Research

Ranked Choice Voting (RCV) adoption is expanding across U.S. elections, but faces persistent criticism that it is too complex for public understanding and vulnerable to strategic manipulation and ballot exhaustion. We develop and deploy computational frameworks to empirically test these concerns using real election data from 110 contests across three diverse contexts: New York City's 2021 Democratic primaries (54 races), Alaska's 2024 primary-infused statewide elections (52 races), and Portland's 2024 multi-winner City Council elections (4 races).

Our findings reveal that despite its intricate multi-round functions and theoretical vulnerabilities, RCV consistently exhibits simple and predictable dynamic in practice, closely mirroring the interpretability of plurality elections. RCV measurably increases competitiveness over prior plurality elections, its strategic behavior is largely immune to strategic manipulation while being transparent and predictable, and ballot exhaustion produces quantifiably minimal impact, suggesting RCV's practical resistance to spoiler potential. These findings provide the first large-scale empirical validation that RCV delivers its promised democratic benefits without succumbing to theoretical vulnerabilities. The computational framework offers election administrators and researchers tools for immediate election-night analysis and facilitates clearer discourse around election dynamics.

As artificial intelligence increasingly automates decision-making in competitive markets, understanding the resulting dynamics and ensuring a fair market mechanism is essential.  We investigate the multi-faceted decision-making of large language models (LLMs) in oligopolistic Cournot-based markets, showing that LLMs not only grasp complex market dynamics—demonstrating their potential as effective economic planning agents—but also engage in sustained tacit collusion, driving prices up to 200% above baseline levels. Our analysis examines LLM behavior across three dimensions—(1) decision type (2) opponent strategies, and (3) market composition—revealing how these factors shape the competitiveness of LLM-based decision-makers. Furthermore, we show that regulating a few dominant agents by enforcing best-response strategies may aid in disrupting collusion and restoring competitive pricing. Our findings identify potential concerns associated with AI integration in competitive market environments and provide regulatory policy recommendations for the era of automation.

Ranked Choice Voting and Single Transferable Voting are widely used. An open question is how candidates should strategize, forming coalitions or lending support to other candidates. Calculating optimal strategies is exponentially complex because of intricate per-round vote transfers, with minor changes in the voter rankings potentially leading to bigger ripple effects – for example, lending support to a losing candidate can prevent their votes from transferring to a more competitive opponent. We study optimal strategies – persuading voters to change their ballots or adding new voters – both theoretically and algorithmically. Theoretically, we study the types of strategies that are effective under both perfect and imperfect polling information. Algorithmically, we develop efficient algorithms to reduce the election instance while maintaining optimization optimality, practically circumventing the computational complexity barrier. Finally, we apply our algorithmic approach to ranked-choice polling data on the US 2024 Republican Primary, finding, for example, that several candidates would have been optimally served by boosting another candidate instead of themselves.

Gerrymandering, the deliberate manipulation of electoral district boundaries for political advantage, is a persistent issue in U.S. redistricting cycles. In this work, we introduce and analyze Votemandering, a strategic blend of gerrymandering and targeted political campaigning devised to gain more seats by circumventing fairness measures. Votemandering leverages accurate demographic and socio-political data, bolstered by advancements in technology and data analytics, to influence voter decisions in pursuit of subtle gerrymandering strategies. We formulate votemandering as a Mixed Integer Program (MIP) that performs fairness-constrained gerrymandering over multiple election rounds. We analyze the influence of various redistricting constraints and parameters on votemandering efficacy. We explore the interconnectedness of gerrymandering, substantial campaign budgets, and strategic campaigning, illustrating their collective potential to generate biased electoral maps. A case study of Wisconsin State Senate redistricting reveals significant votemandering potential. Our findings underscore the need for reforms in the redistricting process beyond enforcing thresholds for specific fairness metrics.

The broad objective of this work is to propose a mathematical model for the study of causes of wage inequality and relate it to choices of consumption, the technologies of production, and the composition of labor in an economy. The paper constructs a Simple Closed Model, or an SCM, for short, for closed economies, in which the consumption and the production parts are clearly separated and yet coupled. The equilibria in SCM are market variables that simultaneously satisfy the revenue maximization and optimal Fisher market conditions and their existence is established by developing a correspondence with a related Arrow-Debreu market. This formulation allows us to identify key combinatorial data that link variables of the closed economy with its equilibria, in particular, the impact of consumer preferences on wages. The combinatorial structures associated allow the formulation and explicit construction of the Consumer Choice Game, where expressed utilities of various labor classes serve as strategies with total or relative wages as the pay-offs. We illustrate, through examples, the mathematical details of the consumer choice game. We show that consumer preferences, expressed through modified utility functions, do indeed percolate through the economy, and influence not only prices but also production and wages. Thus, consumer choice may serve as an effective tool for wage redistribution.

Drawing from the experiences of a healthcare service provider during the Covid-19 pandemic, our research aims to analyze the decision-making processes of a provider managing both an Emergency Department (ED) and medical clinics. Patients, exhibiting diverse levels of severity, engage with the provider through direct ED visits or phone consultations.  Their conditions are susceptible to severity evolution, prompting a reconsideration of their decisions. The decision to enter a facility is determined by patients' risk perception weighed against potential benefits. The overarching goal of the hospital system is to effectively allocate service capacity across its facilities to minimize costs arising from patient fatalities or defections. We develop a comprehensive model for this system using fluid approximation techniques, accommodating multiple periods that may exhibit different demand patterns. Despite the inherent complexity of the problem space, we identify distinct sub-regions within it that allow for individual analysis. Through provably efficient computational methods, we establish that achieving the globally optimal solution is feasible for single periods and more importantly, across multiple periods with varying demand characteristics. Our analytical and computational findings highlight a key insight: the influence of endogeneity introduces intricate and counterintuitive capacity allocation dynamics. This observation pertains to both single and multiple time periods. We illustrate our findings through case studies conducted in collaboration with our hospital partner in Champaign county, grounding our work in real-world data.