This study investigates the small-sample performance of meta-regression methods for detecting and estimating genuine empirical effects in research literatures tainted by publication selection. Publication selection exists when editors, reviewers or researchers have a preference for statistically significant results. Meta-regression methods are found to be robust against publication selection. Even if a literature is dominated by large and unknown misspecification biases, precision-effect testing and joint precision-effect/meta-significance testing can provide viable strategies for detecting genuine empirical effects. Publication biases are greatly reduced by combining two biased estimates, the estimated meta-regression coefficient on precision (1/Se) and the unadjusted average effect.

Meta-Regression Methods for Detecting and Estimating Empirical Effects in the Presence of Publication Selection

Some empirical results are more likely to be published than others. Such selective publication leads to biased estimates and distorted inference. This paper proposes two approaches for identifying the conditional probability of publication as a function of a study’s results, the first based on systematic replication studies and the second based on meta-studies. For known conditional publication probabilities, we propose median-unbiased estimators and associated confidence sets that correct for selective publication. We apply our methods to recent large-scale replication studies in experimental economics and psychology, and to meta-studies of the effects of minimum wages and de-worming programs.

Identification of and Correction for Publication Bias

This paper questions the conventional wisdom that publication bias must result from the biased preferences of researchers When readers only compare the number of positive and negative results of papers to make their decisions, even unbiased researchers will omit noisy null results and inflate some marginally insignificant estimates Nonetheless, the equilibrium with such publication bias is socially optimal The model predicts that published non-positive results are either precise null results or noisy but extreme negative results This paper shows this prediction holds with some data, and proposes a new stem-based bias correction method that is robust to this and other publication selection processes

Publication Bias under Aggregation Frictions: Theory, Evidence, and a New Correction Method

The Effect of Education on Poverty: A European Perspective

This study examines the relationship between firm-level tax evasion and corruption. The literature has considered the corruption-firm tax evasion nexus to some extent, however, the heterogeneous impact of bribes on tax evasion has been largely ignored. We posit that the impact of bribes on tax evasion is conditional on the prevalence of tax evasion. Using firm-level information from the Business Environment and Enterprise Performance Survey (BEEPS) across 25 transition economies for the years 2002 and 2005, we estimate the heterogeneous impact of bribes on tax evasion using the censored quantile instrumental variables technique. The results show that corruption has a larger impact when tax evasion is more widespread. Among other results, firm-level characteristics also show heterogeneous effects across the conditional distribution of tax evasion. In terms of the policy implications, the results suggest that policies that focus on reducing the tax burden and controlling corruption will be more effective when tax evasion is more prevalent.

Corruption and firm tax evasion in transition economies: Results from censored quantile instrumental variables estimation

While numerous studies assess the relationship between education and health, no consensus has been reached on whether education really improves health. We perform a meta‐analysis of 4866 estimates gleaned from 99 published studies that examine the health effects of education. We find that the current literature suffers from moderate publication bias towards the positive effects of education on health. After correcting for publication bias with an array of sophisticated methods, we find that the overall effect size is practically zero, indicating that education generates no discernible benefits to health. The heterogeneity analysis by Bayesian Model Averaging (BMA) and Frequentist Model Averaging (FMA) reveals that the reported estimates can be largely explained by whether the econometric models control for endogeneity of education, the types of data and the differences in health measurements. Our results also suggest that education may not be an effective policy option for promoting population health.

Wangyongxin Zhang

Papers

Does education really improve health? a meta-analysis