School of Public and International Affairs, 1929-2026

Armenia and Azerbaijan face enduring challenges in their post-conflict relations. Both must address structural conditions and resolve ongoing irredentist demands against one another. They must also mitigate the historical animosity that perpetuates their conflict. This thesis investigates three conditions for settlement-level agreements employed by third-party guarantors to foster a durable peace: sustained guarantor commitments, symmetric regional integration, and active – not passive negotiation. Existing comparative case studies on Israeli agreements with Egypt and Jordan and the Dayton Accords were analyzed to determine the effects of various post-conflict mechanisms. These mechanisms include economic incentives, institutional design, and security guarantees. Their impact on the likelihood of sustaining a negative peace was assessed. On this basis, novel empirical data from 12 semi-structured elite interviews were collected. The interviewees included current and former U.S. negotiators, ambassadors, and regional stakeholders. Their insights were used to assess the impact of the U.S.-proposed Trump Route for International Peace and Prosperity (TRIPP) economic framework, asymmetrical interdependence, and mediation postures on this unfolding peace sample. The findings suggest that different mechanisms in peacebuilding and regional integration have played contradictory and overlapping roles in the South Caucasus. The analytical framework developed bridges the gap between classical conflict-resolution concepts of negative peace and liberal institutionalism. The findings predict that unsequenced economic integration without prior institutional architecture is more likely to create coercive leverage in favor of autocratic regimes like Azerbaijan. In contrast, a credible third-party guarantor acting as an active negotiator (the U.S.) is required to balance asymmetries and bind parties to enforcement obligations. The last section offers a concrete recommendation for policymakers.

Within the dominant security studies and energy policy literature, climate ambition has long been treated as necessarily subordinate to national security priorities in moments of geopolitical crisis. This thesis turns this canon on its head. Throughout the 2000s-2010s, the EU was heavily reliant on Russian fossil fuel supplies, representing approximately 40% of the EU’s fuel consumption between 2000-2022. During this period, Russia leveraged this trade relation as a site of geopolitical coercion, arresting fuel flows to destabilize the EU. With Russia’s invasion of Ukraine, European policymakers, wary of the prospect of Russian territorial aggression, suspended all fossil fuel trade with Russia, pursuing regional energy integration to to offset the suspension of Russian fuel supplies. Yet, in the years that followed, Russia continued to exploit European energy systems as a node of coercion, targeting European transboundary infrastructure with covert attacks as part of a broader hybrid warfare campaign against the EU. Responsively, EU policymakers increasingly framed energy security via energy sovereignty as a key priority within the states’ national security doctrines. Specifically, and unexpectedly, EU policymakers looked to decarbonization as a means of enhancing domestic energy autonomy, framing renewable energy transition as a core national security imperative. This surprising divergence from the canonical theory that climate ambition must necessarily be sidelined in times of conflict begs the questions: To what extent did the onset of the Russo-Ukrainian conflict and Russia's accompanying hybrid warfare campaign against the EU drive the securitization of decarbonization, reconstituting energy transition from a climate imperative into a national security asset? Moreover, to what extent did this securitization of decarbonization materially 4 accelerate the energy transition ambitions of EU member states, and what might this imply about the influence of exposure to conflict on the pace of renewable energy transition? To answer this question, this thesis employs a mixed methods approach of historical tracing, empirical energy system analysis (employing International Energy Agency energy mix and fuel import data as well as The Center for the Study of Democracy’s Energy Security and Cliamte Risk Index), and textual policy analysis comparing the pre-2022 and post-2022 national security doctrines and National Energy and Climate Plans (NECPs)’s of six country case studies: Finland, Poland, Romania, The Czech Republic, Germany and France. These case studies were selected to capture the geographical scope of the European Union, differentiated exposure to Russian hybrid energy coercion, and diversity of energy system structure and energy security risk profiles. Textual policy analysis comparing the pre-2022 and post-2022 national security doctrines and National Energy and Climate Plans (NECPs)’s examines whether rhe 2022 energy crisis borne of the Russo-Ukrainian conflict and Russia’s sustained hybrid energy warfare against the EU has compelled EU member states to incorporate decarbonization as a security imperative within their official national security doctrine and whether this framing has resulted in an acceleration of the state’s decarbonization targets under explicit security-based justifications. Ultimately, this study finds that, across Europe, exposure to the 2022 energy crisis and Russian hybrid energy warfare universally motivated the reconstitution of decarbonization as a national security imperative and the acceleration of state decarbonization targets under this securitized logic. The findings of this thesis thus unseat the longstanding epistemological segregation of climate policy and national security, advocating that moments of geopolitical volatility can converge state security and climate interests and thus offer an opportunity for climate progress under security-based incentives.

