Study of medical publications suggests that most published medical research findings are false. Statistical studies in economics (econometrics) typically involve much worse possibilities for study design, data generation, and statistical controls than does medical research. In addition, medical research typically concerns controlled medical interventions on particular persons, while econometric research typically concerns interactions among many persons. Thus, compared to medical research findings, econometric research findings are less likely to be true and less likely to be significant to feasible action.
Attempts to replicate published econometrics results have encountered major obstacles. Even with the Internet’s cheap and powerful means of communication, few researchers make publicly available the data and estimation procedures used for a publication. Even with publication data and code, replication has been unsuccessful for a large share of publications.
Lack of interest in replication covers the inability to replicate most published econometrics findings. As Larry Summers has observed:
In the natural sciences, investigators rush to check out the validity of claims made by rival laboratories and then to build on them. Such efforts are very rare in economics. I find it implausible that this is a consequence of the accuracy and robustness of the conclusions of econometric studies. Instead, the absence of replication attempts for most economic work are a consequence of the fact that the results are rarely an important input to theory creation or the evolution of professional opinion more generally.^
Donohue and Levitt’s study on abortion and crime has attracted considerable empirical scrutiny and replication efforts. But that work isn’t important to theory creation or economists’ professional opinions; rather, that work plays into bitter, enduring public disputes. For most econometrics studies, no one is interested in replicating them, attempts to replicate them would not be easy, and if replication were done, most of the findings would turn out to be false.
The significance of published econometrics results are largely as currency in disciplinary prestige and advancement. They have more importance as citations than as knowledge.