Repository for the Spring 2025 Computational Social Science Workshop
Time: 11:00 AM to 12:20 PM, Thursdays Location: 1155 E. 60th Street, Chicago IL 60637; Room 295
Matthew Elliott is Professor of Economics at Cambridge University and fellow of Jesus College Cambridge. Before that he was Assistant Professor of Economics at the California Institute of Technology and postdoctoral researcher at Microsoft Research, New England Research and Development (NERD). He is associate editor at Econometrica and serves on the editorial board at the Review of Economic Studies. In 2017 he was awarded the British Academy Wiley Prize in Economics, and in 2012 the Aliprantis Prize (joint with Professor Ben Golub) awarded by the Society for the Advancement of Economic Theory. He was educated at Oxford University and then Stanford University, where he received his PhD.
Networked Production and Fragility. Modern production often relies on a complex network of collaborations, be it via supply chains or within large organizations. This complexity raises the possibility of systemic fragility, especially in the short run before the system can rewire and reorganise. Through a series of different but related modelling approaches, this talk explores fragility in the short-run and the medium-run in a variety of contexts, and how it can emerge endogenously through different mechanisms.
Reading List:
Julia Ticona is an assistant professor at the Annenberg School for Communication, where her research investigates the ways that digital technologies shape the meaning of precarious work. She uses qualitative methods to examine the role of tech in the construction of identity and inequality for low-wage workers. Previously, she was a postdoctoral scholar at the Data & Society Research Institute, where she collaborated on an amicus brief on behalf of Data & Society for Carpenter vs. U.S. before the U.S. Supreme Court. She is an Associate Fellow at the Institute for Advanced Studies in Culture. She received her Ph.D. in Sociology from the University of Virginia, where she was a member of the Society of Fellows, and her B.A. from Wellesley College. You can find her work in the Journal of Communication, New Media & Society, the International Journal of Communication, and Information, Communication, and Society. She has also published op-eds and essays in Wired, Dissent, FastCompany, and Slate. She has also been called on as an expert by local and national media outlets, including: The New York Times, NPR, The Nation, and The Philadelphia Inquirer.
Imagining AI in organized labor: Struggles over the value of cultural work. The popularization of generative artificial intelligence (AI) tools has raised concerns about the future of creativity and cultural work. In recent months, AI has become a key sticking point in labor negotiations for workers in media industries, from journalists to actors to film and television writers, many of whom are contending with the possibility of automation for the first time. Through interviews, ethnographic observation, and discourse analysis, this collaborative project investigates how cultural workers – and the unions that represent them – conceptualize generative AI and how these conceptualizations in turn shape 1) labor demands and 2) cultural workers’ understandings of their work and status relative to other occupational groups. The project’s goals are to deepen scholarly and popular understanding of the social processes by which collective meaning is assigned to emerging workplace technologies, and to consider how these assigned meanings have implications for ongoing labor struggles and inter-occupational solidarity. Findings will be disseminated via at least two peer-reviewed scholarly articles, presentations at academic conferences and research-oriented labor convenings, and a written summary of the findings geared toward labor leaders.
Reading List:
- Imagining AI in organized labor: Struggles over the value of cultural work
- Constructing What Counts as Human at Work: Enigma, Emotion, and Error in Connective Labor
Joshua Epstein is Professor of Epidemiology in the NYU School of Global Public Health, and founding Director of the NYU Agent-Based Modeling Laboratory, with affiliated appointments at The Courant Institute of Mathematical Sciences, and the College of Arts & Sciences. Prior to joining NYU, he was Professor of Emergency Medicine at Johns Hopkins, and Director of the Center for Advanced Modeling in the Social, Behavior, and Health Sciences, with Joint appointments in Economics, Applied Mathematics, International Health, and Biostatistics.
His research interest has been modeling complex social dynamics using mathematical and computational methods, notably the method of Agent-Based Modeling in which he is a recognized pioneer.
Generative and Inverse Generative Social Science: Agent_Zero, the Rational Actor, and Beyond. Professor Epstein will discuss (1) the epistemology of generative social science (2) Agent_Zero as a formal cognitively plausible alternative to the Rational Actor, and (3) the use of AI, and Genetic Programming in particular, to evolve families of generative agent architectures computationally.
Reading List:
Michal Kosinski is an Associate Professor of Organizational Behavior at Stanford Graduate School of Business. His research interests encompass both human and artificial cognition. His current work centers on examining the psychological processes in Large Language Models and leveraging Artificial Intelligence, Machine Learning, Big Data, and computational techniques to model and predict human behavior.
