A recent study by Public Citizen reports that it takes about 2.5 years to issue an economically significant rule, starting from the time the rule is first listed in the regulatory agenda. There are major differences between agencies – an economically significant rule takes EPA almost four years, rather than the 2.5 years needed by the average agency.
But if the rule is accompanied by an advance notice of proposed rulemaking (ANPR), the average time is over four years. There’s no way of knowing whether the ANPR process itself slows things down, or whether the issuance of an ANPR is just a signal that a rule is unusually complicated or contentious. Economically significant rulemakings that included both an ANPR and a Regulatory Flexibility Analysis took almost five years.
The causes of the delays are also unclear. We can divide the possible causes into three categories:
- Technical difficulty. The delays could be caused simply by the difficulty of the issues — the need to collect data, conduct the necessary analyses, and make hard technical judgments.
- Procedural issues. Some administrative law scholars have long complained about ossification of the rulemaking process due to the need of an agency to satisfy judicial demands forever more detailed analysis and explanation, which also prompt ever-expanding efforts to compile an exhaustive regulatory record. Others point to the added time needed for OIRA review – OIRA being the White House office that oversees cost-benefit analysis by agencies.
- Political issues. Delays may be caused by pressure from Congress, White House fears about the political impact of rules, and the concomitant need to negotiate with stakeholders in order to limit political fall-out. Some of those stakeholders may be other government agencies. For instance, the Pentagon is responsible for some of the most serious hazardous waste sites and naturally tends to resist anything that would make clean up more expensive.
So far as I know, we really have no hard evidence about the relative weight of these causes. Most of these factors operate behind closed doors of one kind or another. Even if they took place in the open, collecting all the data from across the government would be a daunting effort.
I was curious about one issue, which is the amount of additional delay introduced by OIRA. OIRA review could cause delays in three ways: (1) because of behind-the-scenes involvement even before the formal rulemaking begins; (2) because the agency has to defensively prepare itself to deal with OIRA, slowing the process; or (3) because of the formal OIRA review. We only have data about the third type of delay.
I didn’t attempt a systematic analysis of how much OIRA’s formal review added to delay, but I did take a look at some of the data, combining the data in the Public Citizen report with a study of formal OIRA review. I looked at average times from 2002, 2004, 2006, 2010, and 2012, subdivided by economically significant rules versus those characterized as significant for other reasons by OIRA. For economically significant rules, OIRA review accounted for 3%-9% of the total rulemaking time, with an average of 5.8%. The highest percentages were in 2002 (9%) and 2012 (8%), which might have had something to do with the election cycle. For the “otherwise significant” rules, OIRA review accounted for a steady 6.0-6.5% of the total, except in 2012, when it rose to 10% of the total delay, which again might relate to the election cycle. Thus, formal OIRA review accounted for only a fraction of the total length of rulemaking, though not a negligible fraction.
These are all averages, of course. Some rulemakings, and some OIRA reviews, took much longer than average, and some agencies (OSHA in particular) seem to have especially long delays.
I would draw two general conclusions. First, since we don’t really know the relative weight of the various causes of delay, we need to chip away at all of them wherever we can streamline the process without significant loss of quality. The public is harmed when beneficial regulations are unduly delayed. Second, because there are already many procedural requirements that apply to rulemaking and rulemaking times are long, there should be a heavy burden of proof on advocates of adding even more procedural obstacles.