Risk and Evidence of Bias in Randomized Controlled Trials in Economics

Abstract

The randomized controlled trial (RCT) has been a heavily utilized research tool in medicine for over 60 years. Since the early 2000's, large-scale RCTs have been used in increasingly large numbers in the social sciences to evaluate questions of both policy and theory. The early economics literature on RCTs invokes the medical literature, but seems to ignore a large body of this literature which studies the past mistakes of medical trialists and links poor trial design, conduct and reporting to exaggerated estimates of treatment effects. Using a few consensus documents on these issues from the medical literature, we design a tool to evaluate adequacy of reporting and risk of bias in RCT reports. We then use this tool to evaluate 54 reports of RCTs published in a set of 52 major economics journals between 2001 and 2011 alongside a sample of reports of 54 RCTs published in medical journals over the same time period. We find that economics RCTs fall far short of the recommendations for reporting and conduct put forth in the medical literature, while medical trials stick fairly close to them, suggesting risk of exaggerated treatment effects in the economics literature.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 27 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.