Emergent Properties in Large Language Models. Large Language Models (LLMs) trained to predict the next word in a sentence have surprised their creators by displaying emergent properties, ranging from a proclivity for biased behavior to an ability to write computer code and solve mathematical tasks. This talk discusses the results of several studies evaluating LLMs' performance on tasks typically used to study human mental processes. Findings indicate that as LLMs increase in size and linguistic dexterity, they can navigate false-belief scenarios, sidestep semantic illusions, and tackle cognitive reflection tasks. This talk will explore several possible interpretations of these findings, including the intriguing possibility that LLMs do not merely model language but also the psychological processes underlying how humans use language.
Reading List:
- Kosinski, M. (2024). Evaluating Large Language Models in Theory of Mind Tasks. Proceedings of the National Academy of Sciences (PNAS).
- Hagendorff, T., Fabi, S., Kosinski, M. (2023). Human-like intuitive behavior and reasoning biases emerged in large language models but disappeared in ChatGPT. Nature Computational Science.
We will have two speakers on April 16th.
Daniel Larremore is an associate professor in the Department of Computer Science and the BioFrontiers Institute at University of Colorado Boulder. His research develops statistical and inferential methods for analyzing large-scale network data, and uses those methods to solve applied problems in diverse domains, including public health and academic labor markets. In particular, his work focuses on generative models for networks, the ongoing evolution of the malaria parasite and the origins of social inequalities in academic hiring and careers. Prior to joining the CU Boulder faculty, he was an Omidyar Fellow at the Santa Fe Institute (2015-2017) and a post-doctoral fellow at the Harvard T.H. Chan School of Public Health (2012-2015). He obtained his PhD in applied mathematics from CU Boulder in 2012, and holds an undergraduate degree from Washington University in St. Louis.
Patterns and Determinants of Faculty Mid-Career Moves in US Universities. Despite extensive research on faculty hiring and attrition, relatively little is known about mid-career moves (MCMs) where faculty change institutions while remaining in academia. In this study, we analyze 10 years of annual census data on tenure-track faculty at US PhD-granting institutions to identify patterns in the rates, risks, and consequences of MCMs for careers and the scientific ecosystem more broadly. At a high level, MCMs tend to flow from rural to urban areas and from public to private institutions, with faculty often trading prestige for promotion. MCM rates vary considerably by academic field, even more so by career age and academic rank, by institutional characteristics such as private/public status, geographical location, and perhaps surprisingly, by whether one is employed at one's doctoral alma mater.
Reading List 1:
Aaron Clauset is a Professor of Computer Scinece at University of Colorado Boulder. He received a PhD in Computer Science from the University of New Mexico and a BS in Physics from Haverford College, and, prior to coming to Colorado, was an Omidyar Fellow at the prestigious Santa Fe Institute. Clauset is an internationally recognized expert on network science, data science, and complex systems. He received the 2016 Erdos-Renyi Prize in Network Science, and his research has appeared in prestigious scientific venues like Nature, Science, PNAS, JACM, STOC, AAAI, SIAM Review, and Physical Review Letters. His work has also been covered in the popular press by the Wall Street Journal, The Economist, Discover Magazine, New Scientist, Miller-McCune, the Boston Globe and The Guardian.
Evaluation dynamics in peer review at elite scientific journals Peer review at elite scientific journals plays a unique role in supporting public trust in science, influencing the direction of scientific discovery, and shaping scientific careers. However, the need for confidentiality in peer review makes it difficult to assess how well elite journals meet the community's ideals or to develop and test theories to improve it. Here, we use 110,303 submissions over 5 years to Science and to Science Advances, two elite, general science journals, to answer three questions about elite peer review. First, we ask, how much do author and manuscript characteristics like author team size, institutional prestige, manuscript topic, author gender, and geography have on editorial and peer review outcomes? Second, who has more influence on desk rejection and on acceptance decisions, journal editors and outside experts? Third, does a reviewer's gender influence the sentiment of the reviews they write? Our results shed new light on the complexity of manuscript evaluation at elite journals, and underscore the importance of detailed peer review data for untangling disparities in evaluations to identify and resolve potential sources of bias.
Reading List 2:
Sameer Srivastava is the Ewald T. Grether Professor of Business Administration and Public Policy at UC Berkeley’s Haas School of Business and serves as chair of the Management of Organizations group. He is also affiliated with UC Berkeley Sociology. His research uses computational methods to: (1) unpack the complex interrelationships between group culture, individual cognition, and interpersonal networks; and (2) examine how they jointly relate to individual attainment and organizational performance. He holds AB, AM, MBA, and PhD degrees from Harvard University.
