Editor’s Note:  If the author of the article below had good cause to think that SBA/Advocacy’s landmark study on the costs of federal regulation was lacking in accuracy, objectivity, utility and/or integrity, he could have used the Data Quality Act’s Request for Correction process to seek and obtain any justified changes.  Since the author has declined to use the Congressionally-enacted mechanism for seeking correction of quality-deficient federal information disseminations, readers can only conclude that Mr. Mandelbaum’s objections are to the study’s conclusions, not with how the study was performed.

Questions on a Study of the Cost of Federal Regulation

By ROBB MANDELBAUM

In late March, Representative Tammy Duckworth, a Democrat from Illinois, announced a new bill promising regulatory relief to small businesses. But what caught The Agenda’s eye was Ms. Duckworth’s justification for the legislation. “For businesses with less than 20 employees,” she said in the press release, “the annual cost of federal regulation can be over $10,000 per worker.”

To be precise, it was $10,585 per employee in 2008, according to a 2010 study sponsored by the Small Business Administration’s Office of Advocacy. For large companies, the study says the cost is $7,755 for every employee. And the total cost of federal regulations to the economy was $1.75 trillion in 2008, an impressive figure — if it is true. Certainly it has been accepted as currency by many business and free market advocates; it has been cited by the U.S. Chamber of Commerce, the National Federation of Independent Business, and the Competitive Enterprise Institute. Mitt Romney embraced the notion of a $1.75 trillion price tag in his presidential campaign. And, as Ms. Duckworth shows, the outrage the claim inspires is not strictly partisan.

But the Advocacy study has also come under scrutiny from economists and legal scholars, who have challenged the methodology, the underlying data and the authors’ reluctance to share certain details about how they did their work. And while much of the criticism comes from liberal organizations like the Center for Progressive Reform and the Economic Policy Institute (which weighed in twice), here, too, the outrage is not strictly partisan: the Congressional Research Service has also cast doubt on the study’s findings.

Written by W. Mark Crain and Nicole Crain, economists at Lafayette College in Easton, Pa., the study tried to calculate all of the costs of all federal regulations on the books in 2008, regardless of when they were first adopted. To do this, the Crains (or “Crain and Crain,” as they are known in the academic world) tallied the costs of four categories of regulations and then added them together.

The largest of these categories is economic regulations, which they said carried costs totaling $1.2 trillion in 2008. It also is the source of most of the controversy surrounding the report. The Crains defined economic regulations broadly, to include issues as diverse as trade rules, antitrust policies, consumer product safety obligations and some labor rules. (Labor safety, though, is a separate category.) And because it would be extremely difficult to identify, let alone calculate, costs associated with each economic regulation, the Crains didn’t try.

Instead, they used a mathematical formula to try to predict how increases and decreases in regulation would affect a country’s economic output, an exercise known as regression analysis. And to measure the amount of regulation, the Crains turned to a global index of “regulatory quality” devised by researchers at the World Bank and the Brookings Institution — and here is where the controversy begins. According to the Crains, countries that score higher on the index have less regulation. But the index is not a systematic accounting of regulations in each country. Instead, according to the index’s creators, it captures “perceptions” of how well governments make “sound policies and regulations” that foster economic growth. It does this by surveying experts and relying on polls of business owners and individuals.

But are sound policies the equivalent of less regulation? Not according to one of the index’s creators. Evaluating a regulatory environment “is very different from simply measuring whether it is ‘stringent’ or not,” Aart Kraay, a World Bank economist, wrote in an e-mail to the Crains that was later provided to the Congressional Research Service. For example, the Netherlands and Scandinavia had higher scores on the regulatory quality index than the United States in 2008, and, as the Economic Policy Institute noted dryly, these are “countries not typically associated with a lack of regulation.” It is also possible that the surveys that make up the index absorb sentiment about regulations that the Crains cover elsewhere — like environmental rules, for example — which would effectively double-count the cost of those rules. (Despite repeated requests, the Crains did not agree to be interviewed for this post.)