This thesis examines how extreme-right groups in Europe adopt and use cryptocurrency, with a focus on variation across ideological subcurrents. Drawing on a mixed-methods analysis of 53 far-right Telegram channels, it finds that cryptocurrency adoption is uneven and shaped by both ideological alignment and organizational structure. Groups that emphasize anti-establishment narratives and distrust of centralized financial institutions are more likely to adopt cryptocurrency, particularly when they possess strong digital infrastructures and transnational networks. The findings further show that cryptocurrency is used primarily for organizational rather than operational financing, supporting long-term activities such as propaganda, recruitment, and network maintenance. Variation across subcurrents suggests that ideological compatibility plays a critical role in shaping adoption patterns, sometimes outweighing purely instrumental considerations. By highlighting the interaction between ideology and organizational capacity, this study challenges predominantly functional explanations of cryptocurrency use and offers a more nuanced account of how financial technologies become embedded within extremist ecosystems.

This thesis seeks to examine the impact of the Temporary Protected Status (TPS) 2021, TPS 2023, and the Cubans, Haitians, Nicaraguans, and Venezuelans (CHNV) humanitarian parole programs as instruments of changing legal status markers in their effects on the economic and labor market integration of Venezuelan migrants to the United States. To accomplish this goal, this paper proposes a research design that applies difference in differences estimators with both GLM logistic regressions and OLS linear regressions to 5-year pooled cross-sectional ACS data. The advantages of this design are that it is both highly replicable and provides an innovative solution to imputing legal status among Venezuelan migrants across the last five years, as the three key policies have had direct, temporally defined impacts on that status. These multiple, rapid-fire policies / changes in legal status offer a unique opportunity to address a gap in the existing literature whereby heterogeneities in legal status are underexamined as a complex and textured mechanism impacting migrant integration. The design continues advancing a literature that seeks novel quantitative solutions to the study of economic integration while still offering nuanced implications that are often elusive in the more statistically robust datasets available to quantitative researchers. I find that at least the TPS 2021 policy had a positive impact on the earnings of those non-citizen migrants with income before and after the policy took effect.

In the summer of 2025, the Supreme Court decided in Mahmoud v. Taylor that the use of storybooks in Montgomery County, Maryland elementary schools that included representations of LGBTQ+ characters and of gender diversity without the possibility for parents to opt their children out of reading them for religious reasons was a violation of parents’ Free Exercise rights under the First Amendment. The decision represented a stark departure from precedent that distinguished between “mere exposure” to an action or to content that might violate a person’s religious beliefs and “coercion to perform an action” that violated religious beliefs, especially in a school context. The Court ruled that the very act of teaching the books in class would have a coercive effect on students and would burden parents’ Free Exercise rights in the absence of an opt-out policy. This thesis focuses on a specific aspect of the decision’s indirect effects – the decision-making processes of principals and superintendents on the front lines of the “parents’ rights” movement as they navigate the space between state policy and community input in deciding whether or not to grant requests from parents and guardians for students to “opt out” of learning materials that they may find religiously objectionable. Through a survey sent to principals and superintendents in New Jersey as well as through in-depth interviews with administrators, this thesis identifies that administrators are strongly influenced by the composition of their district, allowing opt-outs beyond the scope of state law and anticipating requests from “frequent flyer” parents and guardians. In their decision-making process, superintendents and principals weigh the burdens of enforcement against the potential for backlash, litigation, and distrust from members of the school community.