Consonance Versus Dissonance: How Exposure to Unfamiliar Colleagues Within and Across Network Communities Affects Social Belonging and Network Change. Organizations vary in the degree to which their members experience social belonging. Interventions designed to boost belonging have typically focused on changing individuals’ mindsets. We instead develop a structural intervention that seeks to foster belonging by exposing people to unfamiliar colleagues—ones they are not in regular contact with. We consider two forms of such exposure: consonant, to colleagues from the same network community as the focal actor; and dissonant, to colleagues from different network communities. We hypothesize that consonant exposure engenders more group solidarity, more persistent relationships, and enhanced social belonging. We test these expectations in a pre-registered field experiment at a non-profit organization. Participants (N=213) engaged in a facilitated professional development experience with unfamiliar colleagues and were randomly assigned to either consonant or dissonant groups. Although the anticipated solidarity advantage of consonant exposure was only marginally significant when assessed immediately following the intervention, consonant-condition participants maintained more intervention-group ties and reported greater social belonging three months after the intervention. Yet, pointing to the potential tradeoffs of consonant versus dissonant exposure, dissonant-condition participants experienced steeper declines in network constraint and greater increases in betweenness and closeness centrality. We discuss implications for research on social networks, workplace belonging, and organizational intervention.
Reading List
Benjamin Golub is an American economist who is a professor of economics and computer science at Northwestern University. His research focuses on the economics of networks. He was named the winner of the 2020 biannual Calvó-Armengol International Prize, which recognizes a “top researcher in [e]conomics or social sciences younger than 40 years old for contributions to the theory and comprehension of the mechanisms of social interaction.”
Incentive design with spillovers. How should contracts be designed to motivate a group of agents to work efficiently toward a risky outcome, such as a scientific advance? This is a fundamental question about optimal contracting for teamwork, and the framework to formalize it is nearly 50 years old. Nevertheless, economics has made little progress on this problem for general team production technologies. We solve the problem by combining insights from contract theory and the theory of network games. The talk will also discuss using recent advances in large-matrix statistics to address realistic cases in which the network of collaborations and spillovers is not known to the designer, and which offer a variety of open avenues for both theoretical and empirical research. The talk will be accessible to an interdisciplinary social science audience.
Reading List
Eleanor Power is an Associate Professor in the Department of Methodology. She completed her PhD in Anthropology at Stanford University in 2015. Prior to joining LSE in 2017, she was an Omidyar Postdoctoral Fellow at the Santa Fe Institute.
Eleanor is an anthropologist interested in how belief, practice, and identity interact with and shape interpersonal relationships. She looks at how people work to discern something of the character, moral being, and intentions of their peers through their actions, and equally how people strive to communicate something of themselves to others, both in dramatic and in subtle ways. She studies how such actions and reactions form the basis not only of people’s perceptions of one another, but also form the substance of their relationships and the emergent structure of their social world.
Generosity and Reputational Concern Across Cultures: Networked Dictator Games in Five Countries. We conduct experimental economic games to study how reputational stake influences people’s decision-making. Players make a series of “Dictator Game” decisions, splitting an endowment between themselves and a recipient. Crucially, recipients are not anonymous strangers but are other community members, presented via photo on a custom Android app. By varying the identity of the recipient and whether they will come to know the identity of the donor, we effectively vary the reputational exposure of the donor's decision. We expect that players will be more generous when their decisions have greater reputational stake. This greater reputational stake could come from: the revelation of the donor's identity, the social proximity of donor and recipient, and their respective network positions. We conduct these games in seven rural communities in five countries (India, Colombia, Nepal, Morocco, and Mexico), where we already have full sociodemographic and social network data. This entails (so far) almost 1400 players and almost 40,000 allocation decisions.
While there is substantial cross-site variation in the average amount given (implying different cultural norms), we find strikingly similar effects of social proximity and revelation across sites. Donors give more of their endowment to friends or friends-of-friends, as opposed to more distant recipients. We further find a small but consistent effect of revelation on Dictator Game allocations: donors give more of their endowment when their identity is revealed, as opposed to being kept anonymous. There is greater heterogeneity in how revelation interacts with social proximity and the network position of donor and recipient, the implications of which we discuss.
Reading List
- Iyer, P., Deschenaux, I., Ross, C., Alami, S., Seabright, E., Hertzog, W., Chauhan, K., & Power, E. A. (n.d.). Reputational concern across cultures. (Unpublished Manuscript, see email link)
- Bursztyn, L., & Jensen, R. (2017). Social image and economic behavior in the field: Identifying, understanding, and shaping social pressure. Annual Review of Economics, 9(1), 131–153. https://doi.org/10.1146/annurev-economics-063016-103625
- Dumas, M., Barker, J. L., & Power, E. A. (2021). When does reputation lie? Dynamic feedbacks between costly signals, social capital and social prominence. Philosophical Transactions of the Royal Society B: Biological Sciences, 376(1838), 20200298. https://doi.org/10.1098/rstb.2020.0298