Current Working Papers, Code and Datasets
|
True Gini Coefficients: A Research Note
Abstract: Paper Dataset |
|
The Institutional Foundations of Growth and
Inequality (with Lewis S. Davis)
Abstract: We investigate the impact of political and economic institutions on income inequality, and find strong evidence that property rights lead to lower levels of inequality. In light of this new evidence, we call into question previous claims regarding the effect of inequality on growth that have treated inequality as exogenous, failing to account for the central role of institutional variables. We find that the omission of property rights explains the negative relationship between growth and inequality commonly found in cross-country growth regressions. Turning to evidence based on fixed effects estimates in panel data, we show that the inclusion of economic institutions allows for unbiased estimation using a more efficient GLS estimator. While we confirm previous evidence of a positive short-run relationship between inequality and economic growth within countries, the random effects model does not support the position that the short run and long run relationships are the same, suggesting that the short run relationship may be driven by business cycle dynamics |
|
Inequality of Opportunity? Evidence on
the Distribution and Returns to Education Across Countries Abstract: This paper develops a simple model of the market for human capital, and tests the implications of the model using data from a cross-section of countries. Using data on the distribution of educational attainment and returns, this paper is able to offer a more direct test of the hypothesis that credit constraints help to explain variation in inequality across countries than previous work using aggregate income inequality statistics. Mixed evidence is found for the effect of private credit on educational attainment, however; expected lifespan appears to be a much more important factor in explaining investment patterns. International trade emerges as a significant factor contributing to higher returns to education in developing countries |
|
The Determinants of Income Inequality: A
Bayesian Approach
Abstract: Economic theory suggests a wide range of potential mechanisms generating inequality, but there is little consensus regarding the most relevant, nor is there a generally accepted empirical specification for use in Gini coefficient regressions. Drawing appropriate inference from the cross-country data requires formal recognition of this uncertainty in the search for model specification. This paper adopts a Bayesian approach to inference that that averages across models using Bayesian posterior model probabilities to assess the marginal impact of specific variables across a range of potentially likely “true” model specifications. The resulting posterior mean and standard error of each coefficient reflects more accurately our uncertainty regarding both parameter values and model specification. Paper Dataset |
|
Inequality and the Distribution of Human
Capital Abstract: Paper Dataset |
|
Providing Global Public Goods: Incentives and
Determinants in Global Collective Action Problems
Abstract: Economic growth, technology change and the global integration of markets have helped to fuel demand across the world for a stable climate, control of infectious disease and protection from terrorism. Activities such as these have been labeled "global public goods" (GPGs) by scholars and government officials who find the public good paradigm useful for raising concerns that these activities will be underprovided in the absence of centralized coordination and cooperation. This paper addresses two prior theoretical questions upon which concerns of under provision must rest. First, what qualities are needed for a good to be counted as both public and global? Second, how should these goods be produced, provided and financed in a non-cooperative environment (one that lacks a global government capable of the coercive and legitimate threat of force)? We address the first question by drawing on familiar arguments in public economics and political economy. The second question is addressed using arguments from non-cooperative game theory, analyzing the criteria for efficiency in provision. Given that GPGs are still principally provided by national governments to fulfill obligations to residents within their jurisdiction, and that public goods can be provided efficiently without government under certain conditions, strategies that assess and address the incentives of candidate providers may help formulate action to enhance supply. Using cross-country data, the determinants of provision are explored using a number of examples of GPGs including security, climate change, and global health. |
|
Academic Freedom, Intellectual Bias, and “the
Truth”: What Can Bayesian Statistics Teach Us? Abstract: Questions that have been raised regarding academic freedom span a wide array of topics, from the scope of what is covered to the manner in which it is taught. The analysis offered in this paper addresses one aspect of the current debate: how to think about and handle the presentation of contested hypotheses. Instances often arise in which uncertainty exists regarding which of two (or more) hypotheses is "correct,” but what is meant by a correct hypothesis: the hypothesis that is more consistent with the data, or the hypothesis that is true? The second definition is clearly more satisfying, but only the first is something that, given some underlying assumptions, can be reasonably and objectively stated. These two concepts of what it means for a hypothesis to be correct are distinct, but not unrelated. This paper argues that Bayesian statistical theory can help us to understand their relationship, and thus to understand the decision problem that facing educators, students and, most importantly, policy makers considering legislation regarding a so-called "student's bill of rights." |