If you have an ACS member number, please enter it here so we can link this account to your membership. (optional)

ACS values your privacy. By submitting your information, you are gaining access to C&EN and subscribing to our weekly newsletter. We use the information you provide to make your reading experience better, and we will never sell your data to third party members.



Improving Accuracy

EPA studies why its cost estimates for pending regulations tend to be high

by Cheryl Hogue
May 7, 2012 | A version of this story appeared in Volume 90, Issue 19

Credit: Shutterstock
EPA is analyzing the pulp and paper industry’s costs to implement a 1998 air and water pollution rule and comparing those figures to the agency’s cost estimate for the regulation.
A paper mill, in Puget Sound, Wash.
Credit: Shutterstock
EPA is analyzing the pulp and paper industry’s costs to implement a 1998 air and water pollution rule and comparing those figures to the agency’s cost estimate for the regulation.

Before it can issue a major new regulation, the Environmental Protection Agency must, as required by law and presidential directives, estimate the costs of the rule. To do so, EPA economists crunch data primarily supplied by industry on the costs of compliance to come up with the estimates.

These appraisals of cost can influence how stringent the agency eventually makes a regulation. When combined with benefit calculations, these cost estimates are also key to informing the public how a regulation is expected to improve society. And these assessments become fodder for opponents of agency regulation, who argue that EPA’s rules cost businesses too much.

The problem is that EPA’s cost estimates hardly ever prove to be accurate. In fact, they are more likely to overstate the actual cost of a rule than to underestimate, published studies show.

Now, with the goal of making its cost estimates more predictive, EPA is trying to determine the reasons they are off so often, according to Al McGartland, director of the agency’s National Center for Environmental Economics. The center formulates the guidelines that the agency follows when analyzing the economic impacts of regulations, McGartland said at an April 19 meeting of the EPA Science Advisory Board’s Environmental Economics Advisory Committee (EEAC). That panel will counsel agency economists as they investigate why EPA’s cost estimates don’t jibe with actual costs.

EPA is conducting a detailed review of 10 major regulations, some dating to the 1990s, to determine the factors that led to inaccuracy of cost estimates prepared before rules were issued. The agency is looking to determine how much the preregulatory cost estimates differ from actual compliance costs, and if the difference is 25% or more, why, McGartland said. The work is being done under a January 2011 directive from President Barack Obama to eliminate or update obsolete regulations and to ensure that the benefits of federal rules justify their costs, he added.


EPA is studying the cost estimates of 10 regulations:

1997 Air pollution limits for nonroad diesel engines, such as those in construction equipment

1998 Air emission limits for locomotives

1998 Air and water standards for pulp and paper mills (cluster rule)

1998 Nitrogen oxide emissions limit for new fossil-fuel-fired power plants

2000 Water pollution limits for waste-handling facilities

2000 Water pollution limits for facilities that clean the interior of tanks hauled by trucks, trains, or ships

2001 Drinking water standard for arsenic

2001 Technology-based air pollution standards at pulp mills

2004 Air pollution standard for automakers to apply coatings to cars and light trucks

2006 Exemption allowing use of the pesticide methyl bromide on strawberries in California


An interim report that the agency prepared for EEAC lays out potential reasons that estimates and actual costs of regulations differ.

To begin with, both regulators and regulated entities may have incentives to overstate costs of rules, the report says. The data EPA uses to calculate the costs of an upcoming rule commonly come from estimates provided by individual companies or trade organizations, it says. These regulated entities may overstate the anticipated costs of compliance for several reasons, it adds.

“Industry may be attempting to thwart what it sees as onerous regulations by providing a signal that costs are prohibitive,” the report notes. Alternatively, the private sector might provide higher cost estimates because of uncertainties over how the details of regulation will eventually turn out, the report says.

Also, industry tends to calculate costs based on the use of existing technologies, such as pollution control equipment already on the market, rather than untested, innovative approaches to compliance. “Industry is likely to respond by describing a plausible way of complying rather than by evaluating all alternatives before identifying that which will minimize their compliance costs,” the report says.

