Tomas Havranek
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Tomas Havranek has a total of 410 co-authors. Here are the top co-authors by number of shared publications:
Zuzana Irsova
27 shared publications- Zuzana Irsova
Institute of Economic Studies, Faculty of Social Sciences, Charles University, Prague, Czech Republic
11 shared publications Jiri Novak
8 shared publications- Ali Elminejad
Nazarbayev University, Astana, Kazakhstan
5 shared publications - Marek Rusnak
Financial Stability Department Czech National Bank Na Příkopě 28 11503 Prague 1 Czech Republic
5 shared publications - Pedro R. D. Bom
Deusto Business School University of Deusto Bilbao Spain
5 shared publications
Latest Publications
- Open Access
- 10 October 2026
- Published by Center for Open Science in PsyArXiv
Effect of Exercise on Cognition, Memory, and Executive Function: A Study-Level Meta-Meta-Analysis Across Populations and Exercise Categories
Background: Physical exercise is widely promoted for cognitive enhancement, yet prior syntheses have been questioned due to methodological limitations. Objectives: This study evaluated the reliability of evidence supporting physical exercise for general cognition, memory, and executive function, and assessed efficacy across specific population-intervention combinations after adjusting for publication bias. Methods: This study is a study-level meta-meta-analysis of randomized controlled trials (RCTs). We included systematic reviews with meta-analyses of RCTs evaluating exercise interventions against non-exercise controls in humans, identified in a meta-meta-analysis published in 2025. We extracted study-level effect sizes and synthesized the data using robust Bayesian meta-analysis which allowed us to evaluate evidence for and against the presence of the effect, between-study heterogeneity, publication bias, while accounting for model uncertainty.Results: We extracted 2,239 study-level effect estimates from 215 meta-analyses. After excluding suspected duplicates, 1,933 estimates from 201 meta-analyses were included in the overall analyses, which revealed extreme evidence of publication bias and between-study heterogeneity. Adjusted average effects were significantly smaller than unadjusted estimates for general cognition (SMD=0.208; 95% credible interval (CrI) [0.075, 0.318]), and executive functions (SMD=0.088, 95% CrI [0.000, 0.168]). For memory the adjustment differed between adjustment specifications (SMD=0.226, 95% CrI [0.169, 0.282] vs SMD=0.021, 95\%CrI [0.000, 0.217]). While broad prediction intervals spanned substantial negative and positive effects, subgroup analyses identified mostly consistent benefits for specific combinations, such as dance and exergaming for general cognition in non-healthy older adults.Conclusion: The primary limitations of the evidence are severe publication bias and extreme heterogeneity, which appear to inflate previously reported benefits. Consequently, broad claims of generalized cognitive enhancement from physical exercise appear premature. The results support a shift from blanket public health advice to recommendations tailored to specific population-intervention combinations.
- 10 October 2026
- Published by Center for Open Science in PsyArXiv
Open AccessEffect of Exercise on Cognition, Memory, and Executive Function: A Study-Level Meta-Meta-Analysis Across Populations and Exercise Categories
Background: Physical exercise is widely promoted for cognitive enhancement, yet prior syntheses have been questioned due to methodological limitations. Objectives: This study evaluated the reliability of evidence supporting physical exercise for general cognition, memory, and executive function, and assessed efficacy across specific population-intervention combinations after adjusting for publication bias. Methods: This study is a study-level meta-meta-analysis of randomized controlled trials (RCTs). We included systematic reviews with meta-analyses of RCTs evaluating exercise interventions against non-exercise controls in humans, identified in a meta-meta-analysis published in 2025. We extracted study-level effect sizes and synthesized the data using robust Bayesian meta-analysis which allowed us to evaluate evidence for and against the presence of the effect, between-study heterogeneity, publication bias, while accounting for model uncertainty.Results: We extracted 2,239 study-level effect estimates from 215 meta-analyses. After excluding suspected duplicates, 1,933 estimates from 201 meta-analyses were included in the overall analyses, which revealed extreme evidence of publication bias and between-study heterogeneity. Adjusted average effects were significantly smaller than unadjusted estimates for general cognition (SMD=0.208; 95% credible interval (CrI) [0.075, 0.318]), and executive functions (SMD=0.088, 95% CrI [0.000, 0.168]). For memory the adjustment differed between adjustment specifications (SMD=0.226, 95% CrI [0.169, 0.282] vs SMD=0.021, 95\%CrI [0.000, 0.217]). While broad prediction intervals spanned substantial negative and positive effects, subgroup analyses identified mostly consistent benefits for specific combinations, such as dance and exergaming for general cognition in non-healthy older adults.Conclusion: The primary limitations of the evidence are severe publication bias and extreme heterogeneity, which appear to inflate previously reported benefits. Consequently, broad claims of generalized cognitive enhancement from physical exercise appear premature. The results support a shift from blanket public health advice to recommendations tailored to specific population-intervention combinations.
