%{search_type} search results

This dissertation consists of three essays in microeconomic theory. They all concern strategic decisions or interactions under incomplete information. The first chapter, Armed Conflict Bargaining, is coauthored with James Fearon. In armed conflict bargaining, states (or governments and rebel groups) may exchange offers while they are fighting, and conflict can end either with a negotiated deal or a military victory. If fighting is driven by private information about military capabilities or costs relative to benefits, why doesn't the possibility of frequent offers lead to rapid learning and convergence on a settlement, as tends to occur in standard buyer-seller bargaining models? We show that when private information concerns military capability, equilibrium war durations can be quite long for reasonable parameter values, and the conditional probability of negotiated settlement is U-shaped. Armed conflict bargaining with private information about capabilities is an interdependent values (aka ``lemons") problem, but one that differs from buyer-seller contexts in that some fighting can directly reveal information about the value of more fighting. We also show that if, as makes sense for this setting, the sides can't contract to implement deals implying ``ex post regret, " then in equilibrium there can be long stretches of pure fighting with no exchange of serious offers. This is typically the case for both interstate and civil wars. The second chapter, Optimal Ordering in Sequential Auctions, is coauthored with Fanqi Shi and Yiqing Xing. It studies the optimal ordering of first-price and second-price (sealed-bid) sequential auctions. In the case of positive one-dimensional signal, we show sellers always want to order items in decreasing spread when such ordering exists. Moreover, the seller can achieve either fully efficient outcome or optimal revenue. Our analysis provides a (partial) explanation for the prevalence of sequential auctions. The third chapter, Robust Contracts with One-Sided Dynamic Impact, studies a repeated moral hazard setting in which the principal has nonquantifiable uncertainty and evaluates contracts by worse-case performances. In each period a distinct agent appears, and the principal knows only that the agent's action set contains a known base action set, but does not know the full action set. In addition, we assume the contracts cannot be legally enforced, and thus have to be self-enforcing, meaning that the principal will not renege on the previously agreed payment and this incentive must entirely come from that the potential benefits from the future contractual relationships outweigh the short-run payoffs from taking away the payment. In this case, a simple linear or affine contract does not work due to the lack of self-enforcement. We show that an also relatively simple form of contracts, ``capped-linear'' contracts, will satisfy the self-enforcement requirement and is optimal.

In recent years, the world has increasingly witnessed international conflict along ideological fault lines. Western policymakers warn that authoritarian countries like Russia and China are seeking to exploit divisions within democratic societies to promote autocratic tendencies, while for decades, authoritarian countries have accused the West of doing the same--of manufacturing domestic uprisings as a way to force liberalism upon them. Of course, these dynamics herald back to the Cold War, where international conflict was defined by ideological competition between capitalist and communist regimes, and before the Cold War, to the Concert of Europe, where conflict often involved clashes between republican and absolutist regimes. And yet, while history is filled with examples of conflicts along these kinds of ideological lines, there is no consensus in existing research on whether these cleavages have any effect on relations between states. Depending on where one looks, one can find arguments suggesting that ideology does or does not matter, or that there may be effects unique to liberalism. This dissertation advances our understanding of the relationship between ideology and international conflict by exploring how it affects a specific aspect of modern interstate conflict: disputes over the leadership and institutions of other countries. I argue that ideological cleavages increase the risk of these regime disputes in two ways. First, ideologically dissimilar states may attempt to change each other's regimes either due to normative preferences or the fear that these preferences exist. Second, they may resort to subversion to prevent political contagion from the success of alternative regimes. By contrast, ideologically-similar states face incentives to prop each other up against domestic threats and to refrain from using subversion to pursue their other foreign policy goals. I employ a variety of research techniques to provide evidence of these arguments, including cross-national statistical analyses, multi-country archival research, and automated text analysis of foreign propaganda. Using large-N quantitative data, I show that pairs of ideologically similar states are less likely to have policy and regime disputes, more likely to provide aid to each other in periods of domestic unrest, and have overall more cooperative relations than pairs of ideologically dissimilar countries. Focused case studies provide additional evidence of the theoretical mechanisms. Taken together, the results illustrate the shortcomings of "black boxing" the types of disputes states have, and they suggest that ideological ties between states are a major driver of patterns of international cooperation and competition

This dissertation investigates how the choice to use economic sanctions undermines threats of force. When faced with a crisis, leaders must choose among a menu of coercive strategies, and the choice of strategy conveys information about the leader's resolve. While some scholars and policymakers look to sanctions as tools for escalation, I argue that sanctions are poor signals of resolve for three reasons. First, sanctions have become attractive instruments for addressing low-level disputes where the use of force would be excessive. Second, sanctions exact pain on opposing states and can be used to ameliorate audience demands to ``do something" in crises where leaders are unwilling to fight. Third, since leaders have the option of directly threatening force, targets should expect the most resolved leaders to do so, and doubt the resolve of leaders who initially apply sanctions as an alternative strategy. This dissertation consists of three papers to support this argument, utilizing formal theory, quantitative analysis, qualitative historical study, and survey experiments to demonstrate why the choice to use sanctions signals a lack of resolve to fight.

