EPA has conducted several ex post assessments of regulatory compliance costs, with the ultimate goal of identifying ways to improve ex ante cost estimation. The work to date has culminated in four case studies that examine five regulations using a common conceptual framework. The standardized framework provides a systematic way to investigate key drivers of compliance costs to see if judgments can be made about why and how ex ante and ex post estimates of costs differ. In addition to describing this conceptual framework, we describe the criteria used to select the rules to be analyzed, summarize the main hypotheses for why ex ante and ex post cost estimates may differ and discuss some of the challenges encountered in conducting these ex post analyses.

This paper compares EPA’s ex ante cost analysis of the Cluster Rule, EPA’s first integrated, multi-media regulation, and MACT II Rule to an ex post cost assessment. The goal of this assessment is to determine if actual costs diverged from ex ante costs and, if so, what factors caused this divergence. We find the EPA ex ante costs overestimated the ex post capital costs for the Cluster Rule by 30 to 100%. Contributing factors appear to be use of cleaner technology, flexible compliance options, site-specific rules, shutdowns and consolidations. Ex ante estimates for the MACT II Rule are found to be overestimated by 25% for capital costs and 200 or more percent for annual costs. The primary reason for the overestimate is the use of the bubble compliance strategy that required fewer paper mills to install pollution abatement equipment than anticipated by EPA.

Methyl bromide (MBr) has been widely used as a fumigant to control pests in the agricultural sector, but it is also an ozone depleting substance. After 2005, methyl bromide could only be produced when a critical use exemption was agreed to by the signatories to the Montreal Protocol. This paper examines how the EPA’s ex ante cost analyses for open field fresh strawberries in California for the 2006–2010 seasons compare to an ex post assessment of costs. A key input into the ex ante cost analysis is the assumed yield loss associated with methyl bromide alternatives. The EPA used conservative assumptions given the wide range of estimates in the literature at the time, but it appears that a number of viable MBr alternatives – either new fumigants or new ways of applying existing fumigants – may have become available more quickly and resulted in lower yield loss than initially anticipated. Likewise, it appears that farmers who substituted away from methyl bromide did so without imposing large negative impacts on production in prime California strawberry growing areas. Ex post evaluation also confirms the effect of California regulatory restrictions in limiting the use of various economically competitive alternatives. It is worth noting that unanticipated complications after switching away from methyl bromide, such as new diseases, slowed the transition to MBr alternatives.

This paper compares EPA’s ex ante cost analysis of the 2001 maximum contaminant limit (MCL) for Arsenic in Drinking Water to an ex post assessment of the costs. Because comprehensive cost information for installed treatment technologies or other mitigation strategies pursued by water systems to meet the new standard is not available, this case study relies upon ex post cost data from EPA Demonstration Projects, capturing a total of 50 systems across the US. Information shared by several states and independent associations on the types (but not costs) of treatment technologies used by systems is also summarized. Comparisons of predicted costs to realized costs using our limited data yield mixed results. Plotting the capital cost data from the Demonstration Projects against the cost curves for the compliance technologies recommended for smaller systems, we find that the EPA methodology overestimated capital costs in most cases, especially as the size of the system increases (as measured by the design flow rate).

While the need to update EPA benefit-cost analysis to reflect the most recent science is broadly acknowledged, little work has been done examining how well ex ante BCAs estimate the actual benefits and costs of regulations. This paper adds to the existing literature on ex post cost analyses by examining EPA’s analysis of the 1998 Locomotive Emission Standards. Due to data limitations and minimal ability to construct a reasonable counterfactual for each component of the cost analysis, the assessment relies mainly on industry expert opinion, augmented with ex post information from publicly available data sources when possible. The paper finds that the total cost of bringing line-haul locomotives into compliance with the 1998 Locomotive Emission Standards rule remains uncertain. Even though the initial per-unit locomotive compliance costs were higher than predicted by EPA, total costs also depend on the number of locomotives affected by the regulation. Over 2000–2009, the number of newly built line-haul locomotives was higher but the number of remanufactured line-haul locomotives was lower than EPA’s estimate.

It has occasionally been asserted that regulators typically overestimate the costs of the regulations they impose. A number of arguments have been proposed for why this might be the case. The most widely credited is that regulators fail sufficiently to appreciate the effects of innovation in reducing regulatory compliance costs. Most existing studies have found that regulators are more likely to over- than to underestimate costs. While it is difficult to develop summary statistics to aggregate the results of different studies of disparate industries, one such measure is the average of the ratio of ex ante estimates of compliance costs to ex post estimates of the same costs. This ratio is generally greater than one. In this paper I argue that neither the greater frequency of overestimates nor the fact that the average ratio of ex ante to ex post cost estimates is greater than one necessarily demonstrates that ex ante estimates are biased. There are several reasons to suppose that the distribution of compliance costs could be skewed, so that the median of the distribution would lie below the mean. It is not surprising, then, that most estimates would prove to be too high. Moreover, Jensen’s inequality implies that the expected ratio of ex ante to ex post compliance costs would be greater than one. I propose a regression-based test of the bias of ex ante compliance cost estimates, and cannot reject the hypothesis that estimates are unbiased. Failure to reject a hypothesis with limited and noisy data should not, of course, be interpreted as a strong argument to accept the hypothesis. Rather, this paper argues for the generation of more and better information. Despite the existence of a number of papers reporting ex ante and ex post compliance cost estimates, it is surprisingly difficult to get a large sample with which to make such comparisons.