Avoiding specifics in this arena (and all other areas of policy arcana for that matter) is probably a wise political move for any presidential campaign, but a lack of specificity means that each new administration has no mandate to reform the regulatory apparatus. What’s more, such reform is difficult to do on the fly: after the busy task of issuing and approving regulations has already begun, an administration intent on imposing limits on the regulatory state invariably gets bogged down fighting the daily battles and does not engage in a new war over the issue. Without a reform plan that’s ready to go from the beginning, it becomes practically impossible to achieve meaningful changes in how we do regulatory policy in the United States.
And to be sure, we need to reform the system: U.S. regulations impose costs of at least $1 trillion on the U.S. economy and their effects on economic activity are inexorably growing. The power of the executive branch agencies to ignore the express will of Congress and pursue an independent regulatory agenda is a troubling situation, and we need to do more to impose additional checks on the system.
What should the next administration do to improve the issuance and oversight of regulations? We want to provide a few organizing principles upon which a new administration could begin to change the regulatory process.
REPEALING REGULATIONS IS OVERRATED
Many politicians like to frame regulatory reform as consisting of a systematic review of existing regulations to determine which cannot survive an honest cost-benefit review and then repeal those that won’t. Although there are myriad regulations currently in place that would not pass such a test, rolling back regulations that companies have already largely complied with will accomplish little.
The problem is one of sunk costs: once companies have spent money to reduce emissions or improve workplace safety, undoing those requirements won’t save them much money. In fact, a regulatory rollback likely would make those firms worse off by allowing new competitors to compete against them, operating with a lower cost structure because of the repealed regulations.
For example, during the presidential campaign, Donald Trump declared his intention to shelve the Obama administration’s Mercury and Air Toxics rule, one of the costliest regulations issued by the president’s Environmental Protection Agency. However, there are minimal savings to be had from its repeal at this stage: companies have already responded to it by dismantling a multitude of coal plants, and those will not return regardless of future regulations. What’s more, as long as natural gas remains plentiful, it will continue to be more cost-effective to burn gas rather than coal to produce electricity; even shuttered coal plants that could be turned on tomorrow would likely remain shuttered if the next president were to repeal the rule.
There are some exceptions to this. For instance, occupational licensing at the state level has expanded greatly over the last few decades, but many of those requirements do little to improve public safety but much to impede the entrance of potential competitors. The current Council of Economic Advisers has noted that these impediments are a significant cost to society in the form of higher prices for consumers and lower wages for workers not in the protected guilds. Repealing these rules would definitely benefit the economy, but would require preemptive federal deregulation or active state involvement.
While repealing regulations is usually pointless, it is still worth having agencies—or better yet, outside institutions—do retrospective cost-benefit analyses to determine how well the agencies did at predicting the costs and benefits of their rules. If nothing else, this would impose a modicum of restraint on the agencies going forward.
GIVE OIRA THE POLITICAL BACKING TO DO ITS JOB
In the 1990s, staffers in the White House Office of Information and Regulatory Affairs (OIRA) constructed a trophy of sorts consisting of a can of shaving cream attached to an x‑ray that suggested said can was unpleasantly inserted into an unfortunate patient. The trophy sat on the desk of the analyst who had most recently lost a battle with an agency because of political machinations.
The award did not disappear with the advent of the George W. Bush administration. While having ostensibly conservative secretaries running the cabinet agencies helped to temper some of the agencies’ more egregious activities, a handful of political appointees in an agency with thousands of career staffers cannot hope to arrest the actions of the entire bureaucracy. What’s more, even the most rock-ribbed Republican quickly realizes that political exigencies can make approving a dubious regulation the path of least resistance in most situations.
Early in the Bush administration, an OIRA analyst dared to push back on a regulation issued by the U.S. Department of Agriculture that would essentially force a group of farmers to tear out their apricot orchards—for which the government would generously compensate them—in an attempt to boost the price of apricots. The program was nowhere close to being a “major” program; the total government cost was less than $20 million, which meant it was not subject to a formal cost-benefit analysis. But it was a lousy program and the OIRA analyst tasked with reviewing it thought it would be a good test case to see whether the ostensibly free-market Republican White House would countenance any challenge to Big Agriculture.