Even if a consensus existed on precisely what the regulatory quality index measured, converting a soft concept into a hard number raises other concerns. The formula the Crains used tried to account for other factors, or variables, that could affect economic growth, such as education. However, the economists initially would not share the precise data they used with other researchers. When the Congressional Research Service tried to replicate the Crains’ findings using similar, but not identical, supporting data and methods, it found that “the regulatory quality index had no discernible independent effect on G.D.P. per capita, suggesting that the analysis is highly sensitive to the choice of control variables and measures.”

The Economic Policy Institute performed a similar experiment and got similar results. “It’s basically a case of garbage in, garbage out,” said John Irons, who analyzed the study for the Economic Policy Institute. “That $1.2 trillion number is not a reliable number in any sense at all.” But Mr. Irons acknowledged that this did not necessarily mean it was too high, only that it was wrong.

The Crains’ treatment of economic regulations accounts for about 70 percent of the total regulatory cost for 2008, so that’s where critics focused most of their energy. Still, they also found much to dispute in how the researchers treated other regulations. For example, to calculate the cost of environmental, labor safety and homeland security regulations, the authors relied heavily on the Office of Management and Budget, which performs cost-benefit estimates for many regulations.

The government figures are presented as a range, but the Crains told the Congressional Research Service that they used only the high-end figure because the government does not track every rule and because the researchers believe “government agencies tend to be conservative in estimating regulatory costs.” But government agencies generally rely on information provided by the regulated industry to estimate costs, and, as the Center for Progress Reform and the O.M.B. itself point out, there is good reason to believe that industries facing regulation tend to overstate those costs.

Nor does the report consider the benefits of regulations, because the Office of Advocacy did not ask the Crains to measure them. The O.M.B. is required by law to tally benefits as well as costs as best it can, and it has concluded that, in general, the benefits of regulations exceed their costs to society. Advocacy is required by law only to “measure the direct costs and other effects of government regulation on small businesses.” Even the Crains have acknowledged that this is a skewed foundation on which to base policy.

Before the study was made public, it was reviewed by a pair of economists. Richard Williams, a conservative scholar, raised several questions about the suitability of relying on the Regulatory Quality Index. But later, he cited the study as authoritative, even though he acknowledged that he did not know whether his concerns had been addressed in the final version. In a recent interview, he said  he understood the Crains’ intellectual impulse. “I think they’re right — you want to know how regulations affect G.D.P.,” he said. “The theoretical concept was good, but at the time there just weren’t good databases.” Asked why he cited the study and the $1.75 trillion figure if he had doubts, he responded, “We only have two estimates of the cost of regulations. One is the O.M.B.’s, which is clearly an underestimate. On the other hand, we have this big number, which clearly has problems, but it is a number.”

The other reviewer was Robert Litan, a former Clinton administration official. His comments at the time, in their entirety, were, “I looked it over and it’s terrific, nothing to add. Congrats.” He recently declined to comment further about the study.

The S.B.A.’s Office of Advocacy also declined further comment. It still defends the paper online, where it says “the study demonstrated that small businesses bear a larger burden from regulations than large businesses,” but “as with almost any academic methodology, it was not intended to be considered a precise finding.” One measure of how independent the Office of Advocacy is from the rest of the S.B.A. and the rest of the Obama administration is that administration officials have publicly, and harshly, criticized the findings. Austan Goolsbee, then the president’s chief economic adviser, called them “utterly erroneous.”

The office of Ms. Duckworth, the Illinois Democrat, did not have much to say. The bill she introduced would allow a small business to escape a fine for “a first-time error in filling out federal paperwork.” Her spokesman, Anton Becker, was happy to talk about that. About the $10,500 figure, he would say only, “The figure cited in the press release is an example of one study’s findings from the Small Business Administration.”