Addressing the Replication Crisis in Accounting

Abstract

The prominence of the “replication crisis” in social sciences has spurred concern regarding the reliability of published research. Accounting researchers share such apprehensions, acknowledging the perceived scarcity of published accounting replication studies. To investigate this belief, we analyze annual replication rates and the number of replicated original articles from three prominent accounting journals: Contemporary Accounting Research (CAR), Journal of Accounting and Economics (JAE), and Review of Accounting Studies (RAST). Two sample periods, spanning from 1970-2015 and 2016-2021, reveal a more than doubled increase in both 1) the average annual replication rate, rising from 3.13 to 8.67 times, respectively and 2) the average number of replicated original articles per year, advancing from 5.33 to 12.77 original articles, respectively. These findings challenge conventional beliefs surrounding the replication crisis in accounting, revealing a growing volume of published accounting replication research and replicated original articles. The significant increase observed, even within this small sample of three journals, suggests that accounting replication is more prevalent than thought. As ongoing analysis extends to a broad selection of prominent and smaller journals, the distinguished pattern emerging within the current sample suggests a potential continuation of this trend across accounting research.