- Open Access
- 4 September 2026
- Published by Elsevier in SSRN Electronic Journal
Trust, Rule of Law, and the Size Premium: Evidence from a Meta-Analysis
Reported estimates of the size premium, the tendency of smaller firms to earn higher average returns than larger firms, vary widely across studies, countries, periods, and designs. We examine whether generalized trust and rule of law help account for that heterogeneity. Small firms are more opaque and more dependent on outside finance, so the enforcement and information environment should matter more for them than for large firms. We study 1,613 reported size-slope estimates from 105 studies and 31 countries. The meta-regressions control for study design, specification, precision, publication context, and market and macro-financial conditions; Bayesian model averaging assesses uncertainty over the control set. The more stable association is with rule of law, and it runs against the intuitive expectation that better legal institutions shrink the premium: stronger rule of law is associated with more negative reported size slopes, hence larger conventional size premia. The trust association is conditional and less precisely estimated: where rule of law is weak, higher generalized trust is linked to less negative reported slopes (and thus a weaker premium), and this link fades as rule of law strengthens. Formal and informal institutions thus help organize part of the disagreement in this literature, although the analysis concerns variation in reported estimates and does not identify causal effects.
- 4 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessTrust, Rule of Law, and the Size Premium: Evidence from a Meta-Analysis
Reported estimates of the size premium, the tendency of smaller firms to earn higher average returns than larger firms, vary widely across studies, countries, periods, and designs. We examine whether generalized trust and rule of law help account for that heterogeneity. Small firms are more opaque and more dependent on outside finance, so the enforcement and information environment should matter more for them than for large firms. We study 1,613 reported size-slope estimates from 105 studies and 31 countries. The meta-regressions control for study design, specification, precision, publication context, and market and macro-financial conditions; Bayesian model averaging assesses uncertainty over the control set. The more stable association is with rule of law, and it runs against the intuitive expectation that better legal institutions shrink the premium: stronger rule of law is associated with more negative reported size slopes, hence larger conventional size premia. The trust association is conditional and less precisely estimated: where rule of law is weak, higher generalized trust is linked to less negative reported slopes (and thus a weaker premium), and this link fades as rule of law strengthens. Formal and informal institutions thus help organize part of the disagreement in this literature, although the analysis concerns variation in reported estimates and does not identify causal effects.
