My research uses statistical, network, and computational methods to study organizational behavior, social movements, social inequality, and demography.
I am particularly interested in two lines of organizational studies. The first seeks to understand the antecendents and consequences of the movement of corporate political transparency and accountability in the U.S., while the second assesses the relationship between corporate political activities and social responsibilities.
Along the first line, I focus upon shareholder activism against corporate political activities (i.e., shareholder political activism) and its influence. I investigate how shareholders select corporate targets, how firms manage shareholder threats, and how shareholder activism influences corporate political and financial outcomes. This project contributes to several scholarly debates such as the business unity debate and corporate political spending as an agency problem debate. Relying on webscraping and big data techniques, it also builds a unique shareholder activism and political spending database for U.S. SP1500 companies in 2000-2018.
Along the second line, I focus upon the configuration of corporate political activity (CPA) and corporate social responsibility (CSR). I study how the configurations of a firm’s CSR and CPA, such as alignment, misalignment, and non-alignment, affect the firm’s reputation and status. Using Chinese Private Enterprise Survey data, I examine the antecedents and consequenes of the CPA and CSR configuration.
In addition to these two lines of work, I have several collaborations related to organizational studies. With Charles Seguin (Penn State University), we examine the policy impact of social movement organizations (SMOs) before congress. Using web-scraping techniques, we build a unique database on the appearance of SMOs before congress in 1870-2016. This work is funded by NSF(Award No.: 1824092). With Andrew Davis (NC State University), I study how terror organizations use civic organization openness to mobilize for attacks (International Journal of Comparative Sociology, 2019). With Alex Kinney and Andrew Davis, I use topic modeling techniques to study how terror organizations recruit potential supporters (Poetics, 2018).
This strand of research focuses upon how different institutions (re)produce social inequalities in the school settings. Specifically, I bring insights from social movement and organizational theories to study how racial and organizational dynamics influence school de(re)-segregation.
With Jeremy Fiel, I use the Age-Period-Cohort method to decompose the change of school segregation over time in the U.S. since the late 1990s (Demography, 2017). This work contributes to the school desegregation debate by providing evidence on the persistent black-white segregation after accounting for students transitioning from elementary, to middle, and to high schools.
With Jeremy Fiel, I also use discrete time event analysis to study the reversal trend of court-ordered school desegregation in the U.S. from 1970 to 2013 (American Journal of Sociology, 2019). The AJS article contributes to the resegregation debate by examining how federal and local environments influence school resegregation. We find that after accounting for federal policy changes and districts’ variable success in desegregating schools, local organizational, financial, political, and racial factors matter. You can click here for data visualization.
Going forward, I plan to use text mining and social network methods to analyze the content of court cases of school desegregation for a better understanding of how micro courtroom dynamics influence school desegregation. I have scraped all federal and state legal cases on school segregation from Nexis Uni database.
This demography line of research uses causal inference methods to study the causes and consequences of cohabitation in China. One project focuses upon the cohabitation effect (Journal of Marriage and Family, 2017). Particularly, I use propensity score matching with survival analysis methods to assess how premarital cohabitation impacts subsequent divorce in post-reform China. I also use an innovative imputation-based sensitivity analysis method to assess whether the omitted variables can alter the significance level of the conclusion. I find that premarital cohabitation has a positive impact on divorce in early-reform period (1980-1994), but this effect disappears in late-reform period (1995-2010). This research provides first evidence supporting the diffusion perspective of the influence of cohabitation on marital instability in the non-Western society.
Another following work elaborates the meaning of cohabitation. I hypothesize that cohabitation was a prelude to marriage in the early-reform period, but now it serves as a test marriage in China. This research will directly speak to the debate on the nature of cohabitation.
In addition to the foregoing projects, I use a variety of computational methods such as topic modeling, machine learning, online experiments, and webscraping techniques to study social phenomenon of interest. With Charles Seguin, one work uses a series of machine learning methods such random forest and convolutional neural networks to assess the stability and dynamics of culture by predicting the gender type of U.S. baby names. Another work uses big data and webscraping techniques to study social scientists testifying before congress.