Papers by Nicholas Weller
SSRN Electronic Journal, 2000
ABSTRACT Case studies are widely used in political science, providing a richness of detail that s... more ABSTRACT Case studies are widely used in political science, providing a richness of detail that standard quantitative studies often lack. Process-tracing case studies have gained significant prominence recently for their promise to use this strength to explicate causal mechanisms behind robust empirical correlations. Effective case selection is obviously central for process tracing to deliver on this promise, but there are no generally applicable methods for case selection when exploring complex relationships that are non-linear and feature multiple causal mechanisms. To fill this gap, we develop guidelines for selecting process-tracing cases that focus on the functional form of the relationship between relevant variables and the distribution of possible causal mechanisms. Through general examples and specific applications, we show how this approach can reveal opportunities for researchers, while failing to account for these factors can result in poor case selection, mistaken inferences, and an inability to generalize.
Journal of Public Policy, 2013
ABSTRACT State capacity is a key concept for research in public policy and political science. Des... more ABSTRACT State capacity is a key concept for research in public policy and political science. Despite its importance, there is no broadly accepted measure of state capacity in the existing literature, and frequently used measures of capacity have not been examined for their validity. We begin with an explicit definition of state capacity – the state's ability to implement public policy – and connect this definition to a measurable outcome of state capacity – the state's taxation of income. We show that this measure, income taxes as a percentage of total tax revenue, is a useful indicator of state capacity and meets higher standards of measurement validity than other tax-based indicators. We also compare our measure to the most common existing indicators of state capacity to show that income taxation is a better theoretical and statistical measure of states’ effectiveness in policy implementation.
SSRN Electronic Journal, 2000
Political consultants play an integral role in modern political campaigns. When candidates hire c... more Political consultants play an integral role in modern political campaigns. When candidates hire consultants to assist with their campaign they create a principal-agent relationship with attendant risks of moral hazard. We argue that the promise of a repeated relationship between a ...
Abstract will be provided by author.
Coordination is a core concern in social science. Problems as diverse as trying to decide where t... more Coordination is a core concern in social science. Problems as diverse as trying to decide where to go to dinner, what political candidate to support or which regulatory policy to adopt all contain coordination as a core element. Most coordination problems arise among actors connected in a network, and these connections can both improve and impede a group's ability to achieve coordination. To model how links influence coordination we distinguish between "constraining edges" that make coordination harder by reducing the number of equilibrium outcomes, and "redundant edges" that make coordination easier by merely increasing communication without affecting the number of equilibria. We show experimentally that the addition of constraining edges reduces coordination, while redundant edges improve subjects' ability to solve a coordination problem.
Proceedings of the ACM Conference on Electronic Commerce, 2011
A growing literature on human networks suggests that the way we are connected influences both ind... more A growing literature on human networks suggests that the way we are connected influences both individual and group outcomes. Recent experimental studies in the social and computer sciences have claimed that higher network connectivity helps individuals solve coordination problems. However, this is not always the case, especially when we consider complex coordination tasks; we demonstrate that networks can have both constraining edges that inhibit collective action and redundant edges that encourage it. We show that the constraints imposed by additional edges can impede coordination even though these edges also increase communication. By contrast, edges that do not impose additional constraints facilitate coordination, as described in previous work. We explain why the negative effect of constraint trumps the positive effect of communication by analyzing coordination games as a special case of widely-studied constraint satisfaction problems. The results help us to understand the importance of problem complexity and network connections, and how different types of connections can influence real-world coordination..
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2013
Quantal response equilibrium (QRE) has become a popular alternative to the standard Nash equilibr... more Quantal response equilibrium (QRE) has become a popular alternative to the standard Nash equilibrium concept in game theoretic applications. It is well known that human subjects do not regularly choose Nash equilibrium strategies. It has been hypothesized that subjects are limited by strategic uncertainty or that subjects have broader social preferences over the outcome of games. These two factors, among others, make subjects boundedly-rational. QRE, in essence, adds a logistic error function to the strict, knife-edge predictions of Nash equilibria. What makes QRE appealing, however, also makes it very difficult to test, because almost any observed behavior may be consistent with different parameterizations of the error function. We present the first steps of a research program designed to strip away the underlying causes of the strategic errors thought to be modeled by QRE. If these causes of strategic error are correct explanations for the deviations, then their removal should enable subjects to choose Nash equilibrium strategies. We find, however, that subjects continue to deviate from predictions even when the reasons presumed by QRE are removed. Moreover, the deviations are different for each and every game, and thus QRE would require the same subjects to have different error parameterizations. While we need more expansive testing of the various causes of strategic error, in our judgment, therefore, QRE is not useful at predicting human behavior, and is of limited use in explaining human behavior across even a small range of similar decisions.
SSRN Electronic Journal, 2000
Rational choice models are commonly used in political science. However, many experimental results... more Rational choice models are commonly used in political science. However, many experimental results find humans do not behave as we would predict from simple rational choice models. In this paper we report on the result of experiments in which subjects both play a standard trust game and make predictions about other players in an experiment, thereby revealing information about how they expect others to behave in the experiment. We find that conditional on their beliefs, subjects play quite rationally in these games. The results suggest we need to understand individuals' beliefs and knowledge when we model behavior in experimental and empirical settings.
