Accountability and Transparency in the Mining Sector of Peru

800 households from 80 communities
2016 - 2016
Target group:
  • Rural population
Outcome of interest:
  • Transparency and accountability
  • Corruption and Leakages
Intervention type:
  • Community participation
  • Information
  • Natural resource management
AEA RCT registration number:

Public training sessions on democratic processes and ideals are a popular tool that aims to improve the performance of governments with “bottom-up” accountability via increased political knowledge and public participation. Researchers evaluated an accountability workshop program, which educated citizens on the distribution of extractive industry tax revenues and the formal means of local political participation. The program had mixed results on local government quality, increasing the likelihood of initiating recall against poor-performing mayors but also increasing civil unrest. The workshops also reduced involvement in participatory budgeting, as well as reducing mayors’ discretionary spending efforts.

Policy issue

How can governance be improved in developing countries? Training sessions for the public on democratic processes and ideals are an increasingly popular tool that aims to improve the performance of democratic governments with “bottom-up” accountability via increased political knowledge and public participation. These trainings could support better governance in several ways. Public education programs could signal to citizens the value of participating in democratic processes. Second, trainings could help citizens learn about the political process and the performance of their elected officials. Third, the meetings may stimulate opportunities for voters to engage in collective political action. However, there are also reasons why providing political information to citizens may not be effective. In a context with poor governance and weak institutions, voters may lose confidence in democratic institutions, leading to disengagement or spurring civil unrest. Poor performing politicians may even perform worse if they believe it too late in their term to satisfy constituents and consequently stop working.

Context of the evaluation

This study was implemented in partnership with Propuesta Ciudadana (PC), a Peruvian NGO consortium that promotes the role of civil society and local government in improving democracy, promoting equity, and supporting social justice. This evaluation focused on three means of important local political engagement in Peru: participatory budgeting, recall of elected officials, and political protests and demonstrations.

Extractive industries represent a major part of the local economy in many rural parts of Peru. By law, the central government must return a portion of royalties and tax revenues from extractive industries to the local governments where the mines are located. Regional governors and mayors are then responsible for implementing development projects using these funds. Each year, municipal governments hold “participatory budgeting” meetings. Communities vote on potential projects and policy priorities, but officials are not legally bound to follow the citizen’s preferences.

Mayors are elected every four years, but citizens can initiate a recall half way through their term. Citizens wishing to establish a recall vote against a mayor must first purchase a “recall kit,” then collect supporting signatures from 25 percent of registered voters. If a recall vote takes place, a mayor is removed from office if 50 percent of ballots cast are in favor of the recall. Recall attempts are common; since 1993 about 5,800 recall kits have been purchased, and 1,700 mayors have been recalled.

In addition to formal politics, Peru has a lively tradition of direct action and protest, including demonstrations against government actions, and protests against private firms (especially in the mining sector). From 2007 to 2014, there were 186 riots and protests against mining, with both private firms and the government targeted by direct action.

Details of the intervention

In collaboration with Propuesta Ciudadana (PC), researchers conducted a randomized evaluation to assess the effect of an accountability workshop program that provided information about the distribution of royalties from extractive industries as well as training on the participatory budgeting and mayoral recall processes. Researchers randomly assigned 80 communities in Piura to a treatment group (40 communities) that were invited to attend the workshops or a comparison group (40 communities), which were not invited to any workshops.

The half-day workshops involved a training video and a facilitated group discussion. The training video provided participants with information about oil and gas production in Piura, the main revenue redistribution mechanism, the process that local governments follow to allocate and spend revenues, and general information about the amount of funding received by different district governments in Piura. In addition, the video highlighted the participatory budgeting process used to select projects for government funding and related accountability meetings where citizens can request details about the progress of projects. Lastly, the video explained the process for carrying out a recall against a district mayor or departmental president, a means of democratically removing an elected official from office. Across treatment communities, 691 people attended the workshops.

Researchers used household surveys with local leaders (identified by the village president in each study community) to evaluate the effect of the program on protests, violence, and political accountability. In addition, researchers used administrative data from the national government of Peru to measure the number of mayoral recall attempts, the attendance at participatory budgeting meetings, the portion of municipal budgets spent (budget execution) and the number of protest actions and civil unrest events in each district.

Results and policy lessons

Accountability workshops had mixed effects on government performance, which suggests that in complicated strategic political environment, information sometimes improves and sometimes undermines citizen participation and ultimately politician performance in office.

Participatory budgeting: The workshops increased knowledge of participatory budgeting, but decreased citizen participation. The workshops increased village leaders’ knowledge of the details of the participatory budgeting process by 5 percentage points. In addition, among village leaders surveyed, individuals who attended the workshops were 33 percentage points (38 percent) less likely to attend participatory budgeting meetings, compared to an 86 percent participation rate in comparison areas that did not hold workshops. At the district level, workshops reduced the number of members in participatory budget committees by 51 percent.

Recalls: The workshops increased knowledge of and incidents of recalls. The workshops increased village leaders’ knowledge of the details of the recall process by 8 percentage points. In districts with low performing mayors (as measured by the share of the budget allocation implemented), workshops increased the probability of initiating a recall by 31 percentage points (a 70 percent change).

Civil unrest: Similar to the impacts on recalls, in districts with low performing mayors, workshops increased individuals’ support for protests by 32 percentage points. There was no impact on individuals in districts with mayors performing at an average or high level. At the district level, workshops increased the probability of civil unrest by 7 percentage points (a 54 percent change).

Together, the results suggest that the information provided in workshops increased participants’ disillusionment with formal political institutions and led them to disengage from personally costly processes like participatory budgeting. Although the intervention increased the initiation of recalls for poor-performing mayors, these mayors responded to the recall threat by further reducing their effort, spending less of their discretionary budget.