Rules-based standards and earnings management
Richard Mergenthaler has a paper on SSRN titled “Principles-Based versus Rules-Based Standards and Earnings Management.”
When I read the title I thought for sure the article would argue that principles-based standards lead to more earnings management because there was more flexibility. To my surprise, the paper presents arguments that rules-based standards lead to more earnings management. To quote the author,
I find a positive relation between rules-based characteristics and the dollar magnitude of earnings management. This result suggests the magnitude of earnings management is greater when executives violate rules-based standards. One potential explanation for this result is that executives are less likely to be penalized when they violate a rules-based standard. Hence, executives who weigh the costs and benefits of managing earnings will manage earnings more in areas where they are less likely to be penalized. Consistent with this notion, the evidence provided suggests detected rules-based violations are less likely to be penalized by the SEC. Finally, there is no evidence of a statistical relation between rules-based characteristics and both the dollar magnitude of SEC fines and the number of SEC sanctions.
These findings are consistent with the notion that a switch to a less rules-based standard will not expand the tendency of firms to manage earnings.
Another thing I found interesting about the paper is the instruments he uses to measure how rules-based a standard is. He uses two measures, both of which are based on four main characteristics although they vary in how they measure the variables and perform the statistics.
1. Bright-line Thresholds–The total number of bright line thresholds.
2. Scope and Legacy Exceptions–The total number of scope and legacy exceptions.
3. Implementation Guidance–The total number of “interpretive pronouncement”
documents.
4. Detail–The total number of words in each standard.
More detail about the instruments he uses can be found in section three of his paper. These instruments could prove useful for other researchers struggling to find ways to measure how rules-based a standard is.