In 2022, Portland, Oregon became the first American city in decades to adopt the Single Transferable Vote (STV) in multi-member districts. STV , the multi-winner form of Ranked Choice Voting (RCV), is a proportional representation (PR) system designed to allocate seats in a legislature in proportional to a party or group’s share of the electorate. STV has grown increasingly popular in recent years as a means of increasing proportional representation and voter choice. Today, Portland and Cambridge, Massachusetts are the only two U.S. cities using multi-member STV in municipal elections. Crucially, the innovation of Portland is to combine multi-member STV with geographic districts in order to increase the voting power of geographically concentrated communities.

However, from adoption to electoral outcomes, STV in Portland has functioned in ways that standard electoral reform theory neither predicted nor anticipated. This thesis argues that STV’s effects are not determined by electoral rules alone, but are fundamentally shaped, and in turn shaped by, structural design features and the political and institutional contexts under which it operates. Drawing on historical analysis, comparative endorsement and vote transfer data, and an adapted Loosemore-Hanby Index, this thesis examines Portland’s City Council across three dimensions: its adoption as defying the party parity hypothesis and shifting-coalitions theory; its organizational effects as shaped by Portland’s urban regime; and that proportionality improved under STV but minority representation declined.

Together, these findings suggest that, in treating electoral rules as deterministic rather than contingent on contextual factors, existing theories and frameworks fail to anticipate patterns of organizational coordination, dynamics of political competition, and variance in the proportionality of representative bodies. As such, understanding STV reform in Portland requires a framework that centers specific political contexts as a defining force shaping electoral outcomes.

International sovereign asset recovery sorts vast sums of looted state wealth into one of three destinations: the hands of the public, where it always should have been; the dubious world of Moneyland, which facilitates the global circulation of illicit assets; or “Asset Purgatory,” where $20 million beachfront villas seemingly sit vacant in perpetuity. This thesis posits that international asset recovery is not an automatic consequence of exposing the illicit holdings of corrupt heads of state. Conventional international mechanisms to recover illicit assets held in foreign jurisdictions often require legal efforts to be initiated by the state seeking repatriation. In response to this frequently inequitable dynamic, this thesis claims that successful international asset recovery through conventional pathways is only feasible under a set of three interdependent conditions: (1) multiple forms of international recognition, (2) state capacity, and (3) political stability. When these conditions are weak, asset recovery either stalls, or proves altogether unsuccessful. This argument is shaped around a normative definition of “success” in asset recovery to mean that restitution is bestowed upon the population from whom these assets were originally extracted. Chapter 1—situated in the context of postcolonial statebuilding—develops a theoretical framework on illicit forms of extraction, which leads us to the contemporary legal terms under which illicit enrichment is defined. Chapter 2 distills a comprehensive map of the procedures and actors under which conventional asset recovery operates, then, crucially, draws attention to the cases it excludes. Within this category lies the Democratic Republic of the Congo, Tunisia, and Equatorial Guinea, each analyzed as comparative case studies by which cautionary lessons or innovative approaches instruct the fourth and final case study: Libya. As the country with the most wealth at stake and the least reported success in reclaiming it, Libya stands out as a unique and urgent focal point of this research. Expert interview research I conducted with representatives from private, NGO, and state-based asset recovery agencies in Switzerland informs the normative and geographic scope of this thesis.

This thesis argues that Neolin, the Delaware Prophet, generated pan-Indian solidarity in 1760- 1763 through a mechanism fundamentally different from diplomatic alliance: the cosmological articulation of shared human suffering. Drawing on multi-source triangulation across four independent primary source accounts of Neolin’s vision (Robert Navarre’s journal of the siege of Detroit, John Heckewelder’s firsthand missionary account, James Kenny’s Quaker trader journal, and John M’Cullough’s captivity narrative), and employing process-tracing methodology applied to administrative correspondence from the Papers of Sir William Johnson, the thesis reconstructs the conditions of life in the Ohio Valley that Neolin’s prophecy addressed: material deprivation following the collapse of the French trade alliance, geographic displacement, the systematic denial of political standing through the weaponization of the Iroquois “women” designation, and the failure of existing diplomatic channels. The thesis demonstrates that Neolin’s vision constituted a complete governance framework, prescribing how communities should eat, drink, form families, conduct spiritual practice, and relate to outsiders. This framework generated solidarity across tribal boundaries not through negotiated interest but through shared recognition of common experience, as evidenced by the cascade of independent, simultaneous military actions in 1763 that occurred without centralized coordination. The uprising compelled the Proclamation of 1763 and the Plan of 1764, policy consequences wildly disproportionate to the formal political power of the communities involved. The thesis engages the scholarly traditions of Parkman, Peckham, White, Dowd, Cave, Richter, Edelson, and Holton, identifying their shared analytical commitment to institutional frameworks as a limitation that obscures the experiential dimension of political mobilization.

