Estimating the Causal Effects of Social Interaction with Endogenous Networks

2012 ◽  
Vol 20 (3) ◽  
pp. 316-328 ◽  
Author(s):  
Jon C. Rogowski ◽  
Betsy Sinclair
Keyword(s):  
Social Networks ◽  
Instrumental Variable ◽  
Empirical Studies ◽  
Causal Effects ◽  
Legislative Behavior ◽  
Roll Call ◽  
Omitted Variables ◽  
Selection Processes ◽  
Endogenous Networks ◽  
Causal Impact

Identifying causal effects attributable to network membership is a key challenge in empirical studies of social networks. In this article, we examine the consequences of endogeneity for inferences about the effects of networks on network members' behavior. Using the House office lottery (in which newly elected members select their office spaces in a randomly chosen order) as an instrumental variable to estimate the causal impact of legislative networks on roll call behavior and cosponsorship decisions in the 105th–112th Houses, we find no evidence that office proximity affects patterns of legislative behavior. These results contrast with decades of congressional scholarship and recent empirical studies. Our analysis demonstrates the importance of accounting for selection processes and omitted variables in estimating the causal impact of networks.

scholarly journals IMPACT OF OFFLINE AND ONLINE SOCIAL NETWORKS ON LONELINESS AND MENTAL HEALTH AMONG ITALIAN OLDER PEOPLE

2019 ◽  
Vol 3 (Supplement_1) ◽  
pp. S529-S529
Author(s):  
Daniele Zaccaria ◽  
Georgia Casanova ◽  
Antonio Guaita
Keyword(s):  
Mental Health ◽  
Social Networks ◽  
Social Network ◽  
Older People ◽  
Online Social Networks ◽  
Social Network Sites ◽  
Well Being ◽  
Controlled Study ◽  
Face To Face ◽  
Causal Impact

Abstract In the last decades the study of older people and social networks has been at the core of gerontology research. The literature underlines the positive health effects of traditional and online social connections and also the social networks’s positive impact on cognitive performance, mental health and quality of life. Aging in a Networked Society is a randomized controlled study aimed at investigating causal impact of traditional face-to-face social networks and online social networks (e.g. Social Network Sites) on older people’ health, cognitive functions and well-being. A social experiment, based on a pre-existing longitudinal study (InveCe - Brain Aging in Abbiategrasso) has involved 180 older people born from 1935 to 1939 living in Abbiategrasso, a municipality near Milan. We analyse effects on health and well-being of smartphones and Facebook use (compared to engagement in a more traditional face-to-face activity), exploiting the research potential of past waves of InveCe study, which collected information concerning physical, cognitive and mental health using international validate scale, blood samples, genetic markers and information on social networks and socio-demographic characteristics of all participants. Results of statistical analysis show that poor social relations and high level of perceived loneliness (measured by Lubben Scale and UCLA Loneliness scale) affect negatively physical and mental outcomes. We also found that gender and marital status mediate the relationship between loneliness and mental wellbeing, while education has not significant effect. Moreover, trial results underline the causal impact of ICT use (smartphones, internet, social network sites) on self-perceived loneliness and cognitive and physical health.

scholarly journals Graphical Models for Quasi-experimental Designs

2015 ◽  
Vol 46 (2) ◽  
pp. 155-188 ◽  
Author(s):  
Peter M. Steiner ◽  
Yongnam Kim ◽  
Courtney E. Hall ◽  
Dan Su
Keyword(s):  
Graphical Models ◽  
Instrumental Variable ◽  
Regression Discontinuity ◽  
Experimental Designs ◽  
Causal Effects ◽  
Confounding Bias ◽  
Randomized Controlled ◽  
Quasi Experimental ◽  
Causal Graphs ◽  
Latent Subgroups