- Open Access
- 23 April 2026
- Published by Springer Nature in Nature
- Vol. 653 (8114) , E7
Publisher Correction: Reproducibility and robustness of economics and political science research
- 23 April 2026
- Published by Springer Nature in Nature
- Vol. 653 (8114) , E7
Open AccessPublisher Correction: Reproducibility and robustness of economics and political science research
- Top Cited Papers
- 1 April 2026
- Published by Springer Nature in Nature
- Vol. 652 (8108) , 151-156
Reproducibility and robustness of economics and political science research
Science aspires to be cumulative. Reproducibility efforts strengthen science by testing the reliability of published findings, promoting self-correction, and informing policy-making1. Computational reproductions, whereby independent researchers reproduce the results of published studies, are an essential diagnostic tool2,3,4,5,6,7,8,9,10. Such efforts should have greater visibility11,12,13,14,15,16. However, little social science reproduction and robustness has been conducted at scale10,13,17,18,19,20,21,22,23. Here we reproduced original analyses and conducted robustness checks of 110 articles that were published in leading economics and political science journals with mandatory data and code sharing policies17,18. We found that more than 85% of published claims were computationally reproducible. In robustness checks, our reanalyses showed that 72% of statistically significant estimates remain significant and in the same direction, and the median reproduced effect size is nearly the same as the originally published effect size (that is, 99% of the published effect size). Additionally, 6 independent research teams examined 12 pre-specified hypotheses about determinants of robustness. Research teams with more experience found lower levels of robustness, and robustness did not correlate with author characteristics or data availability.
- 1 April 2026
- Published by Springer Nature in Nature
- Vol. 652 (8108) , 151-156
Top Cited PapersReproducibility and robustness of economics and political science research
Science aspires to be cumulative. Reproducibility efforts strengthen science by testing the reliability of published findings, promoting self-correction, and informing policy-making1. Computational reproductions, whereby independent researchers reproduce the results of published studies, are an essential diagnostic tool2,3,4,5,6,7,8,9,10. Such efforts should have greater visibility11,12,13,14,15,16. However, little social science reproduction and robustness has been conducted at scale10,13,17,18,19,20,21,22,23. Here we reproduced original analyses and conducted robustness checks of 110 articles that were published in leading economics and political science journals with mandatory data and code sharing policies17,18. We found that more than 85% of published claims were computationally reproducible. In robustness checks, our reanalyses showed that 72% of statistically significant estimates remain significant and in the same direction, and the median reproduced effect size is nearly the same as the originally published effect size (that is, 99% of the published effect size). Additionally, 6 independent research teams examined 12 pre-specified hypotheses about determinants of robustness. Research teams with more experience found lower levels of robustness, and robustness did not correlate with author characteristics or data availability.
- Open Access
- 26 September 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-12
Spurious precision in meta-analysis of observational research
Meta-analysis assigns more weight to studies with smaller standard errors to maximize the precision of the overall estimate. In observational settings, however, standard errors are shaped by methodological decisions. These decisions can interact with publication bias and p-hacking, potentially leading to spuriously precise results reported by primary studies. Here we show that such spurious precision undermines standard meta-analytic techniques, including inverse-variance weighting and bias corrections based on the funnel plot. Through simulations and large-scale empirical applications, we find that selection models do not resolve the issue. In some cases, a simple unweighted mean of reported estimates outperforms widely used correction methods. We introduce MAIVE (Meta-Analysis Instrumental Variable Estimator), an approach that reduces bias by using sample size as an instrument for reported precision. MAIVE offers a simple and robust solution for improving the reliability of meta-analyses in the presence of spurious precision.
- 26 September 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-12
Open AccessSpurious precision in meta-analysis of observational research
Meta-analysis assigns more weight to studies with smaller standard errors to maximize the precision of the overall estimate. In observational settings, however, standard errors are shaped by methodological decisions. These decisions can interact with publication bias and p-hacking, potentially leading to spuriously precise results reported by primary studies. Here we show that such spurious precision undermines standard meta-analytic techniques, including inverse-variance weighting and bias corrections based on the funnel plot. Through simulations and large-scale empirical applications, we find that selection models do not resolve the issue. In some cases, a simple unweighted mean of reported estimates outperforms widely used correction methods. We introduce MAIVE (Meta-Analysis Instrumental Variable Estimator), an approach that reduces bias by using sample size as an instrument for reported precision. MAIVE offers a simple and robust solution for improving the reliability of meta-analyses in the presence of spurious precision.
