Hi, I’m Ruoyu Chen, a Ph.D. student at the Sol Price School of Public Policy, University of Southern California. I hold an M.S. from Peking University and a bachelor’s degree from Huazhong University of Science and Technology, with a semester at Nanyang Technological University in Singapore.


My research interests include transportation, urban economics, and urban analytics. I have extensive experience in leveraging observational (big) data to generate actionable insights using a wide range of advanced analytical methods, including modern econometrics, spatial analysis, and machine learning.


My previous work has been published in flagship journals like Transport Research Part A (TRA) and Journal of Transport Geography (JTG).  I’ve also been recognized with awards such as the Best Poster Award at the World Transport Convention and the Best Student Paper Award from the International Association of China Planning.


Scroll down or visit my Google Scholar to learn more about my research. Interested in collaborating? Feel free to reach out! :)


Email: rchen912 at usc dot edu

Google Scholar | LinkedIn | ORCID

Selected Publications

Keywords: Jobs-housing relationship, Change, COVID-19, Big data, China

ABSTRACT

The outbreak of COVID-19 and subsequent pandemic containment measures have significantly affected our daily life. Scholars conducted empirical studies and uncovered various immediate impacts of these on transportation systems, e.g., decreasing transit ridership, increasing private car usage, and declining long-distance and international trips. However, the long-lasting jobs/housing impacts of these are still unclear. We attempted to fill this gap by looking into the jobs-housing relationship before and amid the COVID-19 pandemic using an excess-commuting approach. The approach allowed us to analyse a series of jobs-housing matrices based on the location-based service big data of around fifty million individuals in the Pearl River Delta (PRD), China before and amid COVID-19. In the PRD, a zero-COVID policy was implemented, which presents a distinct and interesting context for our study. We found that after the COVID-19’s outbreak: (1) residences and employment became more centrally located in downtowns, which is opposite to the trend of suburbanization elsewhere; (2) in the whole PRD, the minimum and maximum commutes became smaller while the actual commute became larger, indicating the co-presences of better spatial juxtaposition of jobs and housing, more compressed distribution of jobs and housing, and longer average actual commutes; (3) more exurban residents started working in exurbs and suburbs whereas more suburban residents chose workplaces in downtowns; (4) the female and senior enjoyed less increase in the average commute. This paper illustrates the potential of big data in the longitudinal study of jobs-housing relationships and excess commuting. It also produces new insights into such relationships in a unique context where zero-COVID policies are in presence. 

Keywords: Transit Fare, Transport Policy Evaluation, Empirical Data, Rail Transit, Unsupervised Learning, Wuhan

ABSTRACT

Fare policy plays an important role in transit operations and management. To better coordinate and achieve the multidimensional goals of a proposed fare adjustment policy (e.g., increasing revenue, managing demand, and improving equity), a fundamental step is to evaluate its travel pattern impacts, which helps us consider the policy in a bigger socioeconomic context. Existing studies rarely investigate the impacts of such a policy on different users’ and user groups’ travel patterns and transit operators’ farebox revenue using longitudinal data from sources such as smartcard data. To fill this gap, we exploit 24 weeks’ smartcard data from Wuhan, China, to empirically quantify those impacts. We find that (a) the fare increase had significant but varying impacts on travel patterns across users and user groups; (b) confronting the fare increase, commuter groups identified by the topic model reduced their trip frequency more but later as compared to other groups; (c) low-accessibility, long-distance, and single-destination metro riders were less sensitive to the fare increase; (d) when there was a system-wide fare increase with a distance-based structure, trip purposes and socioeconomic statuses could better predict the impacts on the travel demand and farebox revenue than spatiality. These findings indicate that increasing average fares while offering discounted tickets for frequent and/or captive riders could maintain the existing ridership and farebox revenue and possibly increase additional ridership.