XBRL (eXtensible Business Reporting Language) is an application of XML (eXtensible Markup Language) intended for use in digital business reporting. Observers predict XBRL will provide benefits to firms that adopt it, such as enhancing information use, facilitating comparability and consistency, and providing technological capability for near‐continuous financial reporting through the Web. However, questions arise regarding how well the proposed taxonomy for financial statements corresponds to firms' preferred reporting practices. We argue that a poor fit may lead to information loss and to subsequent resistance to use or general adoption of the taxonomy. A lack of fit could therefore negate anticipated firm or information‐user benefits.
To address this issue, we assess how well the year 2000 version, for financial reporting by commercial and industrial (C&I) firms under U.S. GAAP, accommodates current financial‐reporting practices of public companies. We attempt to match each line item in the 1999 annual financial statements of 67 companies with an XBRL taxonomy “tag,” employing two measures of the frequency of “special attention” (difficult‐to‐match) items as indicators of the quality of “fit” between the taxonomy and firms' reporting practices. Analyses show a good fit on average, but also indicate significant differences in the number and proportion of exceptions across financial statements and industries. In light of these results, we suggest modifications to the taxonomy and discuss the need for industry‐specific taxonomies.