- Open Access
- 3 March 2025
- Published by Center for Open Science in MetaArXiv
Spurious Precision in Meta-Analysis of Observational Research
Meta-analysis upweights studies reporting lower standard errors and hence more precision. But in observational settings common to much research in social sciences, precision is not given to the researcher. Precision must be estimated, and thus can be p-hacked to achieve statistical significance. Simulations and applications show that spurious precision can invalidate inverse-variance weighting and bias-correction methods based on the funnel plot. Selection models fail to solve the problem, and common cures to publication bias can become worse than the disease. We introduce a novel approach that addresses spuriousness: meta-analysis instrumental variable estimator (MAIVE).
- 3 March 2025
- Published by Center for Open Science in MetaArXiv
Open AccessSpurious Precision in Meta-Analysis of Observational Research
Meta-analysis upweights studies reporting lower standard errors and hence more precision. But in observational settings common to much research in social sciences, precision is not given to the researcher. Precision must be estimated, and thus can be p-hacked to achieve statistical significance. Simulations and applications show that spurious precision can invalidate inverse-variance weighting and bias-correction methods based on the funnel plot. Selection models fail to solve the problem, and common cures to publication bias can become worse than the disease. We introduce a novel approach that addresses spuriousness: meta-analysis instrumental variable estimator (MAIVE).
- Top Cited PapersOpen Access
- 20 February 2025
- Published by Wiley in Journal of Economic Surveys
- Vol. 39 (5) , 2315-2333
Relative Risk Aversion: A Meta‐Analysis
Estimates of relative risk aversion vary widely, but no study has attempted to quantitatively trace the sources of the variation. We collect 1021 estimates from 92 studies that use the consumption Euler equation to measure relative risk aversion and that disentangle it from intertemporal substitution. We show that calibrations of risk aversion are systematically larger than estimates thereof. Moreover, reported estimates are systematically larger than the underlying risk aversion because of publication bias. After correction for the bias, the literature suggests a mean risk aversion of 1 in economics and 2–7 in finance contexts. The reported estimates are driven by the characteristics of data (frequency, dimension, country, stockholding) and utility (functional form, treatment of durables). To obtain these results, we use recently developed techniques to correct for publication bias and Bayesian model averaging techniques to account for model uncertainty.
- 20 February 2025
- Published by Wiley in Journal of Economic Surveys
- Vol. 39 (5) , 2315-2333
Top Cited PapersOpen AccessRelative Risk Aversion: A Meta‐Analysis
Estimates of relative risk aversion vary widely, but no study has attempted to quantitatively trace the sources of the variation. We collect 1021 estimates from 92 studies that use the consumption Euler equation to measure relative risk aversion and that disentangle it from intertemporal substitution. We show that calibrations of risk aversion are systematically larger than estimates thereof. Moreover, reported estimates are systematically larger than the underlying risk aversion because of publication bias. After correction for the bias, the literature suggests a mean risk aversion of 1 in economics and 2–7 in finance contexts. The reported estimates are driven by the characteristics of data (frequency, dimension, country, stockholding) and utility (functional form, treatment of durables). To obtain these results, we use recently developed techniques to correct for publication bias and Bayesian model averaging techniques to account for model uncertainty.
- Open Access
- 9 April 2024
- Published by Center for Open Science in MetaArXiv
Where Have All the Alphas Gone? A Meta-Analysis of Hedge Fund Performance
We examine the factors influencing published estimates of hedge fund performance. Using a sample of 1,019 intercept terms from regressions of hedge fund returns on risk factors (the alphas) collected from 74 studies, we document a strong downward trend in the reported alphas. The trend persists even after controlling for heterogeneity in hedge fund characteristics and research design choices in the underlying studies. Estimates of current performance implied by best practice methodology are close to zero across all common hedge fund strategies. Additionally, our data allow us to estimate the mean management and performance fees charged by hedge funds. We also document how reported performance estimates vary with hedge fund and study characteristics. Overall, our findings indicate that, while hedge funds historically generated positive value for investors, their ability to do so has diminished substantially.
