Abstract. Empirical research on culture and institutions in economics often relies on cross-cultural data to examine historical or contemporary variation in traits across ethnolinguistic groups. We argue that this work has not adequately addressed the well-known problem of cultural non-independence due to common ancestry and show how phylogenetic regression, along with newly available global language trees, can be used to directly account for this issue. Our analysis focuses on Murdock's Ethnographic Atlas, a widely used database of preindustrial societies, with broader implications for any cross-cultural study. First, we show that various economic, institutional, and cultural characteristics in the Atlas exhibit phylogenetic signal- they tend to be more similar among societies with closer ancestral ties - highlighting the non-independence of observations. Second, through simulations in a sample resembling the Ethnographic Atlas, we demonstrate that phylogenetic correlation leads to severe inefficiency of the standard OLS estimator compared to phylogenetic generalized least squares (PGLS). Furthermore, although clustered standard errors partially mitigate the problem, OLS estimation yields unacceptably high type I error rates, frequently detecting a statistically significant relationship where none exists. Finally, we revisit some of the recently published results in a phylogenetic regression framework. In many specifications, PGLS estimates differ markedly from their OLS counterparts, indicating a smaller magnitude and weaker statistical significance of relevant coefficients.
Abstract. Headhunting – the practice of acquiring human heads for ritual purposes – was historically widespread around the world. We hypothesize that headhunting represented a cultural response to frequent inter-tribal warfare and served as a mechanism to train warriors ready to defend their community. The practice was effective since, first, it allowed verification of warrior quality based on performance in headhunting raids and, second, it offered a system of rewards for men to develop and refine warfare skills. We use phylogenetic comparative methods and ethnographic data to empirically investigate this hypothesis in a sample of preindustrial Austronesian societies. Headhunting turns out to be substantially more prevalent in societies exposed to frequent warfare, accounting for shared cultural ancestry and a host of potentially confounding characteristics. Furthermore, Bayesian estimation of correlated evolution models suggests that, consistent with our hypothesis, the adoption of headhunting typically followed increases in warfare frequency and the decline of this practice was preceded by reduced intergroup conflict.
Abstract. Identifying effective treatments and policies early in a pandemic is challenging because only limited and noisy data are available, and biological processes are unknown or uncertain. Consequently, classical statistical procedures may not work or require strong structural assumptions. An information-theoretic approach can overcome these problems and identify effective treatments and policies. The efficacy of this approach is illustrated using a study conducted at the beginning of the COVID-19 pandemic. An information-theoretic inferential approach with and without prior information was applied to the limited data available in the second month (April 24, 2020) of the COVID-19 pandemic. For comparison, a second statistical analysis used a large sample with millions of observations available at the end of the pandemic’s pre-vaccination period (midDecember 2020). Even with limited data, the information-theoretic estimates performed well in identifying influential factors and helped explain why death rates varied across nations. Later experiments and statistical analyses based on more recent, richer data confirm that these factors contribute to survival. Conclusions: An information-theoretic statistical technique is a robust method that can overcome the challenges of under-identified estimation problems in the early stages of medical emergencies. It can easily incorporate prior information from theory, logic, or previously observed emergencies.
Effect of Universal TB Vaccination and Other Policy-Relevant Factors on the Probability of Patient Death from COVID-19
(with Amos Golan and others) SSRN Working Paper, June 2020.
Abstract. The possibility of reoccurring waves of the novel coronavirus that triggered the 2020 pandemic makes it critical to identify underlying policy-relevant factors that could be leveraged to decrease future COVID-19 death rates. We examined variation in a number of underlying, policy-relevant, country-level factors and COVID-19 death rates across countries. We found three such factors that significantly impact the survival probability of patients infected with COVID-19. In order of impact, these are universal TB (BCG) vaccination, air pollution deaths and a health-related expenditure. We quantify each probability change by age and sex. To deal with small sample size and high correlations, we use an information-theoretic inferential method that also allows us to introduce priors constructed from independent SARS data.