Harnessing the Crowd to Improve Accountability for the Delivery of Public Services

Title	Harnessing the Crowd to Improve Accountability for the Delivery of Public Services
Post date	11/03/2015
C1 Background and Explanation of Rationale	This study aims to generate reliable evidence about the provision and quality of citizen monitoring of public services. Governments around the world are building platforms to collect information from citizens. As part of these programs, agencies increasingly seek to leverage mobile technologies to engage beneficiaries in monitoring at lower costs and with the potential to generate more useful data. Unfortunately, existing research — which tends to be conceptual or descriptive — has provided little guidance about whether mobile technologies can enhance public engagement with government by facilitating regular, high-quality, and useful feedback about the quality of public services. This experimental study of citizen monitoring of solid waste services is uniquely suited to answer foundational questions about citizen engagement with governments as mobile technologies become universally adopted. Because solid waste services are visible, we can verify the quality and timeliness of reports. We can also assess how citizen monitoring affects service delivery as part of a longer-term project.
C2 What are the hypotheses to be tested?	H1a (Provision): More nominated reporters will respond to at least one prompt than randomly recruited reporters. H1b (Provision): Nominated reporters will respond to more prompts than randomly recruited reporters, measured as a count both over the entire 8-week experiment and within individual weeks. H1c (Provision): Nominated reporters will respond to more open-ended prompts than randomly recruited reporters, measured as a count both over the entire 8-week experiment and within individual weeks. H2 (Attrition): Of reporters who respond to at least one prompt in the first two weeks of the experiment, fewer nominated reporters will discontinue reporting than randomly recruited reporters, measured as a lack of reporting for at least two weeks that continues through the end of the 8-week experiment. H3 (Accuracy): Conditional on reporting, nominated reporters will provide information that is closer to independent measures of the severity of solid waste problems than randomly recruited reporters. H4 (Representativeness): Nominated reporters will provide information indicating a larger variance of solid waste conditions than randomly recruited reporters.
C3 How will these hypotheses be tested? *	Provision of reporting. Differences in means for the measures of reporting for H1-H2 will be estimated using randomization inference under the sharp null hypothesis assumption and an exact replication of our randomization procedure. Quality of reporting. If a group of people reports the same objective conditions accurately, we will observe low variance of reported outcomes for the same objective conditions. We exploit this implication to estimate the accuracy of reporting through the KCCA platform. For each reporting outcome, we will divide the reporters according to their recruitment method. We will then run a nonparametric kernel regression with automatically chosen bandwidths to model the reporting outcome as a function of the objective outcomes measured during the baseline household surveys and audits. We will compare the mean squared error of the nominated reporters versus the randomly recruited reporters. We will then form a sampling distribution of difference in mean squared errors using randomization inference by repeatedly fitting such nonparametric models while exactly replicating our randomization procedure.
C4 Country	Uganda
C5 Scale (# of Units)	1080
C6 Was a power analysis conducted prior to data collection?	Yes
C7 Has this research received Insitutional Review Board (IRB) or ethics committee approval?	Yes
C8 IRB Number	ESMS-BU-MA-031-5N
C9 Date of IRB Approval	6/3/2015
C10 Will the intervention be implemented by the researcher or a third party?	Researchers, The recruitment will be implemented by the research team. The experimental prompts for feedback will be implemented by the Kampala Capital City Authority.
C11 Did any of the research team receive remuneration from the implementing agency for taking part in this research?	No
C12 If relevant, is there an advance agreement with the implementation group that all results can be published?	not provided by authors
C13 JEL Classification(s)	not provided by authors