Github: https://github.com/ljanastas/.

CV: Download here.

Google Scholar: https://scholar.google.com/citations?user=I9yQIZ0AAAAJ&hl=en.

Research Statement

My research explores how technological change reshapes political institutions.

Each major technological revolution has resulted in an expansion of bureaucratic and administrative capacity. And once that capacity grows, it rarely shrinks.

I call this phenomenon the "technology-state ratchet," and part of my research examines how this mechanism works and how it has contributed to the growth and structure of the administrative state in both democracies and authoritarian regimes. Some questions that my current research addresses are: as an innovation spreads, who benefits politically? How does the relationship between citizens and the state evolve? And what new equilibria do political institutions settle into?

Artificial intelligence may be among the most politically important innovations of our time, and large chunk of my recent work focuses on it. Some questions that I’m pursuing now include: will AI strengthen authoritarian regimes or weaken them? Does the incorporation of AI into government processes tend to centralize or decentralize political power? Does the use of AI for political ends empower citizens and strengthen democracy, or amplify the influence that corporate interests and lobbyists already hold?

A growing stream of my research applies these same tools to science itself. Scientific institutions face their own principal-agent problems: funders, journals, and researchers each respond to incentives that can make unreliable work individually rational. My current projects include a mechanism design account of the replication crisis and a computational research integrity pipeline that combines LLMs with statistical forensics to detect problematic empirical findings at scale, benchmarked against Retraction Watch.

In addition to these substantive interests, my work in political methodology focuses on causal and computational methods to measure institutional change. My work has appeared in the American Political Science Review, Political Analysis, the Journal of Public Administration Research and Theory, the Journal of Democracy, and Public Administration Review, among others.

Publications and works in progress

BOOKS

  • Causal Inference and Counterfactuals with Machine Learning Under contract, Cambridge University Press (Elements Series)

  • Inventing the Leviathan: How Technology Created the Administrative State In preparation — to be submitted to Cambridge University Press (Historical Political Economy series)

WORKING PAPERS/WORKS IN PROGRESS

  • Rational Irreproducibility: Why Partial Reforms to the Replication Crisis Don’t Work

  • Computational Forensics for Research Integrity: A Forensic Funnel Approach (In Progress)

  • Simulating Public Opinion with Large Language Models (w/ Ryan Powers and Akshat Lakhiwal) (In Progress)

  • Artificial Intelligence, Bureaucratic Discretion, and Democratic Administration (Revise and Resubmit at Public Administration Review)

  • Delegation, Discretion, and Regime Type: A Principal-Agent Theory of AI Adoption (w/ Jie Lian) (In Progress)

  • The State-Builder's Dilemma: Rural Electrification, Southern Oligarchy, and the Limits of the New Deal Administrative State (In Progress)

  • Adverse Destruction? Federal Workforce Reductions and Bureaucratic Capacity (w/ Gregory Porumbescu) (In Progress)

PUBLICATIONS