Randomized controlled trials (RCTs) and quasi-experimental designs like regression discontinuity (RD) designs, instrumental variable (IV) designs, and matching and propensity score (PS) designs are frequently used for inferring causal effects. It is well known that the features of these designs facilitate the identification of a causal estimand and, thus, warrant a causal interpretation of the estimated effect. In this article, we discuss and compare the identifying assumptions of quasi-experiments using causal graphs. The increasing complexity of the causal graphs as one switches from an RCT to RD, IV, or PS designs reveals that the assumptions become stronger as the researcher’s control over treatment selection diminishes. We introduce limiting graphs for the RD design and conditional graphs for the latent subgroups of compliers, always takers, and never takers of the IV design, and argue that the PS is a collider that offsets confounding bias via collider bias.

scholarly journals Social networks and unemployment exit in Great Britain

2018 ◽  
Vol 45 (8) ◽  
pp. 1205-1226
Author(s):  
Panos Sousounis ◽  
Gauthier Lanot
Keyword(s):  
Social Networks ◽  
Social Interactions ◽  
Empirical Studies ◽  
Unemployment Duration ◽  
Empirical Literature ◽  
Job Seekers ◽  
Content Type ◽  
Job Attainment ◽  
Discrete Time Hazard Model ◽  
Direct Measures

Purpose The purpose of this paper is to examine the effect employed friends have on the probability of exiting unemployment of an unemployed worker according to his/her educational (skill) level. Design/methodology/approach In common with studies on unemployment duration, this paper uses a discrete-time hazard model. Findings The paper finds that the conditional probability of finding work is between 24 and 34 per cent higher per period for each additional employed friend for job seekers with intermediate skills. Social implications These results are of interest since they suggest that the reach of national employment agencies could extend beyond individuals in direct contact with first-line employment support bureaus. Originality/value Because of the lack of appropriate longitudinal information, the majority of empirical studies in the area assess the influence of social networks on employment status using proxy measures of social interactions. The current study contributes to the very limited empirical literature of the influence of social networks on job attainment using direct measures of social structures.

Bridging Diversity across Time and Space

2011 ◽  
pp. 136-158
Author(s):  
Violina Ratcheva
Keyword(s):  
Social Networks ◽  
Social Context ◽  
Empirical Studies ◽  
Background Knowledge ◽  
Multidisciplinary Teams ◽  
Team Members ◽  
Social Communities ◽  
Knowledge Diversity ◽  
Geographic Separation ◽  
Multidisciplinary Teamwork

The uniqueness of multidisciplinary teamwork is in its potential to integrate different bodies of knowledge into a new synergy. However, previous empirical studies have shown that member heterogeneity and geographic separation hinder effective sharing and use of team knowledge. The chapter explores how such teams interact to overcome the barriers and take advantage of their “built in” knowledge diversity. The findings indicate that often teams lack common background knowledge at the beginning of the projects, and in order to resolve differences members rely on their external intellectual and social communities. The reported research establishes a positive correlation between team members’ participation in multiple professional and social networks and teams’ abilities to successfully build on their knowledge diversity. The findings also suggest a need to reconceptualize the boundaries of multidisciplinary teams and to consider the processes of sharing diverse knowledge in a wider social context.

Das Handwerk – Zur definitorischen und statistischen Abgrenzung eines Querschnittsbereichs

2019 ◽  
Vol 67 (1) ◽  
pp. 67-79 ◽  
Author(s):  
Katarzyna Haverkamp
Keyword(s):  
Empirical Research ◽  
Market Entry ◽  
Empirical Studies ◽  
Causal Effects ◽  
Economic Research ◽  
Entry And Exit ◽  
Definition Of

