Brief 73: Adoption of Community Monitoring Improves Common Pool Resource Management Across Contexts

BrazilChinaCosta RicaLiberiaPeruUganda
Contextual features of CPR
ResourceGroundwaterSurface waterGroundwaterForestForestForest
CommunityRural villagesUrban microneighborhoodsRural villagesVillagesIndigenous communitiesVillages
Primary threat to resourceDrought, overuseIndividual, industrial pollutionDrought, overuseOvercutting by residentsExtraction by outsidersOvercutting by residents
Components of harmonized interventions
Community workshops
Monitor selection, training, incentives
Monitoring of the resource
Dissemination to citizens
Dissemination to management bodies(Alternate arm)***
Experimental design
Alternate treatment armConservation plan makingDissemination to governmentNegotiation trainingSMS reminders
Experimental designThree-arm2×2 factorialTwo-arm2×2 factorialTwo-armThree-arm
No. of monitoring communities (NM)808081603960
No. of nonmonitoring communities (N¬M)408080603750
Common outcome measurement
Duration of implementation, mo121512121312
Primary compliance measureSMS reports receivedDissemination postersReports submittedMonitoring walks completedReports submittedReports submitted
Primary resource outcomeWell electricity usagePollutant concentration in waterWell electricity usage, water qualityDeforestationDeforestationDeforestation, forest quality
Endline citizen survey
Table 1. Features of the research contexts and experimental designs. NM denotes the number of communities assigned to any treatment condition with community monitoring, and N¬M denotes the number assigned to any treatment condition without community monitoring.
* In the forest studies, the community constitutes at least one of the possibly overlapping management bodies.
 In both three-arm designs, communities assigned to the alternative treatment arm received both monitoring and the alternative treatment.
  1. Introduction: The monitoring program was introduced through community workshop(s) led by implementing partner organizations, usually NGOs, and community members
  2. Selection: Individual monitors were then selected and trained to monitor CPR
  3. Technology: Each program used some form of technology to aid in monitoring or dissemination of findings (e.g. water-level sensors in Brazil)
  4. Incentives: Monitors were given locally-appropriate monetary incentives to provide either monthly or quarterly reports for at least one year
  5. Dissemination: Data from monitors was disseminated to citizens through fliers and/or community meetings, and to at least one management body overseeing the CPR.
Figure 1. Proportion of communities engaging in monitoring, by quarter (3 months) of the intervention. Note that monthly monitoring was incentivized in Brazil, Costa Rica, Peru, and Uganda. Records of monitoring in these sites are collapsed to quarters (three month periods) for comparability with the records from China and Liberia.
  1. Resource Use: Community monitoring reduced the extraction of community pooled resources. Extraction is measured by forest loss in Liberia, Peru and Uganda; water pollution in China and Costa Rica; and water use in Brazil and Costa Rica. Importantly, monitoring interventions yield the highest effects where resource extraction problems were the most acute.
  2. Satisfaction: Though community members reduced their short-term use of resources, they reported high levels of satisfaction with the uptake of community monitoring.
  3. Knowledge: Monitoring increased knowledge, including in areas such as issues related to CPR, the level of CPR in communities, and causes of resource degradation. Monitoring also increased knowledge of CPR management authorities
  4. Stewardship:Monitoring did not have a consistent effect on resource stewardship. Monitoring increased stewardship in Liberia while decreasing it in Peru. The authors attribute these divergent outcomes to different existing institutions for CPR management.
Figure 2. Estimated site-level Intention-to-treat effects (top panels) and mean Intention-to-treat effects (µ) across sites (bottom panels) for each of the main hypotheses. The thin segments represent 95% confidence intervals. The thick segments indicate the direction of the pre-specified one-tailed hypothesis; where these segments do not bound zero, we reject the null hypotheses at the α = 0.05 level.