- 9 April 2024
- Published by Center for Open Science in MetaArXiv
Open AccessWhere Have All the Alphas Gone? A Meta-Analysis of Hedge Fund Performance
We examine the factors influencing published estimates of hedge fund performance. Using a sample of 1,019 intercept terms from regressions of hedge fund returns on risk factors (the alphas) collected from 74 studies, we document a strong downward trend in the reported alphas. The trend persists even after controlling for heterogeneity in hedge fund characteristics and research design choices in the underlying studies. Estimates of current performance implied by best practice methodology are close to zero across all common hedge fund strategies. Additionally, our data allow us to estimate the mean management and performance fees charged by hedge funds. We also document how reported performance estimates vary with hedge fund and study characteristics. Overall, our findings indicate that, while hedge funds historically generated positive value for investors, their ability to do so has diminished substantially.
- Open Access
- 8 April 2024
- Published by Center for Open Science in MetaArXiv
Beauty and Professional Success: A Meta-Analysis
Common wisdom suggests that beauty helps in the labor market. We show that two factors combine to explain away the mean beauty premium reported in the literature. First, correcting for publication bias reduces the premium by at least a third. Second, controlling for cognitive ability negates the premium for all occupations except sex workers, a point further underscored by the similarity of the beauty effect on earnings and productivity. The second factor implies a positive link, perhaps genetic, between beauty and intelligence. We find little evidence of substantial attenuation bias that could offset publication and omitted-variable biases. The empirical literature is inconsistent with discrimination based solely on tastes for beauty. To obtain these results we collect 1,159 estimates of the effect of beauty on earnings or productivity reported in 67 studies and codify 33 aspects that reflect estimation context, including the potential intensity of attenuation bias. We employ recently developed techniques to account for publication bias and model uncertainty.
- 8 April 2024
- Published by Center for Open Science in MetaArXiv
Open AccessBeauty and Professional Success: A Meta-Analysis
Common wisdom suggests that beauty helps in the labor market. We show that two factors combine to explain away the mean beauty premium reported in the literature. First, correcting for publication bias reduces the premium by at least a third. Second, controlling for cognitive ability negates the premium for all occupations except sex workers, a point further underscored by the similarity of the beauty effect on earnings and productivity. The second factor implies a positive link, perhaps genetic, between beauty and intelligence. We find little evidence of substantial attenuation bias that could offset publication and omitted-variable biases. The empirical literature is inconsistent with discrimination based solely on tastes for beauty. To obtain these results we collect 1,159 estimates of the effect of beauty on earnings or productivity reported in 67 studies and codify 33 aspects that reflect estimation context, including the potential intensity of attenuation bias. We employ recently developed techniques to account for publication bias and model uncertainty.
- Open Access
- 3 January 2024
- Published by Center for Open Science in MetaArXiv
Conventional Wisdom, Meta-Analysis, and Research Revision in Economics
Over the past several decades, meta-analysis has emerged as a widely accepted tool to understand economics research. Meta-analyses often challenge the established conventional wisdom of their respective fields. We systematically review a wide range of influential meta-analyses in economics and compare them to ‘conventional wisdom.’ After correcting for observable biases, the empirical economic effects are typically much closer to zero and sometimes switch signs. Typically, the relative reduction in effect sizes is 45-60%.
- 3 January 2024
- Published by Center for Open Science in MetaArXiv
Open AccessConventional Wisdom, Meta-Analysis, and Research Revision in Economics
Over the past several decades, meta-analysis has emerged as a widely accepted tool to understand economics research. Meta-analyses often challenge the established conventional wisdom of their respective fields. We systematically review a wide range of influential meta-analyses in economics and compare them to ‘conventional wisdom.’ After correcting for observable biases, the empirical economic effects are typically much closer to zero and sometimes switch signs. Typically, the relative reduction in effect sizes is 45-60%.