November 8, 2010

Does Haste Make Waste in Regulatory Analysis?


Most federal agencies must conduct economic analysis when proposing major regulations. This paper uses a new dataset scoring the quality of analysis that accompanied proposed regulations in 2008 to assess whether some types of regulations receive more thorough analysis than others. Previous scholarship speculates that midnight regulations receive less thorough consideration, and the Office of Management and Budget asserts that agencies rarely estimate benefits and costs of transfer regulations. We test these hypotheses and find that the evidence supports both. We find that two classes of regulations have significantly lower-quality analysis: “midnight” regulations proposed after June 1, 2008, and budget or “transfer” regulations that describe how federal agencies will spend or collect money.

1. Introduction

Not all regulations are created equal. Since 1974, a series of executive orders have required executive branch regulatory agencies to produce some form of economic analysis when promulgating significant regulations. However, the quality of regulatory analysis varies considerably across agencies (Ellig and McLaughlin 2010, Hahn and Dudley 2007, Hahn et. al. 2000). Quality may vary systematically across time as well, possibly due to incentives created by election cycles. Numerous scholars note that regulatory activity spikes during an outgoing president’s lame-duck period; national news media dubbed these regulations “midnight regulations” (Cochran 2001, Davies and de Rugy 2008, Howell and Mayer 2005). One documented criticism of midnight regulations is that they may suffer in quality because they are rushed both during their creation at the promulgating agency and throughout the review process (Brito and De Rugy 2009, Brito and McLaughlin 2008, Morriss et. al. 2003). McLaughlin (2010) finds that midnight regulations receive shorter reviews at the Office of Information and Regulatory Affairs (OIRA). This result raises the question of whether midnight regulations receive lower-quality analysis than regulations that are reviewed for longer periods.

Using a new data set that measures the quality of agencies’ regulatory analyses (Ellig and McLaughlin 2010), this paper finds that midnight regulations tend to receive lower-quality analysis than other regulations. We also show that one particular type of regulation—budget or “transfer” regulations that define how an agency will spend or collect money—tends to be reviewed for shorter periods and receive lower-quality analysis. Holding other things constant, a midnight regulation would score about 19 percent below the mean score for all 2008 regulations, and a transfer regulation would score more than 60 percent below the mean. Review time at OIRA is a significant determinant of quality only when we do not control for whether regulations are transfer regulations.

Our findings point out a critical relationship between political pressure, legal deadlines, and the quality of regulatory analysis. While midnight regulations have often been criticized in the news media and pulled back for review by the incoming administration, the finding that midnight regulations tend to receive lower quality analysis than other regulations makes the criticism more than mere political showmanship. Our results provide an additional reason an incoming administration would want to give midnight regulations a second look. Even if the new administration agrees with a predecessor’s midnight regulation, it may feel additional scrutiny is warranted because midnight regulations are not thought out as carefully as “daylight” regulations. Our findings also help explain why an outgoing administration would seek to limit its own midnight regulations—as the G.W. Bush administration explicitly did in 2008 (Dudley 2009). Minimizing midnight regulation reduces the likelihood that a new administration could reject its predecessors’ regulations because of shoddy analysis. Instead, the incoming administration would need to articulate the policy differences that led it to make a different decision. By limiting midnight regulation, an outgoing administration forces its successor to be more transparent about its reasons for altering regulatory policies.

The reduction in analytical quality associated with transfer regulations is even bigger than the reduction in quality associated with midnight regulations. Transfer regulations are apparently treated differently by OIRA. Several former OIRA officials have told us that most OMB review of transfer regulations is conducted by budget analysts, whose main concern is ensuring that agencies correctly estimated the effects on the federal budget, rather than focusing on the economic analysis. OMB (2010, 18) observes that although transfer regulations generate social costs via mandates, prohibitions, and price distortions, agencies do not usually estimate the social benefits and costs of transfer regulations. For whatever reason, agencies apparently feel free to propose transfer regulations with much lower-quality economic analysis.

The quality of regulatory analysis is quite likely correlated with the quality of regulations themselves. For example, one important component of regulatory analysis is the consideration of alternative regulatory approaches to achieve the outcomes desired. Many low-quality analyses fail to consider even a single alternative approach. If other alternatives were not even considered, how can an agency be confident that its regulatory approach represents the best one, however “best” may be defined? Similarly, many of the low-quality analyses did a poor job of articulating the systemic problem the regulation is supposed to solve and presenting evidence that the problem is indeed significant. It is difficult to believe that the regulation will effectively solve the problem when the agency cannot even articulate the problem or prove that it exists (Ellig and McLaughlin 2010). Thus, this research documents what numerous other scholars have suspected: certain factors related to election cycles (such as midnight periods) and political processes (such as judicial and statutory deadlines) may diminish the effectiveness or efficiency of federal regulations.

A revised version of this article was published in the Administrative Law Review’s special symposium issue on the 30th anniversary of the Office of Information and Regulatory Affairs.  The full citation is: “Patrick McLaughlin and Jerry Ellig, “Does OIRA Review Improve the Quality of Regulatory Analysis? Evidence from the Final Year of the Bush II Administration,” Administrative Law Review 63, Special Issue (2011), pp. 102-23.