Infrastructure development, land, human resources development, regional integration, local government and development, public private partnership

Big data technologies are profoundly changing how people live and how society funtions. Many areas including digital economic, smart city, sustainability and public health have been actively adopting smart technologies while facing challenges created by smart technologies such as data capacity, data sharing and citizen pravicy. There is a great need for public administrators and policy makers to understand the changes brought by cutting edge technologies and adapt into a smarter and more aglie governance approach. Meanwhile, the new information era has brought us the oppurunity to utilize abundant data to further understand policy implications in a finer spatial and temporal scale.

The Big Data for Public Policy cluster seeks to analyze the impact of technology and big data on public policy and public administration, connect traditional governance theories with cutting-edge machine learning methodologies, design and evaluate people-centered smart city solutions.

Relevant Projects

Does Goal Setting Improve Responsiveness of Technology-Enabled Coproduction?

This study utilizes case-level 311 service requests data from 13 U.S. cities in 15 years, this study investigates the relationship between goal-setting and response time of service requests. Specifically, I ask two questions: (a) does having a case-level service target decrease response time of service requests? (b) does shorter target time decrease response time of service requests? I select 5 common service types across cities. In total, more than 4 million service requests are aggregated into 343 city-service observations. Utilizing case-level data, three measurements mean, median, 90th percentile are calculated and compared. The results confirmed both of the hypotheses, that having a target decreases service response time, and among cities with service targets, a shorter target is associated with shorter response time. This study also explores issues in matching service types and choosing performance measurements.

The Interconnections of Decisionmaking Tools in the Era of Smart Governance: A Fine-Scale Analysis.

This study investigates whether using technology-enabled coproduction affects orga- nizational service delivery routine. I select the case of disaster recovery for the 2018 Hurricane Michael in the City of Tallahassee (Florida, U.S.). Interviews with city officials were conducted to guide the theoretical framework and provide insights on the results. I conduct Poisson regres- sion to investigate whether citizen requests affect organizational service delivery time. The results show that although organizational routine and citizen requests do help reduce restoration time independently, there is no evidence indicating that citizen requests directly affect organizational routine.

Toward Individual Equity in the Era of Big Data: A Machine Learning Approach.

This sutdy proposes an evidence-based and needs-based service delivery approach by using big data at the household level and machine learning techniques. This study uses 311 public service requests data to estimate the household-level needs for public services. I identify citizens’ requesting behavior as a function of six factors: needs, awareness, accessibility, service condition, bystander effect, and trust in government. Firstly, 5 types of machine learning model were trained using requesting behavior as the dependent variable and select the model with best performance. Secondly, I estimate household-level score of needs by setting other factors to an ideal scenario. I use survey data from 907 households and archival household-level data for 53,685 households in the City of Tallahassee (Florida) during 2018 Hurricane Michael to illustrate this method. The estimated score of needs was designed to reveal needs for services of those who do not utilize the citizen requests platform. The estimated score also shows a positive relationship with various socioeconomic indicators, which is consistent with previous studies on citizen request behavior. Our approach can help government decision makers better allocate resources based on fine-scale analysis of needs, which echoes New Public Administration’s call for equitable, proactive, client-central, and responsive decisionmaking.