The policy implications address the persistent gap between formal government-to-government consultation frameworks (NHPA Section 106, Executive Order 13175) and the experiential registers through which Indigenous community solidarity actually forms, drawing on the Standing Rock movement and the trajectory from the 1974 International Indian Treaty Council through the 2007 UN Declaration on the Rights of Indigenous Peoples to demonstrate the contemporary relevance of this eighteenth-century case.

This thesis investigates the institutional drift of the Foreign Agents Registration Act (FARA), a transparency and disclosure statute enacted in 1938 that requires individuals and organizations acting on behalf of foreign principals to register with the Department of Justice. Despite its original intent to serve as a neutral, universal disclosure mechanism, this thesis finds evidence suggesting that enforcement and registration patterns of the FARA have instead clustered around countries that are U.S. adversaries. This observation suggests evidence of institutional drift. Using a mixed-method approach including archival research, a hand-coded enforcement dataset, and a panel fixed effects regression using the Department of Justice’s bulk registration dataset, this thesis finds consistent evidence suggesting institutional drift away from FARA’s original universal disclosure intent. This thesis furthermore observes that authoritarian governments have cited the FARA in enacting repressive foreign agent laws stigmatizing civil society and independent media. The end of this thesis features two policy recommendations.

Investigating the 2014 West Virginia Water Crisis provides an opportunity to explore where limited research has been conducted. Due to schools in 9 counties being directly impacted by ‘do not drink’ orders, many students were kept out of the classroom for multiple weeks in January of 2014. While a decrease in test scores relative to the rest of the state was to be expected, this thesis uses a mixed methods approach to investigate the impacts of the crisis and understand why there may have been a discrepancy between how many weeks of school students missed and how much learning was predicted to have been lost. For the quantitative aspect, I used a Difference-in-Differences (DID) model to estimate the magnitude to which there was some divergence in mean standard deviations from the rest of the state’s standardized assessment scores. Based on the magnitude of the coefficient for the DID model being -0.062 standard deviations, there is an indication that across the nine counties impacted most by the chemical spill, students’ learning losses were worth approximately four weeks of missing in-class instruction compared to students in non-impacted counties. In reality, schools in the impacted counties lost between two and three weeks when accounting for harsh winter weather that afflicted some of the same counties impacted by the chemical spill. To account for the estimated additional week of lost time in the classroom, the qualitative content analysis of this thesis theorizes that there was another reason for students to miss school in the impacted counties; The breakdown of the social contract between citizens, the government, and industrial powers led to a substantial lack of trust amongst civilians. When applying this dynamic to social capital theory, I posit that there may have been an increase of absenteeism, due to the erosion of the trust component of the social contract, amongst some families in the impacted counties that would then encourage them to keep their children out of schools beyond their reopening; this would lead to a trend of lower test scores after 2014 compared to counties in West Virginia that were not impacted.

The People’s Republic of China has developed tremendously through the US-led international order. It has become the world’s second largest economy, become included in international institutions, and is an international stakeholder. However, Beijing has used its newfound economic and military power to coerce other countries. This paper aims to understand how two other international stakeholders, the United States and Japan, viewed China during its ascent to power, in order to understand some of the limitations of viewing this relationship through a liberalist lens. By analyzing the National Security Strategies of the United States, and the Diplomatic Bluebooks of Japan, we are able to see that the belief that China would liberalize and become a supportive member of the international order led to economic co-binding and an oversight on the indications that the PRC would not do so. Under the Trump and Biden Administrations, the United States and Japan turned to balancing China’s rise through increased military and economic cooperation. This paper concludes that the United States and Japan have shifted towards a neo-realist perspective towards China, aiming to constrain, deter, and detach rather than integrate with Beijing. The policy proposed within this paper is for the US-Japan integration to move into US Naval Shipyards.

