Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Dissertation Wei Cheng 7-5-2017.pdf (756.83 KB)

ETD Abstract Container

Abstract Header

Essays on the Economics of Social Networks

Cheng, Wei
http://rave.ohiolink.edu/etdc/view?acc_num=osu14992071424818

Abstract Details

2017, Doctor of Philosophy, Ohio State University, Economics.
A variety of connections exist between people, such as friends, neighbors and colleagues. These connections constitute a web of networks. There is a long standing literature in economics that studies the effect of network interactions on an individual’s behavior. A major challenge in investigating this network effect lies in the endogeneity of social networks, which is the focus of this dissertation. In the first essay, I develop statistical tests to determine whether a network is endogenous in the spatial autoregressive framework (SAR). SAR models have been widely used in studying spatial interdependence and social interaction. Previous research have assumed the spatial weight matrix (also known as sociomatrix or adjacency matrix) to be exogenously given. However, ignoring the possibility that adjacency matrices can be endogenous may lead to inconsistent estimates. Therefore, it is desirable to test whether spatial weights are endogenous or not. This paper constructs a Hausman specification test and one simple equivalent test under a general setting. I also provide LM test statistics for two representative SAR models with endogenous spatial/network relationships in the literature. I summarize their testing features and establish their asymptotic properties under the null and the alternative hypotheses. Monte Carlo simulations are conducted to examine the finite-sample performance of those tests. In the second essay, I take a novel approach to estimate productivity spillover among scientific workers. Specifically, I estimate the effect of the number of collaborators and average collaborator productivity on a researcher’s own productivity using a spatial autoregressive (SAR) framework. To obtain a causal relationship, I address the endogeneity issue from people’s selection in collaborators by estimating pairwise network formations in the first step. Identification comes from common research interests that affect a researcher’s choice of collaborators but not his productivity directly. First stage results indicate that scientists prefer to collaborate with other researchers who have similar research interests, but too close proximity in research areas actually decreases the probability of collaborating. Estimation of the productivity equation confirms the existence of a positive spillover effect from previous coauthors through the quantity dimension but not the quality dimension. On average, having 1 more collaborator increases an individual’s impact factor weighted prorated number of publications by 5%. Moreover, the spillover effects are not uniform across different individual characteristics in terms of career age, gender and level of productivity. The third essay studies the productivity spillovers from colleagues and coauthors simultaneously for scientific workers. I use department fixed effects and individual fixed effects to control for departmental common shocks and individual selection. In addition, I model scientists’ mobility choices and use citation breakout as the exclusion variable to address the potential endogenous year-to-year change in colleague networks. Having solved the endogeneity issue of colleague networks, I construct instruments for coauthor quality by exploiting the characteristics of the (estimated staying) colleagues of a scientist’s noncolleague coauthors. Our results provide strong evidence coauthors generate significant spillover effects while colleagues do not. Increasing the number of a scientist’s coauthors by one on average increases his productivity by 2%. Our findings highlight the important of being exposed to a wider spectrum of ideas and knowledge in the production of science. Additionally, spillovers from colleagues remain insignificant across departments of different sizes or quality, and the estimated magnitude of coauthor spillovers is robust to different specifications and subsamples.
Lung-fei Lee (Committee Chair)
Bruce Weinberg (Committee Co-Chair)
Stephen Cosslett (Committee Member)
131 p.
Economics
Economics of Social Networks

Recommended Citations

Citations

  • Cheng, W. (2017). Essays on the Economics of Social Networks [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu14992071424818

    APA Style (7th edition)

  • Cheng, Wei. Essays on the Economics of Social Networks. 2017. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu14992071424818.

    MLA Style (8th edition)

  • Cheng, Wei. "Essays on the Economics of Social Networks." Doctoral dissertation, Ohio State University, 2017. http://rave.ohiolink.edu/etdc/view?acc_num=osu14992071424818

    Chicago Manual of Style (17th edition)

osu14992071424818
342
© 2017, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.