How do political institutions affect wealth redistribution initiatives and their efficacy? While influential current theory holds that redistribution should be both higher and more targeted at the poor under democracy, many newly democratic states have failed to implement redistributive policies that would benefit the majority of newly enfranchised voters. Using original data, primarily though not exclusively on land reform and bank and natural resource expropriations in Latin America, I find that redistribution has actually been greatest during periods of autocratic rule. I demonstrate empirically that where institutional constraints to rule are higher, as in democracy, large-scale redistribution is more difficult to implement. But why do some autocratic rulers choose to redistribute while others do not? I argue that when there is a split between a dictator's support coalition and elites out of government that can pose a threat to his rule, the dictator may choose to expropriate rival elites. Simultaneously redistributing much of the assets of those elites to the poor can gain the support of lower classes and reduce potentially destabilizing pressure from below. One important consequence is that dictators who have expropriated tend to survive longer in office than those that do not. Although redistribution is often more likely under autocracy, there are nonetheless cases of redistribution under democracy. When elites are politically weak during the democratic transition process, as during revolution, there is a long-run relationship between democracy and redistribution. So although democracy may sometimes be a credible commitment to redistribution, it is more often captured by elites and does not induce redistribution.

Why do some ethnic groups involve themselves in civil wars, fighting in rebel groups against the state, while others do not? In particular, what explains variation in ethnic group involvement within the same country? The conventional wisdom is that poverty and political grievances are to blame for this involvement and participation in political violence more generally. Such conclusions tend to be based on studies of single cases or large-N studies that analyze a biased sample of ethnic groups. Countering these hypotheses, I propose a theory of ethnic group rebellion that predicts ethnic group involvement in civil war based on sub-national variation in the ability of the state to exercise a monopoly of control over territory and people, and its administrative strength in an area. Using original geographic, economic, and political data collected on an exhaustive set of ethnic groups from 48 African countries, I test hypotheses emanating from my theory and those of traditional explanations in the context of African civil wars between 1980 and 2006. I find that the distance between an ethnic group's inhabited region of the country and the capital city -- an important measure of the state's presence and capacity for control -- is positively related to the probability of involvement in rebellion. The periphery is a high risk region for involvement in both territorial (separatist/secessionist) rebellion and rebellion designed to take over the state and oust the government. The goals and organizational requirements for these two types of rebellion are quite different, and my theory proposes that these differences should attract ethnic groups from opposite ends of the economic spectrum. Territorial rebellions are likely to involve the poorest ethnic groups living in the periphery, as these movements do not require a high degree of organization or funding to continue a low-level insurgency against the state. In contrast, wealthier groups, which are likely to be better organized and better able to act collectively, are more likely to be involved in government takeover rebellions, which require a higher degree of rebel competence and strength to confront the forces of the state directly. The evidence from statistical analysis and case study examination is strongly supportive of the elements of my theory, suggesting that traditional emphasis on poverty, political representation, and transnational ethnic ties as central factors in this type of political violence is misguided and more focus should be placed on the ability of the state to administer its territory effectively and ethnic group responses to variation in this capacity.

This dissertation focuses on how military preparations made 'in peace' shape the dynamics of deterrence and the ability to prevent a crisis from occurring. Resolve has received most of the attention in past studies of deterrence, yet military capability and how it is developed has major implications for deterrence as well as peacetime force posture and security assistance through military alliances. I assemble empirical evidence and provide formal theoretical logic that highlight the role of military preparations in deterrence. In doing so, this dissertation contributes to an area of deterrence studies that has been understudied by past scholars and is of immediate relevence to policy-makers confronting problems of interstate tensions and conflict

States are motivated to do what they can in the present to prepare for long-term threats to their survival. The possibility of facing an overwhelming threat encourages interstate cooperation in the present, but ensuring long-term commitment among states is non-trivial, especially when the cause for cooperation is distant and ill-defined. Despite the challenges posed by both anarchy and uncertainty, states have developed institutions to enhance long-term cooperation, process uncertainty and create a group of like-minded states ready and willing to come to their defense. Some of these security integration institutions, such as alliances, are well known. Others, such as joint military exercises and defense cooperation agreements, are less commonly analyzed. Two of the papers in this three paper dissertation identify understudied security institutions, weapons coproduction and overseas military basing, and present a mix of data sources and methods to show how states utilize these institutions to increase their security ties. The third paper in the dissertation presents a novel measure of security alignment that aggregates existing data on security cooperation and accurately captures security relationships in the international system. Taken together, the papers in this project increase our understanding of how states act in the present to create long-term security commitment among like-minded states