Local jails have emerged as central sites for immigration enforcement under the 287(g) program, which trains officers to perform federal immigration enforcement. As of December 2025, 806 sheriff’s offices across the country operated under one or more of the three 287(g) models, with 74 sheriffs adopting the Jail Enforcement Model in 2025. The Jail Enforcement Model (JEM) empowers local law enforcement to issue immigration detainers, with ICE approval, to people already in jail, holding them for pick up by Immigration and Customs Enforcement [ICE]. Federal officials describe this program as a “force multiplier” because thet reinforce the ability of ICE officers to identify and detain potentially removeable noncitizens who have come in contact with the criminal justice system. This thesis examines whether JEM adoption could be associated with differences in custody transfer percentages, defined as the percentage of issued detainers on people in jail resulting in federal custody, across 29 Texas county jails. I asked two research questions: (1) Do offices that have entered into JEM 287(g) agreements with the federal government (JEM-only) experience higher custody transfer percentages compared to non-287(g) counties? (2) What does the immigration detainer process look like in JEM-only versus non-287(g) counties in Texas? I hypothesize that JEM-only counties have higher custody transfer percentages than non-287(g) counties because of the “force multiplier” logic. To answer these research questions and test this hypothesis, I employed a mixed-methods approach, combining quantitative analysis of ICE detainer and detention data with qualitative interviews with two sheriffs, a sergeant, and four jail officers across six of the 29 counties. For the quantitative analysis, I used ICE detainer and detention data from 29 Texas counties between September 1st, 2023 and October 15th, 2025 based on the most recent ICE data available, comparing nine JEM-only counties with 20 non-287(g) counties. This analysis found that JEM-only counties and non-287(g) counties demonstrate statistically similar custody transfer percentages: 73.32% vs 78.90%, respectively. The qualitative analysis is based on six semi-structured interviews to investigate potential explanations for my quantitative findings. This analysis identified two mechanisms that might account for why JEM-only counties and non-287(g) counties experience similar custody transfer percentages: (1) Under Texas law, Texas county jails must honor immigration detainers, no matter 287(g) adoption, (2) ICE’s prosecutorial discretion to close certain cases. However, the small number of interviews and non-random selection based on access and responsiveness introduce significant limitations to generalizability.
This thesis ends with a discussion of implications for understanding JEM and 287(g) as immigration enforcement policy as well as outlining avenues for future research and recommendations aimed at standardizing delegated powers and protecting vulnerable populations.

This thesis asks whether the Finch West Light Rail Transit line, or Line 6, meaningfully advances transit equity in Toronto by expanding the real opportunities available to residents in historically underserved neighborhoods. It situates the project within the broader context of the city’s growing income polarization and argues that these patterns are not just economic but spatial, shaping who can access reliable mobility and who cannot. Drawing on the idea that inequity is embedded in the built environment, and using Neighbourhood Improvement Areas as a policy lens, the thesis treats the Finch West corridor as a critical case for understanding how transit investments operate in places where need is already concentrated.

Rather than relying only on standard transit measures such as ridership or travel time, the analysis uses a framework grounded in the capability approach to ask whether infrastructure actually translates into reliable, safe, and dignified access to everyday destinations. It considers both the promises of Line 6 and the realities of how it has been delivered, including the disruptions experienced during construction and the ongoing constraints created by suburban form and service design. The findings suggest that while Line 6 represents a significant investment in an underserved part of the city and may improve mobility along the corridor, its impact on real freedom is more limited than its framing might suggest. These limits are tied to broader structural conditions, including land use patterns, gaps in service integration, and the uneven ways infrastructure is experienced on the ground. The thesis concludes by outlining policy recommendations and reflecting on what this case reveals about the challenges of using transit investment as a tool for addressing urban inequality.