The OIRA administrator, John Graham, was wary of picking a fight with the Agriculture Department early in his tenure, especially after Congress had just passed (and the president signed) a large increase in agricultural subsidies that amounted to a complete abandonment of earlier efforts to end price supports. Eventually, the fight was kicked to Office of Management and Budget director Mitch Daniels, who in turn kicked it to Vice President Dick Cheney. Cheney then met with the agriculture secretary, Ann Veneman, to negotiate a resolution—which amounted to letting the USDA go ahead with the program, but with a promise that OIRA would “win the next one.”
Agencies pass regulations that cost $20 million every day of the week without any political chieftain batting an eye. But attempting to stop a $20 million boondoggle required two cabinet officials and the vice president to sign off—in a Republican administration. That such a minor effort to roll back a government program failed—and triggered an intervention at the highest level of our government—gave OIRA analysts a lesson on how politics trumps good policy.
Solving this particular moral hazard is easier said than done. It’s folly to assume that a future administration—regardless of the party—might take a less politicized approach to the job. Even removing the political aspect to these decisions wouldn’t change the current system’s bias toward over-regulation, given that most of our bureaucrats have a natural inclination to do more—not less—regulating.
The only solution we offer is, lamentably, eternal vigilance on the part of regulatory economists, scholars, and others who understand how cost-benefit analysis is done and can easily recognize when a bureaucrat is putting his thumb on the scale to tilt the results in favor of a rule. We should encourage the next administration to give more weight to OIRA’s declarations and allow its deliberations to be given precedence over short-term political exigencies. This will always be difficult for an administration to countenance. If we do a better job of highlighting those incidents where agencies escape with regulations that don’t meet an objective cost-benefit analysis, we can help to accomplish this.
TAKE COST-BENEFIT ANALYSIS OUT OF THE AGENCIES
The basic problem with how the government does regulatory review is that it allows the defendant to prepare the plaintiff’s brief.
For major regulations—defined as those that will have an economic effect of greater than $100 million—Executive Order 12866 requires that an analysis be done to demonstrate that the benefits clearly exceed the costs of the regulation. These analyses are done by the agency proposing the regulation.
At one level this makes a modicum of sense. Since agency staffers (presumably) have intensively studied the regulation, they have the most information at their disposal to complete the analysis. However, the agency also has an inherent conflict, given that it has a huge incentive to ensure that its regulations get approved. The staff economists from the agency are fully aware that it’s best for their career to help the agency’s mission and they do their best to ensure their analysis supports a regulatory decision that’s already been made. It’s a fatal conflict of interest that generally renders any analysis by the regulating agency suspect.
The answer to this problem is to assign another entity the task of performing the cost-benefit analysis. We propose moving the regulatory economists who are now within the agencies to an entirely new agency, modeled after the Congressional Budget Office, to perform such analysis. (See “Toward a New and Improved Regulatory Apparatus,” Spring 2013.) This would not add to the level of government spending because the regulatory agencies would be absolved of the burden of doing cost-benefit analysis. Their budgets could be commensurately lowered and the savings transferred to the new office. The agencies would also be required to turn over all relevant data to the new office to allow it to do a proper accounting of the costs and benefits of proposed regulations.
MAKE AGENCIES RECOGNIZE THE FULL COST OF REGULATION
For federal agencies, there are obvious incentives to publish cost-benefit analyses that emphasize the latter while downplaying the former. Rarely will agencies publish a final rule with costs exceeding benefits, although the Obama administration has issued at least 23 regulations with this “upside down” ratio. For a cynic, a Regulatory Impact Analysis is more of a marketing document for agency action than an honest assessment of costs and benefits.
