Social Networks and Individual Labor Market Outcomes - Evidence from Linked Employer-Employee Panel Data

Ilyés, Virág (2023) Social Networks and Individual Labor Market Outcomes - Evidence from Linked Employer-Employee Panel Data. PhD thesis, Budapesti Corvinus Egyetem, Szociológia és Kommunikációtudomány Doktori Iskola. DOI

PDF : (dissertation)
PDF : (draft in English)
PDF : (az értekezés tézisei magyar nyelven)


The dissertation focuses on the impact of social networks on the economic and employment opportunities of individuals. It examines the direct and indirect effects of professional ties on various labor market outcomes, such as job finding probabilities, wages, job stability, and upward mobility. Following recent trends in the measurement of network effects, the empirical work takes a relatively novel approach to quantify the economic benefits of social ties by utilizing a large Hungarian linked administrative employer-employee panel dataset. While the dataset does not contain direct information on social ties, it does have anonymous firm and individual identifiers, which can be used to identify different segments of networks, such as former co-workers and university peers. Building on studies using similar data, elaborating on empirical approaches already used to measure network effects, and incorporating recent developments in panel data methods, the dissertation presents three empirical studies on the role of social ties in the labor market. The first study contributed to the literature of co-workers, employee referral, and wage differences in multiple ways. First, by being the first to document the presence of wage gains commonly attributed to the referral activity of former co-workers through the estimation of a two-way fixed effects wage equation on starting wages. We found a 2.1% premium for men, which originates from the direct (“presence”) effect of referrers and match selection (that is co-workers facilitate the creation of better employer–employee pairings). Second, by jointly assessing those selection channels that may also contribute to the generation of the overall wage gains and by examining their relative importance. We demonstrated that there is an additional 1.7% and 0.9% wage advantage, reflecting individual and firm selection, respectively. The presence of such wage elements implies that former co-workers might help individuals getting into high-wage firms, and they also induce the selection of better individuals to firms. Fourth, by slightly extending upon the decomposition method of Woodcock (2008), and this way, by showing an even more nuanced picture on the actual drivers of given selection channels. We revealed that the individual and firm selection terms are mostly driven by their respective within components. That is, individuals with contacts tend to be hired by companies with superior worker pools compared to other firms that rely solely on formal hiring, and that the unobserved quality of linked workers is usually better than the quality of their new firms’ other hires. Finally, we contribute to the literature by presenting various endeavors to link differences in the estimated wage components to theories in the referral and co-worker literature. The second study investigates the differential effects of former co-workers on job finding probabilities and upward mobility by gender. The results of the hiring analysis showed that (1) men benefit more from the help of former co-workers; (2) gender homophily in network effects is only present due to gendered patterns of labor market segregation; and (3) the hiring benefits of women are mostly driven by the help of contacts higher up in the occupational ladder. By focusing on job entries to new firms, we have also shown that informal contacts can also influence career development through job mobility. However, the benefits are unevenly distributed both across and within genders. The returns to social ties are greater for men, who tend to realize meaningful benefits regardless of their prior job and firm quality. On the contrary, among women, only those with average or worse labor market positions receive such gains. The results reflect a duality in network effects: while social ties enhance the limited opportunities of women in worse positions, they also contribute to preserving existing gender differences at the top segments of the labor market. The final study examines the effect of former university peers on job finding and the quality of the first job. The results suggests that acquaintances from university influence the labor market outcomes of graduates at the start of their careers: they increase the chances of their peers of getting into given firms and contribute to getting more prestigious and better paying jobs. However, the measured benefits are primarily attributable to ties from bachelor’s studies, while the impact of master’s peers is mainly driven by the selection of individuals along existing pathways between university master’s programs and given firms. These findings may suggest that too much similarity in the educational and career paths of former university peers, especially early in their careers, may limit the chances of individuals providing help to each other and may even be accompanied by crowding out effects. Also, that dissimilarity, to a given extent, could be associated with increased information on available jobs and better economic opportunities. The results of the studies, taken together, advance our understanding on the relationship between networks and individual economic opportunities and provide essential insights for the disciplines of sociology, economics, and social policy as well. By facilitating discussion on the dual nature of networks as sources of economic benefits and amplifiers of inequalities, hopefully the thesis will inspire further academic work on the topic.

Item Type:Thesis (PhD thesis)
Supervisor:Bartus Tamás, Lőrincz László
Subjects:Human resource management
ID Code:1272
Date:30 March 2023
Deposited On:26 Jan 2023 08:17
Last Modified:01 Jun 2023 08:35

Repository Staff Only: item control page


Downloads per month over past two year

View more statistics