In a world where growing numbers of people are turning to AI chatbots for various forms of advice, victim-survivors of domestic violence and sexual assault are also beginning to seek out AI chatbots for emotional support and legal advice. The National Domestic Violence Hotline now offers an AI chatbot “Ruth” — in addition to their existing telephone, virtual chat, and text-support options — to specifically support victim-survivors as they navigate DV/SA situations. Other DV/SA-specific AI chatbots have found favor among victim-survivors for their assistance in gathering evidence, completing legal documents, offering empathetic emotional support, organizing and recording timelines of incidents, and connecting victims to available human resources.

As the use of AI chatbots in DV/SA contexts only becomes more widespread, policymakers have a responsibility to investigate 1) how victim-survivors are utilizing these AI chatbots, 2) how competent the AI chatbots are in providing accurate information, and 3) what gaps or issues remain for policymakers to address. This thesis tests both general-purpose chatbots (ChatGPT and Gemini) and DV/SA-specific AI chatbots (Aimee Says, Sophia, Hope Chat AI, and Ruth) to evaluate their effectiveness at providing emotional support and legal assistance in common DV/SA contexts. In comparison to the expected guidance that would be offered by a trained, human DV/SA advocate, this research found that AI chatbots do not currently offer services and information at an equivalent level. In some tests, the AI chatbots in fact offer insufficient information or improper assistance, potentially leaving victim-survivors vulnerable to a number of safety and support gaps.

An interview with a frontline DV/SA social worker — from a DV/SA nonprofit organization in New Jersey — is also conducted, summarized, and analyzed. The interviewee discusses how victim-survivors are largely utilizing AI chatbots for extra support in conjunction with formal support from counselors and attorneys — rather than as the sole source of emotional or legal support. However, there remain major legal obstacles (related to confidentiality and privacy concerns) which restrict DV/SA organizations from adopting and encouraging the use of AI chatbots by victim-survivors.

Drawing from the results of the AI chatbot experiments and my interviewee’s discoveries, three policy recommendations are made to assist victim-survivors, policymakers, and frontline DV/SA response teams alike in preparing for future AI chatbot use in DV/SA contexts. In order to reduce overall implementation costs, protect victim-survivor’s confidentiality, and serve the greatest number of victim-survivors possible, this thesis suggests a final policy proposal to make changes to the current National Domestic Violence Hotline’s AI Chatbot, “Ruth,” adopting the strengths of the other chatbots and seamlessly incorporating into the websites of DV/SA nonprofit organizations across the country. A general recommendation is also made to train current DV/SA advocates and frontline responders on victim-survivors’ uses of AI chatbots, so that they can be prepared to properly guide victim-survivors when they predictably encounter victim-survivors’ use of AI tools in the future.

The primary purpose of this senior thesis was to explore how the federal government can improve the effectiveness of multijurisdictional wildfire investigations. This study began by engaging in a brief examination of the previous century of wildfire policy in the United States. It then reviewed the existing literature on methods and strategies for collaboration in wildfire governance, the current policy context of wildfire investigations, and the extent of the wildland arson problem in the United States. Following this review, two case studies were conducted on the investigations into two high-profile wildfires that occurred over the last 20 years. These fires included the 2006 Esperanza Fire in California and the 2021 Marshall Fire in Colorado. By comparatively analyzing the successes and challenges encountered during both of these investigations, it was found that two main factors heavily influence the effectiveness and reliability of multijurisdictional wildfire investigations, regardless of their cause: 1) the quality and successful interpretation of scientific evidence; and 2) the ability of investigators to collaborate across different agencies, which is dependent on the amount of resources available and the way agency relationships are defined in cooperative agreements. Based on these findings, four specific recommendations were proposed for Congress to shape the future direction of federal policy towards multijurisdictional wildfire investigations. These recommendations include: 1) authorizing the creation of a separate division dedicated solely to wildfire investigations within the recently established United States Wildland Fire Service; 2) incorporating investigative functions into the primary duties of the Wildfire Intelligence Center outlined under the 2025 Wildfire Intelligence Collaboration and Coordination Act to support the work of the proposed wildfire investigation division; 3) allocating funding for federal fire science research initiatives related to the study of fire pattern indicators within the proposed wildfire investigation division; and 4) approving President Donald Trump’s effort to consolidate federal wildland firefighting capabilities across the United States Department of the Interior and the United States Department of Agriculture Forest Service to grant the United States Wildland Fire Service a statutory base.