For the affected industries, trumpeting the costs of complying with a proposed regulation and highlighting the possible job-loss implications are two of the few public tools available to fight new regulation. However, despite these dueling cost-benefit analyses, few often capture the true costs of federal regulation.
Agencies frequently portray regulatory burdens as consisting mainly of transition costs or paperwork hours multiplied by median hourly salaries. Every major regulation should have at least a perfunctory discussion of deadweight loss, especially labor market regulations. A deadweight loss is the amount of consumer and producer surplus lost because of the increased price of a good or service caused by the imposition of a new regulation. For instance, a hot dog tax of $1 would mean that consumers will consume fewer hot dogs and pork producers will sell less meat. A portion of that reduction amounts to a loss of profits and consumer enjoyment that can’t be made up elsewhere—the deadweight loss. Even though the government gets some money from the tax, on net the economy loses.
A review of every final rule in the Federal Register from 1994 to present found just five rules (out of 1,400 major regulations issued over the past 20 years) that monetized the value of deadweight loss to the economy. A mere 23 even mentioned the term. And when agencies did monetize the value, the total among the five rules was just $16.4 million. We suspect that the reason these rules offered estimates of the deadweight loss was because it was small relative to other costs, making it safe to address the issue without a chance that it would affect the cost-benefit analysis.
Agency analysis on the full burden of recordkeeping and reporting costs often lacks a necessary thoroughness. For example, a National Labor Relations Board (NLRB) regulation from 2011 that would have forced all employers to notify employees of their union rights estimated familiarization would take two hours per company, at an hourly wage rate of $32.20. Although that may represent a trifle to a single firm, with more than six million affected employers, this regulation imposes an estimated compliance cost of $386.4 million, more than enough to label it a major rule. However, as an independent agency, the NLRB was not required to comport with the requirements of E.O. 12866. The lack of legal authority to issue the rule complemented the incomplete cost-benefit analysis, though, and a federal judge later struck down the rulemaking.
In general, the hourly wages attached to paperwork compliance activities are problematic. To assume that a particular rule will solely require the affected companies to reallocate labor a bit to ensure compliance is facile. It is increasingly common for companies to deal with myriad compliance burdens that can be complicated or costly by outsourcing them to a company with expertise in such an area, which may ensure compliance but results in a cost appreciably higher than any agency estimate.
There is an easy and practical way for the government to derive a more realistic estimate for labor costs based on this new reality and its own actions. For instance, the federal government routinely needs to estimate the full cost of labor when it pays for contract support. These are often referred to as “load” costs, or burdens beyond just direct wages. An examination of contracts from the General Services Administration shows that the government typically pays $189 to $225 per hour for contracting services. This is 3 to 3.6 times higher than $63 per hour of regulatory compliance government regulators often use to monetize the value of an hour, and 5.9 to 7 times higher than the $32.20 per hour estimate in the 2011 rule.
The choice of this number matters a lot. Using a true load factor of 3.6 turns what was a $386 million rule into a $1.3 billion regulation. At $200 an hour, its estimated cost approaches $2.4 billion. For perspective, there have been only 38 billion-dollar rules since 2001.
Finally, scholars have noted the paucity of regulatory analyses that estimate job loss. While it is true that the newfound emphasis on this has more to do with political pressures than a quest to inform the debate, it is a datum that should be considered in any cost-benefit analysis. However, the agencies that have incorporated a job loss estimate in their analyses have approached it with the perspective that regulatory activities actually create jobs by forcing companies to spend more money than they would have otherwise, which results in new jobs for people who produce or install safety and emissions reduction equipment, as well as those who inspect and have oversight over such activities. It amounts to a form of Regulatory Keynesianism and utterly ignores that increasing operating costs invariably result in higher prices, fewer sales, reduced productivity growth, and concomitant reductions in wages and employment. To tout regulations as job creation tools is fundamentally dishonest.
Even when agencies cop to a regulation causing job dislocations, they never monetize the effect of this loss. The University of Chicago’s Jonathan Masur and Eric Posner argue that these dislocations are costly to both workers and society, and derived an estimate of $100,000 per job lost. Even that number excludes the full socioeconomic effect of a job loss. Costs that high would render any number of proposed regulations currently on the dockets uneconomical.