Zusammenfassung In der empirischen Wirtschaftsforschung zeigt sich ein zunehmendes Interesse an der Untersuchung der Fragen der Gründungsdynamik und des Gründungserfolgs im Kontext der deutschen Handwerkswirtschaft. Eine besondere Herausforderung für diese Analysen besteht jedoch darin, dass eine statistische Abgrenzung des juristisch definierten Handwerkssektors in den vorliegenden Sekundärdatensätzen meist nur mit Einschränkungen möglich ist. Vor diesem Hintergrund analysiert dieser Beitrag Möglichkeiten und Grenzen einer statistischen Abgrenzung des Handwerks in Mikrodatensätzen und untersucht unterschiedliche, bislang verwendete Identifikationsverfahren im Hinblick auf die Repräsentativität der jeweils gewonnen Stichproben. Im Ergebnis zeigt der Beitrag die Stärken und Schwächen unterschiedlicher Identifikationsverfahren und formuliert Empfehlungen hinsichtlich ihrer Verwendung in der Entrepreneurship-Forschung. Abstract Recently, several empirical studies investigate the causal effects of regulation on market entry and exit using the example of the German crafts sector. However, since the definition of the sector is made on legal- and not statistical basis, the identification of crafts companies and employees in microdata records is an intricate process. This paper examines different identification strategies that have been used so far in empirical research and investigates whether the resulting samples are consistent with the overall population in question. The paper contributes to existing economic research by providing an understanding for the potential pitfalls when analyzing sub-groups in larger datasets and by formulating an explicit recommendation for the case of the research on regulation and entry in the German crafts sector.

scholarly journals Roll-Call Vote Selection: Implications for the Study of Legislative Politics

2020 ◽  
Vol 114 (3) ◽  
pp. 691-706
Author(s):  
CAITLIN AINSLEY ◽  
CLIFFORD J. CARRUBBA ◽  
BRIAN F. CRISP ◽  
BETUL DEMIRKAYA ◽  
MATTHEW J. GABEL ◽  
...  
Keyword(s):  
Formal Model ◽  
Legislative Behavior ◽  
Legislative Politics ◽  
Roll Call ◽  
Roll Call Vote ◽  
Voting Procedures ◽  
Strategic Incentives ◽  
Theoretical Results ◽  
Representative Samples ◽  
Quantities Of Interest

Roll-call votes provide scholars with the opportunity to measure many quantities of interest. However, the usefulness of the roll-call sample depends on the population it is intended to represent. After laying out why understanding the sample properties of the roll-call record is important, we catalogue voting procedures for 145 legislative chambers, finding that roll calls are typically discretionary. We then consider two arguments for discounting the potential problem: (a) roll calls are ubiquitous, especially where the threshold for invoking them is low or (b) the strategic incentives behind requests are sufficiently benign so as to generate representative samples. We address the first defense with novel empirical evidence regarding roll-call prevalence and the second with an original formal model of the position-taking argument for roll-call vote requests. Both our empirical and theoretical results confirm that inattention to vote method selection should broadly be considered an issue for the study of legislative behavior.

The Influence of Social Networks on the U.S. Senate Roll-Call Voting

2010 ◽  
Vol 1 (4) ◽  
pp. 22-41 ◽  
Author(s):  
Jang Hyun Kim ◽  
George A. Barnett ◽  
K. Hazel Kwon
Keyword(s):  
Social Networks ◽  
Communication Networks ◽  
United States Senate ◽  
Roll Call ◽  
Party Membership ◽  
Roll Call Voting ◽  
Complex Models ◽  
Network Properties ◽  
Committee Membership ◽  
Relational Resources

Along with individuals’ ideological factors, various network properties play a crucial role in the process of legislators’ political decision making. Social networks among legislators provide relational resources through which communication occurs, exerting social influence among the members in a network. This paper examines six social relationships among the members of the 109th United States Senate as predictors of senatorial voting (roll call votes), shared committees, co-sponsorships, party membership, PAC donation, geographical contiguity, and internet hyperlinks, which may be considered as direct or indirect representations of communication networks. The six networks are modeled using MRQAP, and results suggest that roll call voting was predicted by party membership, co-sponsorship, geographical proximity, and PAC donation networks, while shared committee membership did not contribute significantly. As for hyperlinks, results were mixed, showing a small variance of contribution in a simpler model but not significant with more complex models.

Sign in / Sign up

Export Citation Format

Share Document