This thesis contributes to the ongoing debate over global AI leadership by providing a quantitative assessment of the performance gap between the United States and Chinese large language models (LLMs) and the macroeconomic factors underlying this gap. As artificial intelligence becomes increasingly embedded into society, from economic to military domains, understanding which technology ecosystem, Western or Eastern, will shape its future is becoming increasingly important. To address this question, the study constructs a Composite Performance Index (CPI) to provide a standardized measure of model performance across multiple benchmarks and domains, enabling direct cross-country comparison. Using this framework, the analysis evaluates both frontier LLM performance and average performance trends over time. The results show that while the United States maintained an early lead in frontier LLM development, Chinese models rapidly converged and closed the frontier gap by mid-2025. At the same time, the average performance of Chinese models overtook that of US models by mid-2024, suggesting a broader convergence of the LLM ecosystem. However, following this period of convergence, Chinese frontier model performance shows signs of plateauing, which is attributed to a shift in national AI priorities alongside ongoing restrictions on access to advanced semiconductors. The study finds that access to advanced semiconductor chips is the primary constraining factor on frontier performance, while other inputs, including energy, investment, and human capital, function as enabling factors rather than binding constraints. Moreover, the findings suggest that differences in policy priorities and institutional strategies, rather than resource availability alone, play a central role in shaping the trajectory of LLM development. The thesis is structured as follows. Chapter 1 introduces the research question and key hypotheses. Chapter 2 provides background on LLM development and differences between the US and Chinese LLM ecosystem. Chapter 3 outlines the conceptual framework and methodology of both parts of this study. Chapter 4 presents the construction and measurement of the US-China LLM performance gap, while Chapter 5 evaluates its underlying drivers. Finally, Chapter 6 concludes with a summary of findings and policy implications.

This thesis examines the evolution, transformation, and decline of federal place-based economic policy in the United States through the case of Central Appalachia It asks how policymakers have understood the relationship among geography, economic development, and the role of the state, and how successive policy regimes attempted to resolve persistent regional inequality. I argue that the trajectory of place-based policy reflects a fundamental shift from a mid-twentieth-century framework that treated regional inequality as a problem of underdevelopment rooted in place to a late-twentieth-century framework that treated it as a problem of individual adjustment to structural economic change. This shift was poorly matched to regions such as Central Appalachia, where immobility, institutional weakness, and economic concentration constrained adjustment and contributed to cumulative decline. I employ a qualitative historical methodology, drawing on policy documents, archival sources, secondary literature, and case analysis to trace the intellectual, political, and institutional evolution of federal interventions. I integrate insights from political economy with economic and sociological perspectives on spatial inequality. Chapter I argues that Central Appalachia’s persistent poverty has its roots in extractive political and economic institutions that have dominated the region since its settlement, limiting diversification, weakening state capacity, and constraining subsequent policy effectiveness. Chapters II and III analyze the New Deal and War on Poverty, focusing on the Tennessee Valley Authority and the Appalachian Regional Commission as major place-based interventions. These programs’ large-scale public investments achieved durable gains by expanding infrastructure and enabling market activity. Chapter IV documents a later shift toward a federal policy apparatus featuring both localized, market-oriented place-based policies and people-based adjustment programs such as job training and income support. I argue that the new regime proved incomplete: place-based policies functioned as inefficient subsidies to mobile capital, while adjustment policies were never sustained at a scale sufficient to facilitate meaningful adaptation. In Central Appalachia, this approach failed to produce convergence. The decline of place-based policy was not the result of inherent ineffectiveness, but of a broader shift that misunderstood both mobility and regional economic dynamics. Major coordinated place-based interventions have been effective when they address structural constraints and are supported by sustained political commitment. By contrast, both the late-century adjustment regime and its localized place-based components failed to resolve spatial inequality. Recent revival of place-based policy recognizes these failures and the social and political consequences of geographically concentrated economic distress. Effective future policy will require combining place-based investment with credible support for individual adjustment, while recognizing the enduring role of place in shaping economic opportunity.

