I am a computational social scientist studying how social contexts and emerging technologies shape individual life chances and broader patterns of inequality. My research is organized around three interlocking strands: how social networks structure access to opportunities and influence, how artificial intelligence reorganizes participation, production, and power, and how quantitative and computational methods can help us understand the emergence and transformation of inequality. Across these strands, I ask who gains meaningful access to opportunities and influence, who benefits and who bears the risks, how these differences scale up into broader patterns of inequality, and whether and how policies and interventions can change them.
Networks and Inequality
Social networks channel information, support, opportunity, and influence, but access to these resources depends on whom people encounter and which encounters develop into relationships. My research examines how unequal networks form and persist, and how they shape mobility and political influence. A central distinction is between exposure and integration: sharing a school, neighborhood, workplace, or platform does not necessarily provide access to valuable ties within those settings. Across educational, professional, and political contexts, I study when sorting, social preferences, and relational processes such as triadic closure reinforce boundaries of class and race—and when cross-group ties allow integration to spread.
Current Work
- Social Closure in U.S. High Schools? Patterns and Determinants of Socioeconomic Segregation in Adolescent Friendship Networks (solo-authored)
- Social Networks and Intergenerational Mobility in the Netherlands (with Javier Garcia-Bernardo and Samuel Plach)
- When Friends-of-Friends Divide Us: Triadic Closure and the Making of Segregated Networks (with Mohamad Altrabulsi)
- Not All Bridges Are Equal: Weak Ties, Social Boundaries, and Early-Career Outcomes (with Mario Molina and Manu Munoz)
- Bridging the Divide: How Cross-Socioeconomic Friendships Shape Educational Aspirations (with Malavika Krishnamachari)
- The Precious Networks of the Rich: How the Ultra-Wealthy Prevent Progressive Tax Reforms (with Alison Schultz, Yufang Sun, and Marti Medina Hernandez)
Artificial Intelligence and Inequality
Artificial intelligence is changing not only what people produce but also how they collaborate, how work is organized, and how economic value is distributed. My research asks when AI expands meaningful participation and when it creates new forms of cumulative advantage, dependence, and exclusion. Using online experiments, labor-market data, administrative records, and policy simulations, I trace how changes in collaboration and task composition scale up to affect labor markets, public revenue, and inequality. Together, these projects examine how technological design and public policy can support collective action, preserve fiscal capacity, and distribute the gains from AI more broadly.
Current Work
- AI and Human Collaboration in Social Dilemmas (with Hanan Salam, Hua Shen, Vedant das Swain, and Ashmit Mukherjee)
- An Occupation Is More Than the Sum of Its Tasks. AI Exposure, Task Composition, and Labor-Market Outcomes (solo-authored)
- Rebuilding Tax Systems for an AI Economy: Evidence, Reform Design, and Public Tools for Low- and Middle-Income Countries (with Alison Schultz and Jordan Norris)
Methods for Inequality Research
Inequality is more than a difference in average outcomes. My methodological research develops tools for studying how experiences and interventions reshape entire distributions, how causal effects unfold in connected populations, and how characteristics and relationships at one level combine to produce outcomes at another. Across this work, I treat heterogeneity, dependence, and aggregation not as statistical complications but as substantive features of the processes that produce inequality. By modeling these features directly, I connect individual-level experiences and interactions to relational and population-level patterns of inequality.
Current Work
- Graph Neural Networks for Causal Network Analysis (with Rae Yu, Yuchen Wu, and Matias Cattaneo)
- Treatment Effects on Within-Group and Between-Group Inequality: A Causal Decomposition Approach (solo-authored)
- From Micro Relations to Macro Patterns: Composition, Moderation, and Emergence in Regression Analysis (solo-authored)
Published Work
- Rosche, B. (2026). A Multilevel Model for Coalition Governments: Uncovering Party-Level Dependencies Within and Between Governments. Political Analysis.
- Rosche, B., Bons, H., Hox, J., de Leeuw, E. (2025). The Survey Attitude Scale as an Indicator of Survey Attitude and a Predictor of Nonresponse and Panel Dropout. Survey Research Methods, 19(3), 279-293.
- An, W., Beauvile, R., Rosche, B. (2022). Causal Network Analysis. Annual Review of Sociology, 48.