Three essays on the public participation and government performance

Public participation in governance has been perceived as a way to address growing dissatisfaction with how democracy works in practice and the demands for more dynamic participation in policymaking, to have greater inclusion and equity, and improved government performance. Nevertheless, the effect of public participation on government performance is questionable. There have been conflicting theoretical arguments, and existing studies examining different forms of public participation yielded mixed results. This dissertation extends literature by examining three different forms of public participation. First, it examines the effect of participatory budgeting on service performance, using different timing of participatory budgeting adoptions in New York City Council Districts. Applying the difference-in-differences estimator that is robust to heterogeneous treatment effects across groups and over time, I find that participatory budgeting improves public service performance, indicated by the decline in the total 311 complaint rates. However, I find no evidence that districts with PB funds related to school maintenance, potholes, and new trees improve the performance of these specific services, possibly due to the low shares of the capital budget for these services. The second essay investigates the effect of democratization on economic performance, using the case of democratization in Indonesia. Applying the synthetic control method to construct the synthetic unit of Indonesia as the counterfactual, evidence suggests that democratization has a negative effect on economic performance, indicated by GDP per capita. Rather than a linear relationship, however, the trajectory of economic performance shows a decline in GDP per capita in the first few years after democratization, but then it increases over time as democracy is developed toward consolidation. Lastly, I explore equity in ICT-based coproduction using the 311 non-emergency system in San Fransisco City. Applying logistic regressions and the Heckman model to correct sample selection bias, I test the hypotheses that disadvantaged individuals are less likely to know about opportunities for coproduction, to participate in coproduction and in using ICT in coproduction, and to satisfy with their experience in coproduction. I find disparities in the knowledge about 311 -- as a means for coproduction -- in terms of race and education level, gaps in coproduction participation in terms of education level and age, inequity in using ICT in coproduction in terms of education and income levels, and bias in satisfaction in ICT-based coproduction in terms of age. Since the 311 service requests are increasingly submitted through ICT -- especially mobile apps, it indicates inequity in service benefits.