the signal could not reach. Attitude change on a range of gender issues, however, was sporadic.

Donald P. Green, Dylan W. Groves, Constantine Manda, Beatrice Montano, and Bardia Rahmani


Perspective Taking Through Partisan Eyes: Cross-National Empathy, Partisanship, and Attitudes Towards International Cooperation (Forthcoming, Journal of Politics) pre-analysis plan | paper | replication materials (forthcoming)

How does cross-national empathy influence public attitudes towards international cooperation? Few studies have considered whether the capacity to see the world from the perspective of other actors promotes international cooperation or how partisanship may condition empathy’s influence. In this paper, we argue that cross-national empathy increases support for international agreements because seeing issues through the eyes of other states expands the range of considerations that individuals use in forming their attitudes. However, partisan attachments undercut this effect. Across three waves of an original survey experiment covering 6,292 respondents, we find that cues to “step into” the perspective of other states modestly increase aggregate support for international cooperation. But this effect is concentrated entirely among those with weak partisan attachments, regardless of the issue area (climate change or nuclear nonproliferation) and potential partner country (China or India). Our results speak to both the promises and shortcomings of empathy as a device for encouraging cooperative outcomes in international affairs.

Don Casler and Dylan W. Groves


A Radio Drama’s Effects on Attitudes Toward Early and Forced Marriage: Results from a Field Experiment in Rural Tanzania (Forthcoming, Comparative Political Studies) pre-analysis plan | paper | replication materials (forthcoming) | project website

Early and forced marriage (EFM) is an increasing focus of international organizations and local non-government organizations. This study assesses the extent to which attitudes and norms related to EFM can be changed by locally tailored media campaigns. A two-hour radio drama set in rural Tanzania was presented to Tanzanian villagers as part of a placebo controlled experiment randomized at the village level. A random sample of 1,200 villagers was interviewed at baseline and invited to a presentation of the radio drama, 83% of whom attended. 95% of baseline respondents were reinterviewed two weeks later, and 97% fifteen months after that. The radio drama produced sizable and statistically significant effects on attitudes and perceived norms concerning forced marriage, which was the focus of the radio drama, as well as more general attitudes about gender equality. Fifteen months later, treatment effects diminished, but we continue to see evidence of EFM-related attitude change.

Donald P. Green, Dylan W. Groves, Constantine Manda, Beatrice Montano, and Bardia Rahmani


Community-based Rangeland Management in Namibia Improves Resource Governance but not Environmental and Economic Outcomes (Nature - Communications Earth & Environment, February 2022) pre-analysis plan | paper | replication materials | project website | press coverage

Classic theories suggest that common pool resources are subject to overexploitation. Community-based resource management approaches may ameliorate tragedy of the commons effects. Here we use a randomized evaluation in Namibia’s communal rangelands to study a comprehensive four-year program to support community-based rangeland and cattle management. We find that the program led to persistent and large improvements for eight of thirteen indices of social and behavioral outcomes. Effects on rangeland health, cattle productivity and household economics, however, were either negative or nil. Positive impacts on community resource management may have been offset by communities’ inability to control grazing by non-participating herds and inhibited by an unresponsive rangeland sub-system. This juxtaposition, in which measurable improvements in community resource management did not translate into better outcomes for households or rangeland health, demonstrates the fragility of the causal pathway from program implementation to intended socioeconomic and environmental outcomes. It also points to challenges for improving climate change–adaptation strategies.

(* denotes equal authorship): D. Layne Coppock*, Lucas Crowley, Susan L. Durham, Dylan Groves*, Julian C. Jamison*, Dean Karlan*, Brien E. Norton, R. Douglas Ramsey


A Radio Drama’s Effects on HIV Attitudes and Policy Priorities: A Field Experiment in Tanzania (Health Education and Behavior, May 2021) pre-analysis plan | paper | replication materials

A growing body of evidence investigates how entertainment education influences knowledge about HIV, stigma toward those with HIV, and openness to disclosing one’s HIV status. The present study shows that in addition to these effects, mass media interventions may influence audiences’ policy priorities, such as their demand for local access to HIV/AIDS medical care. A condensed (2 hours) version of a popular Swahili radio drama was presented to rural Tanzanians as part of a placebo-controlled experiment, clustered at the village level. A random sample comprising 1,200 participants were interviewed at baseline and invited to attend a presentation of the radio drama, and 83% attended. Baseline respondents were reinterviewed 2 weeks later with a response rate of 95%. In addition to increasing listeners’ knowledge and support for disclosure of HIV status, the radio drama produced sizable and statistically significant effects on listeners’ preference for hypothetical candidates promising improved HIV/AIDS treatment.