SSRN Electronic Journal, 2000
Predictions drawn from game theoretic models of human behavior involve many assumptions about sub... more Predictions drawn from game theoretic models of human behavior involve many assumptions about subjects' beliefs and how those beliefs translate in to action. Despite the importance of beliefs to predicting behavior there has been little attention to understanding actual beliefs. In this paper we utilize an experimental approach to study both beliefs and behavior in a trust game. We find that in a standard trust game many people expect to benefit financially from passing money in the trust game. Subjects in the experiment are relatively good at predicting the behavior of others, often by using their own behavior as a guide. Overall, players' beliefs and actions rarely match game theoretic expectations. By looking at both beliefs and behavior we are able to shed light on what people actually believe and do. We suggest that to improve game theory's ability to predict behavior we must understand how people actually reason.
SSRN Electronic Journal, 2000
Non-cooperative game theory is at its heart a theory of cognition, specifically a theory of how d... more Non-cooperative game theory is at its heart a theory of cognition, specifically a theory of how decisions are made. Game theory's leverage is that we can design different payoffs, settings, player arrays, action possibilities, and information structures, and that these differences lead to different strategies, outcomes, and equilibria. It is well-known that, in experimental settings, people do not adopt the predicted strategies, outcomes, and equilibria. The standard response to this mismatch of prediction and observation is to add various psychological axioms to the game-theoretic framework. Regardless of the differing specific proposals and results, game theory uniformly makes certain cognitive assumptions that seem rarely to be acknowledged, much less interrogated. Indeed, it is not widely understood that game theory is essentially a cognitive theory. Here, we interrogate those cognitive assumptions. We do more than reject specific predictions from specific games. More broadly, we reject the underlying cognitive model implicitly assumed by game theory.
SSRN Electronic Journal, 2000
In this paper we focus on a neglected aspect of common pool resource problemswhether or not actor... more In this paper we focus on a neglected aspect of common pool resource problemswhether or not actors in a group can find a solution to the underlying coordination issue. Using a simple networked model of coordination we demonstrate that coordination problems are very difficult to solve when there are few solutions and actors possess local information. However, if the number of solutions does not change, then increases in the number of connections in the network can make coordination more likely to occur. These results suggest connections can influence the ability to reach solutions in two different ways. First, if connections reduce the number of solutions to a problem, then coordination may be less likely. Second, if it is possible to build connections between actors without reducing the number of solutions, then these connections facilitate coordination. These insights can help shed light on how institutions may facilitate international policy coordination.
SSRN Electronic Journal, 2000
In legal, political, and social settings, people must reach a consensus before particular outcome... more In legal, political, and social settings, people must reach a consensus before particular outcomes can be achieved and failing to reach a consensus may be costly. Although many scholars identify conditions that make consensus more or less likely, they typically ignore key features of communication that may affect a group's ability to reach a consensus. In this paper, we conduct experiments that take into account the costs associated with communicating, as well as the difficulty of the decisions that groups make. We find that when there is even a small cost (relative to the potential benefit) associated with speaking and/or listening, groups are much less likely to reach a consensus, primarily because they are less willing to communicate with one another. We also find that difficult problems significantly reduce group members' willingness to communicate with one another and, therefore, hinder their ability to reach a consensus.
ABSTRACT Collective political action, such as protests, riots or social movements, requires the r... more ABSTRACT Collective political action, such as protests, riots or social movements, requires the resolution of both cooperation and coordination problems. Solutions to these problems are widely seen to depend on the network that connects individuals to each other, because this network is a way for individuals to learn about the actions of others and decide if they want to participate. Although there is a general sense that networks and communication are important for collective political action there is little research exploring the relationship between network structure and group behavior. We address that gap in the literature by utilizing an experimental approach that combines both coordination and cooperation in a networked setting. We find that there are considerable differences in collective behavior based on the network structure that connects individuals. In particular, we find in our experiments that more connections in a network and the presence of highly connected nodes can both facilitate solutions to collective problems. This suggests that in building networks it would be useful to either build many connections or create recognizable leaders that can facilitate coordination.
Social Networks, 2014
ABSTRACT Scholars in the social sciences use network theory to study a range of collective action... more ABSTRACT Scholars in the social sciences use network theory to study a range of collective action problems. Often the goal is to identify how the structure of the network affects efforts to coordinate or cooperate, and research suggests that adding connections to a network can improve the performance of groups faced with such tasks. On the other hand, theory and empirics also suggest that additional connections can degrade the performance of a network. If connections can have negative effects then it is important to consider if there are alternatives to adding connections to a network that can also improve network performance. Because a primary function of connections in a network is to disseminate information, providing individuals with more information about the network may act as a substitute for adding connections to a network. We test experimentally whether providing subjects with more information about the structure of networks can improve coordination. We find that a more complete view of the network leads to faster coordination, but the magnitude of this effect depends on network structure. These results suggest that changing what actors know about a network can improve outcomes without having to add connections that may impede overall performance.
The Computer Journal, 2013
ABSTRACT
Journal of Empirical Legal Studies, 2010
In many legal, political, and social settings, people must reach a consensus before particular ou... more In many legal, political, and social settings, people must reach a consensus before particular outcomes can be achieved and failing to reach a consensus may be costly. In this article, we present a theory and conduct experiments that take into account the costs associated with communicating, as well as the difficulty of the decisions that groups make. We find that when there is even a small cost (relative to the potential benefit) associated with sending information to others and/or listening, groups are much less likely to reach a consensus, primarily because they are less willing to communicate with one another. We also find that difficult problems significantly reduce group members' willingness to communicate with one another and, therefore, hinder their ability to reach a consensus.
Uploads
Papers by Nicholas Weller