This dissertation extends theoretical debates about the effectiveness of counterinsurgency strategy by looking at how these theories were implemented in practice. Focusing on a particularly important case, this dissertation studies how the US Army implemented population-centric counterinsurgency in Afghanistan from 2009 to 2014 and asks three closely related questions. First, how consistent were military units in their implementation of counterinsurgency strategy? Second, what factors explain variation in soldiers' and officers' preferred counterinsurgency approach? Third, did this variation in counterinsurgency approach have an impact on the capabilities of Taliban forces? These questions are explored in three empirical papers that combine a variety of quantitative methods including Bayesian hierarchical modeling, original surveys, and instrumental variables with qualitative historical analysis. The first paper motivates the dissertation by demonstrating that military units varied significantly in their counterinsurgency approaches using Bayesian hierarchical models and a comprehensive dataset documenting the areas of operation for every NATO battalion deployed to Afghanistan. The second paper examines the causes of this variation and uses an original survey to demonstrate that soldiers' exogenous pre-conflict personal beliefs and values impacted their preferred counterinsurgency approach. The third paper examines the consequences of this variation and uses an instrumental variable approach to demonstrate that, all-else being equal, units which incurred more friendly casualties were more likely to employ enemy-centric tactics which then increased insurgent capabilities and escalated the cycle of violence. The broad take away from this research is that implementation on the ground often diverged significantly from how strategy was conceived by civilian and military leaders. Most importantly, the experiences and personalities of individual commanders often colored how they perceived and reacted to local conditions that were otherwise largely similar. As a result, some military units conducted operations that seemed to align closely with the official strategy while others diverged from this guidance. While some degree of flexibility is certainly necessary to conduct effective operations in such complex and ever-changing environments, this inconsistency may have undermined the effectiveness of international efforts to contain insurgency in Afghanistan. The results stress the importance of screening, selecting, and training soldiers and officers for specific mission requirements. This lesson is also likely to be applicable in other organizational settings where leaders must balance flexibility with consistency, such as domestic policing

This dissertation is a collection of three essays on historical political institutions and contemporary economic issues in China. In the first chapter, titled "State Formation and Bureaucratization: Evidence from Pre-Imperial China", I study the mechanisms of state-building in pre-imperial China. I develop an incomplete contract model of land ownership and military conflict, which produces three predictions. First, external military threat dampens state-building as the ruler needs to incentivize his agent to defend against invasion by giving out ownership to land. Second, military attacks are conducive to state-building as centralized states are more efficient at offense. Third, a decrease in the bargaining position of administrators also facilitates state-building. Empirical tests and historical examples are consistent with model predictions. In the second chapter, titled "Media Attention and Choice of Major: Evidence from Anti-Doctor Violence in China", we study the effect of media persuasion in the context of educational choices. We exploit plausibly exogenous variations in Chinese newspaper articles on anti-doctor violence to identify how media reports affect students' decision to study medicine in college. We collect articles from over 1,200 newspapers and match them with an administrative dataset containing the universe of students enrolled in Chinese colleges from 2005 to 2011. We find that one additional article on violence against doctors leads to a 0.5 percent decrease in the number of students choosing medicine majors, especially majors that train physicians and nurses. Students attending elite colleges and with above-median test scores are more responsive to media attention. Consequently, articles on violence reduce the quality of medical students, measured by their rank in the college entrance examination. In the third chapter, titled "Marital Preferences and Housing Prices in China: Evidence from Online Dating", we estimate the effects of China's surging house prices on individuals' marital preferences and equilibrium assortative matching patterns. Using data from China's largest dating website, we estimate mate preferences based on users' decision to reply to a first-time message from a contact. We find that (1) site users, in particular women, have strong preferences for home-ownership, and increases in housing prices are associated with higher reply rate by women; (2) compared to non-homeowners, homeowners have stronger preferences for home-ownership; (3) home-ownership increases users' competitiveness and this effect becomes more pronounced for men when as housing prices increase; (4) there is weak evidence on the impact of housing prices on equilibrium assortative matching patterns.

What role do negotiations play in the midst of interstate wars? Lacking detailed intra-war data, extant scholarship has largely treated negotiations as an activity that mirrors hostilities on the battlefield. This view of negotiations as a reflection of war is not well-supported by the historical record, particularly for contemporary conflicts. To address these gaps, I gather new daily level data on 1,700 battles and 10,000 diplomatic acts during all interstate wars since 1816. These new resources reveal new insights on the calculated and intertwined relationship between fighting and bargaining. I find that pre-1945 wars are so costly in signaling weakness that tend to be very infrequent but quickly end wars. But after 1945, diplomatic bargaining becomes much more common, irregular, and less predictive of war termination. I contend that international actors and beleaguered war targets use these negotiations to moderate the battlefield, regroup militarily, and mitigate an initiator's first-mover advantage. Negotiations are used as an instrument of war. In addition, a computational analysis of negotiation transcripts and United Nations military reports from the Korean War demonstrates that the costs of war should not be viewed as exogenous information, but rather choices that are endogenous to the stakes of the conflict. These original data and analyses establish a new and dynamic intra-war framework that enables a more policy-relevant study of conflict management, and speaks to the utility of viewing diplomacy as an equal partner to---not a mere consequence of---fighting in understanding war.