Replication and Scientific Standards in Economics a Decade Later: The Impact of the JMCB Project

Abstract

Scientific inquiry embodies skepticism. Researchers are trained to scrutinize every result, doubting not only the truth but also the tests of every hypothesis. Research papers in professional journals typically present only summaries of results, however, providing neither the programs nor data that a reader requires to fully understand-and question-the authors' tests. The Journal ofMoney, Credit, and Banking project a decade ago was the first attempt by the editor of a major journal to furnish readers with the data and programs used by the journal's authors. The project revealed the futility of proposing that readers obtain data and programs directly from authors, since data often were lost during the long interval between completion of the research and appearance of the published article. The project also established that professional journals were a low cost mechanism for collecting data from authors and distributing it to readers. A decade later, although the JMCB no longer requests data from authors, 2 journals have recently begun collecting such data and distributing it via the Internet.