We propose that the way to force the federal government to do a full accounting of the cost of any regulation is to take the job of cost-benefit analysis out of the hands of the agencies themselves and hand it over to an independent entity, as we suggested above. There is some precedent for this: the U.S. Department of Transportation had Harvard University’s Volpe Center do the analysis for a few rather involved regulations back in the 1990s and early 2000s, and the OIRA staff who worked with the DOT on those regulations came back impressed at the quality of the analysis.
Some combination of a new agency that does some analysis in-house and seeks outside expert advice for specialized regulations might be the best of all possible worlds.
“PRO-MARKET” AND “PRO-BUSINESS” ARE TWO DIFFERENT THINGS
The argument that regulation X will reduce the profits of business Y and therefore should be opposed by all right-thinking Republicans is nonsense. The purpose of regulatory actions should be to ensure that businesses pay any and all discernible negative external costs that they impose on society in some way. If a job carries a modicum of risk, the Occupational Safety and Health Administration compels the employer to take steps to reduce that risk if doing so saves business, workers, and the government money. It does not require a business to eliminate all risks—an impossible and undesirable outcome given the enormity of such a task. It is cost-effective for all of us to accept some risk in life, but it’s unacceptable for workers to not be aware of those risks.
If a business is imposing a substantial societal cost in some form, and the cost of remediating is too high for the business to stay open, then it ought to go out of business. The market should decide which businesses and industries survive and which will disappear. Negative externalities interfere with market discipline by allowing businesses to escape certain costs. Regulations should fix those external costs.
However, regulations can and often do go too far, beyond anything that can be limited by any OMB. That leaves businesses with burdensome regulations that cost them mightily but do little to solve any discernible problems that may exist.
LEGISLATIVE APPROACHES TO REGULATORY REFORM
Members of Congress interested in regulatory reform seem to understand some of the problems with our regulatory environment and have tried to reform the current system. The two most prominent efforts at regulatory reform are the proposed Regulatory Accountability Act (RAA) and the “Searching for and Cutting Regulations that are Unnecessarily Burdensome” (SCRUB) Act. Each would not only limit regulation in the present, but also place restraints on future presidents. The RAA would require advanced notice for a class of “high-impact” rules (those with costs of greater than $1 billion), tighten judicial review of agency actions from “arbitrary and capricious” to “substantial evidence,” and require agencies to adopt the “least cost” rulemaking.
The SCRUB Act would establish an independent commission to review major past rulemakings. It would set a goal to reduce cumulative regulatory costs by at least 15% and require all new agency actions to produce a retrospective review plan to measure success during implementation. There are roughly 80 major rules issued annually and only a few are reviewed after five or 10 years to determine effectiveness ex-post. The SCRUB Act would address this information discrepancy.
Our only complaint with these legislative approaches is that lawmakers impose somewhat arbitrary and indirect methods to address the problem of a rulemaking process that needs more accountability.
HARD WORK AND ETERNAL VIGILANCE
For politicians, the promise of repealing an unpopular rule may move the polls with certain constituencies, but it isn’t necessarily good policy. Adopting the precepts of a comprehensive regulatory reform that pays close attention to the true aggregate costs of society would help U.S. businesses be more competitive while boosting productivity, wages, and employment.
But there is no sword out there that will slice through the Gordian knot of our regulatory morass. Even the reforms we lay out here can very easily be subject to regulatory capture in the future and result in a process just as unaccountable as the one that currently exists. The way to obtain a more enlightened regulatory policy is the same as it is for obtaining liberty: eternal vigilance.
Readings
- “Unemployment and Regulatory Policy,” by Jonathan S. Masur and Eric A. Posner. University of Chicago Institute for Law & Economics Olin Research Paper No. 625 and University of Chicago Public Law Working Paper No. 412, December 9, 2012.