Cesarean delivery accounts for roughly a third of all births in the United States, at proportion that continues to grow annually. Yet, the degree to which its use reflects a justified use on the grounds of clinical necessity, versus an improper use motivated by non-clinical forces remains contested. This thesis investigates whether the probability of cesarean delivery differs by insurance type among low-risk patients in New York State, and whether that difference arises from the hospitals where Medicaid and commercially insured women deliver or from how they are treated within the same facility. Drawing on roughly 2 million inpatient delivery discharge records from the New York SPARCS database spanning 2016 through 2024, then matched to institutional data from the CMS Provider of Services file, the study constructs three analytic cohorts at varying levels of clinical stringency and estimates hospital fixed-effects linear probability models to decompose the observed payer gap.

The results indicate that commercially insured patients undergo cesarean delivery at meaningfully higher rates than Medicaid patients across all three cohorts, with the primary low-risk cohort showing a gap of 5.5 percentage points. Contrary to what a pure sorting explanation would predict, the gap does not attenuate when hospital identity is held constant; instead, the within-hospital estimate slightly exceeds the aggregate figure, suggesting that between-hospital sorting marginally masks, rather than inflates, the disparity. The gap has also widened over the study period, driven almost entirely by rising cesarean rates among commercially insured patients while Medicaid rates remain flat within the same institutions. Among hospital characteristics, the urban-rural distinction emerges as the only statistically significant moderator: the payer gap is roughly two percentage points narrower at rural facilities than at urban ones.

These findings point to a disparity rooted in the clinical encounter, one that existing Medicaid-focused reduction policies are not structured to reach. The thesis concludes with recommendations for patient visibility initiatives, cross-payer incentive alignment, reimbursement reform, and standardized labor management protocols designed to address the commercial-side trajectory that current policy overlooks.

This thesis examines how the political coalitions governing Los Angeles have shaped the city's housing policy, tracing the structural shift from publicly financed housing to market-mediated intervention. Drawing on Clarence Stone's regime politics, Harvey Molotch's growth machine theory, and Brenner & Theodore's concept of actually existing neoliberalism, I argue that a pro-growth coalition of real estate interests, media figures, and homeowner associations systematically dismantled the city's capacity for direct public housing production over the course of the twentieth century. Without this capacity, the city turned to incentive zoning policy, adopting the Transit Oriented Communities (TOC) program in 2017, which offers density bonuses and streamlined approvals to developers who include affordable units in projects near major transit stops. Using logistic regression models on a block-group-level dataset that merges parcel records, TOC project approvals, and American Community Survey data, I test whether TOC development follows the geographic logic of the growth machine. The results indicate that TOC projects are primarily driven by transit tier incentives and market conditions, concentrating in areas with high home values and large renter populations rather than in the block groups with the most severe housing cost burdens. Cost burden is negatively associated with TOC development, and a selection bias test shows that the TOC zones themselves are drawn in areas where market conditions already favor development. While the program has generated meaningful affordable housing production, it cannot direct development to where need is greatest, producing housing where the growth machine finds it profitable to build. These findings demonstrate that the disjunction between TOC's output and the geography of housing need is not a design flaw but a structural consequence of the political history that produced the program, with implications for any city relying on incentive-based tools in the absence of public institutional capacity for housing production.

Over the past decade, Egypt has diversified its strategic partnerships to decrease its overreliance on the United States and pursue its national interests. These partnerships have deepened in the context of shifting U.S. influence and interests, increased Chinese and Russian engagement in the Middle East, regional instability, and great power competition. To understand how Egypt has maintained the partnerships throughout its foreign policy, this thesis applies hedging as a framework to explain Cairo’s strategic partnerships from 2016 to 2025. This study shows that Egypt’s relationships with the United States and rising powers are driven by three goals: risk mitigation, maximizing benefits, and strategic autonomy. Furthermore, the strategy is not only driven by interests, but by the need to navigate tensions within each partnership and regional instability that threatens national security. As a result, this study finds that Egypt’s hedging is sustained by mutual benefits across all three partnerships but is constrained by external pressure and creates long-term dependency that undermines the strategic autonomy it pursues. Furthermore, this thesis provides policy recommendations for the United States on how to strengthen the mutually beneficial partnership within the context of Egypt’s hedging strategy and its potential consequences.