Donald P. Green, Dylan W. Groves, and Constantine Manda


The Mismeasurement of Cattle Ownership in Namibia's Northern Communal Areas (Nomadic Peoples, 2020) paper

The standard approach to measuring livestock ownership in pastoralist communities relies on an assumption of uniformity that does not reflect the diverse concepts of ownership held by pastoralists themselves. In Namibia's Koakaveld Region, Himba and Herero pastoralist communities have a rich vocabulary for categorising the origins, usage rights and cultural valence of their cattle. Drawing on both authors' experience overseeing a large-scale rangeland management programme evaluation in Namibia's Northern Communal Areas – and one author's experience growing up in and keeping cattle in a Himba pastoralist community – we show how the standard approach to measuring cattle ownership undermines accurate estimates of livestock wealth, off-take and inequality, and obfuscates pastoralist's strategies for turning ecological variability to their advantage. We conclude with lessons about how multi-dimensional data collection methods improve upon the standard approach to livestock ownership measurements.

Dylan W. Groves and Venoo Tjiseua


The Influence of Two Levels of Debushing in Namibia's Thornbush Savanna on Overall Soil Fertility, Measured Through Bioassays (Namibian Journal of the Environment, 2017)

A healthy and productive rangeland depends on well-functioning ecosystem services such as effective cycling of water and nutrients. After rangeland has degraded, bushes may encroach in nature's attempt to restore water and nutrient cycling. When bush encroachment is addressed by debushing, with harvested bush wood sold off the land, then nutrient cycling is disrupted and soil fertility is likely to decline. Former debushing activities on different parts of farms of the Cheetah Conservation Fund in central Namibia provided the opportunity to assess the influence of debushing on overall fertility of soil. Sites were selected on the farm representing nine examples of each of uncleared, partially cleared and totally cleared land. The debushing had taken place at different times, varying between two and 13 years previously. Soil was collected from each of these 27 sites and subjected to bioassay by growing barley (Hordeum vulgare) and Moringa oleifera. Seedling emergence and height at five weeks for both species were greatest in uncleared soil and lowest in totally cleared soil, indicating the loss of soil fertility as debushing intensifies. There was no evidence of restoration of soil fertility, even 13 years after debushing. Nutritious grass is unlikely to grow well after debushing, and more bush is likely to regrow in nature's attempt at restoring fertility over the long term. If faster restoration is sought, then the full spectrum of minerals removed in harvested wood should be replaced on the land.

Ibo Zimmermann, Mattis Nghikembua, D Shipingana, Dylan W. Groves, T Aron, L Marker


Under Review


The Persuasive Power of Supreme Courts: Evidence from Tanzania (Under Review)

Do constitutional court rulings shape social attitudes and perceived social norms? Rigorous evidence from outside of industrialized democracies is scarce. This paper evaluates the influence of a recent ruling by Tanzania's supreme court which held that laws permitting marriage for girls under the age of 18 are unconstitutional. We randomly assigned 1,950 respondents in rural Tanzania to hear a radio report about the ruling or to a control condition. Respondents who heard the report were 8 percentage points more likely to reject all forms of early marriage and 5.4 percentage points more likely to say they would report early marriage to authorities. However hearing the report did not influence perceptions of community norms or willingness to speak out against early marriage, and the persuasive effect of the report was attenuated when presented alongside a countervailing signal from Tanzania's Attorney General.

Salma Emmanuel, Dylan W. Groves, Constantine Manda, Donald P. Green, Beatrice Montano, Bardia Rahmani


Working Papers

Information or Motivation? Local Media and Local Government Responsiveness in Tanzania

Do local government officials fail to respond to their constituents because they are uninformed or unmotivated? In many low-income countries, local governments fail to deliver the public goods and services that are most important to their constituents. I conducted two randomized experiments with 340 local government officials across 109 Tanzanian villages to evaluate competing explanations for local government unresponsiveness. In the ``information'' experiment, enumerators provided 180 randomly selected village leaders with accurate information about their constituents' social attitudes and policy priorities. In an overlapping ``motivation'' experiment, I partnered with local journalists to inform political leaders in 55 randomly selected villages that the journalists would broadcast reports about leaders' responses to pre-specified village priorities. I collected survey measures of leader priorities and a costly behavioral measure of leaders' prioritization of the targeted development issue. I find that the information treatment influenced leaders' self-reported prioritization of the targeted issue in the short term (7 pp, RI p-value = 0.02) and increased the likelihood that leaders submitted a budget request to the District council one month later (5 pp, RI p-value 0.04). I observe weaker evidence for the motivation treatment on leader behavior, and no evidence for an interaction between the information and motivation treatment.

Dylan W. Groves


Seeing Cattle like a State: Animal Tracing System and State-Pastoralist Relations (invited submission for Nomadic Peoples)

Livestock tracing systems are an increasingly prominent component of national livestock development policies around the world. In principle, livestock tracing systems allow governments to track and respond to disease, livestock theft, and ecological vulnerability. In practice, livestock tracing systems are suffused with sedentist assumptions, and their application to pastoralist communities has had a range of unanticipated deleterious consequences. Drawing on qualitative interviews with policymakers, implementing bureaucrats, and affected pastoralist communities, we review the sedentist assumptions that animate the Namibian Livestock Identification and Traceability System (NamLITS) and trace how those assumptions have affected the economic, social, and political lives of pastoralist communities. Finally, we show the range of strategies that mobile peoples have deployed to circumvent livestock tracing programs, and conclude with lessons for governments and development practitioners considering livestock tracing systems as well as mobile communities affected by them.