Yet once a rule is finalized, companies often come up with novel, cost-saving approaches. Plenty of cases have occurred where technological innovation led to compliance costs well below EPA’s estimates, the report says. One example given is the phaseout of chlorofluorocarbons, the stratospheric-ozone-depleting synthetic compounds once widely used as refrigerants. The agency’s cost estimates did not account for process changes and the deployment of new blends of chemicals or substitute compounds that reduced or eliminated uses of CFCs. The actual costs of the CFC phaseout were 30% less than predicted, according to the report.

Economic analysts generally agree that EPA’s cost estimates are inaccurate because the agency doesn’t take into account technological innovations that ultimately make many rules less expensive than anticipated, said David R. Simpson, an economist in EPA’s Benefits Assessment & Methods Development Division.

Another possible explanation for inaccurate cost estimates, Simpson told EEAC, is the slow pace at which the federal government issues regulations. EPA prepares economic analyses of rules years before they take effect, he said. The agency’s interim report acknowledges this, stating: “Cost estimates based on early versions of a rule may no longer apply to the rule that eventually emerges.”

Meanwhile, regulators may have little incentive to sharpen cost estimates that lean to the high side if the benefits of a pending rule are already found to outweigh the costs, the report says. “If we’re going to implement the rule anyway, why bother to parse costs carefully?” Simpson asked.

In its research thus far, EPA has found that its analytical techniques for estimating the costs of rules weren’t necessarily faulty, McGartland said. In some cases, results turned out to be off because the economy changed in unanticipated ways, he told EEAC. For example, the agency relies on Energy Information Administration forecasts for future costs of fuels. But those projections have sometimes turned out to be wrong, either over- or underestimating future prices, he said.

The agency’s research into the accuracy of its cost estimates faces its share of difficulties. In some cases, it’s hard to tease out industry’s actual cost of compliance with a particular rule from other factors, McGartland said. For instance, a factory might install new pollution control equipment during a shutdown for repairs unrelated to environmental requirements, making it difficult to determine how much of the lost productivity should be attributed to an EPA rule. McGartland compared this to replacing a water pump on a car when changing the timing belt. In most cars, the water pump and timing belt are replaced at the same time because most of the cost of this work isn’t the parts—it’s the labor to access the area of the engine.

One of the regulations EPA is analyzing in this work is a set of standards and guidelines the agency issued in 1998 to control hazardous air pollution and toxic water discharges from pulp and paper mills. Collectively, these are called the cluster rule. Although EPA has run into some analytical difficulties, its preliminary findings are that the agency overestimated the capital costs of the cluster rule by 30 to 100%, Ronald J. Shadbegian, an economist in EPA’s Benefits Assessment & Methods Development Division, told EEAC.

EPA doesn’t have enough detailed data to determine the reasons for the inaccuracy, Shadbegian said. The agency did not get much feedback about actual costs from industry until recently, when the pulp and paper industry released an analysis of its own about the cluster rule, he said. Those data were not received in time to inform EPA’s preliminary findings.

But Jerry Schwartz, senior director for energy and environmental policy at the American Forest & Paper Association, told EEAC that the industry study of the cluster rule came to a conclusion different from EPA’s preliminary finding. The industry analysis determined that the cost estimates the agency made before the cluster rule was issued approximated the actual compliance costs, Schwartz said.

In any case, cost estimates play an important role in the government’s cost-benefit analysis of pending regulations and need to be as accurate as possible, Schwartz said, endorsing EPA’s effort to improve its analyses. This, he said, is especially important during difficult economic times and for highly regulated industries in competitive markets.

But determining how much a regulation costs industry is a difficult task, and data are essentially impossible to cross-check, points out Randy Rabinowitz, director of regulatory policy for OMB Watch, a group that tracks rule-making agencies and regulatory gatekeeping at the White House Office of Management & Budget. She tells C&EN that companies generally don’t have systems that can account for the actual cost of complying with regulations.


This article has been sent to the following recipient:

Chemistry matters. Join us to get the news you need.