Research Papers

To stimulate further work on social interactions that moves the field towards the consideration of aggregate phenomena, first, we review some of the theory and econometrics of social interactions. This part of the chapter will represent a brief synthesis of a large literature. Second, we consider how this body of work may be related to macroeconomics.

While the importance of social interactions has been argued for a range of interesting behaviors, far less attention has been paid to the role of social interactions in aggregate phenomena. There has yet to be any systematic examination of whether social interactions may help explain the standard macroeconomic phenomena of growth and fluctuations, although they may be connected to social interactions.

Within economics, social interaction research constitutes a growing area of study. This research represents a good-faith attempt to introduce substantive sociological factors into economic modeling. There is now an impressive range of applications of social interaction models in microeconomic contexts.

We described a framework for meritocratic and affirmative action policies from the perspective of standard economic notions of efficiency in the context of admissions to public universities.

We wanted to explore the efficiency side of admissions policies and so address the presumption that admission policies should be justified by meritocratic admissions rules.

Affirmative action constitutes a redistribution of college slots toward minority students, particularly for more exclusive institutions, and it affects a large number of minority students.

We provided a general description of the relationship between individual decision problems and aggregate crime regressions. The analysis is designed to elucidate the behavioral and statistical assumptions implicit in the use of aggregate crime regressions for both the analysis of crime determinants and counterfactual policy evaluation.

We wanted to examine the construction and interpretation of aggregate crime regressions to highlight some limitations of using these regressions and, in particular, indicate how empirical findings may be overinterpreted when careful attention is not given to the link between the aggregate data and individual behavior.

Despite recent efforts to employ microeconomic data and natural experiments, aggregate crime regressions continue to play a significant role in criminological analyses.

We describe a set of trade-offs that policymakers face with respect to fluctuations at different frequencies.

Our analysis differs from previous studies in that our objective is to understand to what extent a policymaker is forced, when choosing a feedback rule to stabilize the economy, to exacerbate volatility at some frequencies to reduce them at others, or whether it is possible to reduce volatility at all frequencies.

The theory of design limits provides a mathematical framework for understanding the full range of temporal implications of alternative policies. The design limits in the control theory literature are important in understanding what stabilization policies do and also possess implications for policy design in the presence of policymaker ignorance.

We examined the relationship between the use of regressions and variance contrast methods to uncover social interactions. We illustrated how these methods use different identifying assumptions and are thus complementary. Additionally, we provided formal identification results that extend existing ones for both methods.

We aimed to clarify the complementary nature of regressions and variance contrast methods in identifying social interactions. By doing so, we sought to enhance the robustness and reliability of empirical analyses in this field.

Understanding the complementary strengths of different methods for identifying social interactions helps improve the accuracy and depth of social science research. Our formal identification results provide a stronger foundation for empirical studies, leading to more nuanced and reliable insights into social dynamics.

We investigated the empirical evidence for various growth theories under conditions of model uncertainty. Using model averaging methods, we assessed the roles of fundamental growth theories, macroeconomic policy effects, regional heterogeneity, and parameter heterogeneity in explaining aggregate growth.

We aimed to address the uncertainty regarding the correct growth model and to evaluate the strength of empirical evidence for different growth theories. By doing so, we sought to understand the limitations of cross-country growth regressions in adjudicating these theories.

We highlight that cross-country growth regressions have limited capacity to determine the relative importance of alternative growth theories. Our findings underscore the significance of macroeconomic policy effects and regional heterogeneity, suggesting a need for more nuanced models and approaches in growth economics research.

We surveyed the recent literature on growth empirics, comparing modern empirical findings and methodologies to the traditional growth accounting analyses from the 1960s and 1970s. We evaluated new methodologies, findings, and the critical role of Total Factor Productivity (TFP), statistical convergence, and the search for robust growth determinants.

We aimed to assess the progress and limitations of modern growth empirics compared to traditional growth accounting. By doing so, we sought to identify the areas where modern methodologies have advanced our understanding and where substantive conclusions remain elusive.

Bridging the gap between modern empirical claims and growth theories is crucial. Unifying statistical findings into a coherent growth model and connecting them to economically meaningful notions of convergence can enhance our understanding of growth dynamics and inform effective policy-making.

We provided a set of results on the econometric identifiability of binary choice models with social interactions. Our analysis extended beyond parametric identification to consider identifiability when the distribution of random payoff terms is unknown and examined the effects of unobservable payoff terms and endogenous group membership on identification.

We aimed to address limitations in existing identification results by considering more general scenarios where the distribution of random payoff terms is unknown and unobservable payoff terms and endogenous group membership are present. This approach allows for a broader understanding of identification in binary choice models with social interactions.

Achieving at least partial identification under plausible assumptions enhances the robustness and applicability of binary choice models in social interactions. Our results contribute to more accurate and comprehensive econometric modeling, improving empirical research’s ability to analyze and understand social phenomena.

This paper compares the performance of different policy rules, focusing on simple feedback rules versus rules that are optimal, given knowledge of the correct economic structure and the appropriate loss function for the policymaker. First, we compare rule performance when the correct model is not known. Second, we compare rule performance with respect to the frequency-specific behavior for variables of interest.

We are interested in assessing the case that can be made for optimal rules as opposed to simple rules for policy, focusing on monetary policy rules.

The case for optimal policy rules is weakened when one considers the performance of the rules in the presence of model uncertainty and with respect to frequency-specific effects.

We attempted to develop a framework for the overall evaluation of racial profiling as a public policy.

We took a new approach by focusing on ethical issues that arise when profiling occurs while considering the efficacy of racial profiling with respect to law enforcement goals in a context in which law enforcement objectives represent a subset of the desiderata for public policy.

Our goal was to explicitly address the full range of implications that racial profiling has for society.