Dylan W. Groves and Venoo Tjiseua


Who is Susceptible to Media Influence? Applying Machine Learning to a Multi-Message Research Design in Tanzania

A growing research literature shows that mass media can meaningfully influence politically relevant attitudes and behaviors, but little attention has been paid to treatment effect heterogeneity. Are some people more susceptible to media influence than others? Is susceptibility to influence message-specific, or are certain people highly susceptible to influence across messages? To assess whether susceptibility to influence manifests across messages, we implement a two-message experimental placebo design that exposes Tanzanian villagers to persuasive radio dramas about either forced marriage or HIV stigma. Finding substantial average effects on attitudes, political priorities, and voting preferences, we use machine learning to detect heterogeneous effects across messages as a function of respondents' background attributes. Our results indicate that the factors that predict treatment responsiveness are weakly correlated across these messages. These results suggest that susceptibility to influence is largely message-specific.

Donald P. Green, Dylan W. Groves, Constantine Manda, Beatrice Montano, Bardia Rahmani


Religious Elite Messaging and Women's Political Participation

Can religious leaders use their standing to promote women's political participation (WPP) in the developing world? We present the results of a survey experiment conducted in northeastern Tanzania that estimates the effect of a pro-WPP audio message from a progressive religious elite on behavioral intentions, attitudes, and norms relating to female office-seeking. Across two studies, we find that the progressive religious elite's message makes villagers more likely on average to say they would encourage their daughter or niece to run for political office. These effects on behavioral intentions persisted a month later. Moreover, we find suggestive evidence that the children of respondents who received the progressive elite's message became more interested in running for political office. By contrast, we find muted or inconsistent effects of the progressive message on attitudes and norms, and of conservative messages on all three outcomes. The results suggest that interventions that leverage local religious elites can effectively reduce familial gatekeeping of women's political participation in socially conservative communities.

Bardia Rahmani, Donald P. Green, Dylan W. Groves, Beatrice Montano, and Francis Ngatigwa


Scarcity and Collective Action: Evidence from a Severe Drought in Namibia

Political outcomes ranging from mass migration to leadership change to violent conflict have been linked to adverse climatic conditions associated with climate change. While large scale cross-country panel datasets can provide substantiation of such relationships, relatively little rigorous evidence is available to identify micro-level behavioral and collective responses that could help to explain these links. In this paper we explore the micro-level impact of climate change by pairing an exogenous climatic shock -- a severe and spatially-varying drought -- with “lab-in-the-field” behavioral games and surveyed measures of collective action from more than 1,000 livestock managers in 123 grazing areas across northern Namibia. First, we leverage repeated measurements of the same livestock managers over time to test whether play in public goods games differs between drought and non-drought conditions. Second, in the same villages, we explore how observed measures of collective behavior, including participation in water resource management and community grazing institutions, vary with the experience of drought. Initial results provide evidence that, in this context of acute vulnerability to climate change, severe climatic shocks tend not to affect but to increase people’s propensity to act pro-socially and collectively. Where present, such counter-veiling collective responses to the devastation of drought may enhance resilience in the face of climate change and limit the rise of disruptive future political outcomes, including mass migration to conflict.

Garrett Adler and Dylan W. Groves


A Radio Drama's Effects on Environmental Protection: Experimental Evidence from Tanzania

We report the results of an experiment conducted in rural Tanzania in which 1,360 respondents from 34 villages were randomly assigned to attend a screening of one of two radio dramas. The treatment drama depicts a corrupt bargain between a business developer and a public official to sell off the natural resources of a fishing village; the hero of the story rallies villagers to reject the deal in favor of environmental preservation. A placebo drama focused on an unrelated topic. Outcomes were assessed through a survey conducted four weeks later with 98.5\% of original participants. Participants who were randomly exposed to the treatment drama became more knowledgeable about climate change, more likely to cite environmental protection as a political priority, and substantially more supportive of pro-environmental policies and candidates. No effects were found on other outcomes, such as respondents’ proclivity to engage in or punish illegal logging.

Bardia Rahmani, Dylan W. Groves, Beatrice Montano, and Donald P. Green, Beatrice Montano


Works in Progress

Afya Yako: A National-Scale Field Experiment on Local Media and Public Health Services in Tanzania

with Salma Emmanuel, Constantine Manda, and John Marshall (Preparation)

The Effects of Mass Media on Political and Social Attitudes: Evidence from a Radio Distribution Experiment in Tanzania

with Donald P. Green, Beatrice Montano, and Bardia Rahmani (Intervention ongoing)

Household and Community Spillover Effects of Persuasive Messaging: Experimental Evidence from Tanzania

with Donald P. Green, Beatrice Montano, and Bardia Rahmani (Analysis complete)

A Radio Drama's Effects on Gender Based Violence Awareness and Responses: Experimental Evidence from Tanzania

with Donald P. Green, Beatrice Montano, and Bardia Rahmani (Analysis complete)