Putting Regulators on a Budget

Jeff Rosen

Spring 2016

In nearly every type of organization, whether public or private, budgeting is a routine and unavoidable necessity. Could anyone imagine a unit of Walmart being told it did not need to budget and instead could spend any amount it wanted? Could any household function that way? Government surely can't, and since the 1920s the annual fiscal budget process has limited what federal agencies can spend each year. An even older federal statute known as the Antideficiency Act makes it unlawful for a federal-agency employee to spend more money than his agency was authorized. Violations carry significant consequences, including the possibility of criminal penalties.

But what about federal regulators? In making rules, what constrains the total level of costs that agencies can impose on private parties and on the economy, and the amount of non-Treasury resources they can require to be spent on government programs? The answer is nothing: No such budget exists.

There are many ongoing debates about regulation in the United States, and there are competing views about how much regulation is beneficial and worthwhile. But there is no dispute that the regulatory system is costly and has an effect on virtually every American. In a democracy like ours, who decides how much regulatory cost is "enough" or "the right amount"? Right now, the answer is that the federal agencies themselves decide, and very little has ever been done about this state of affairs. Since at least the 1940s, there have been efforts to require agencies to follow procedures dictating how they develop rules, but none to put limits on the costs they can impose.

In our time, allowing agencies to decide on the appropriate amount of costs to impose on the economy increasingly means that the president actually makes that decision. And in an environment in which the president is unable to persuade Congress to enact his priorities, it should be no surprise when regulation and its costs continue to increase. The regulatory process effectively provides the president with a way around congressional resistance to his agenda and its cost. Particularly when one form of resource allocation (spending) is limited by Congress while the other form (regulation) is not, there is a hugely skewed incentive to use regulation as the chief instrument of a more activist government.

Given the significance that regulations now have for our economy, our system of government cannot reasonably continue to work like this. It is plainly time to impose the same kind of overall discipline on the regulatory system as is already in place on the fiscal system, however imperfect it may be. Proponents and opponents of more regulation often picture the other side in extreme terms, as though the right prefers to have no regulation at all or the left favors infinite regulation. When pressed, most will usually agree that there is a need for some regulation and that there can also be excessive regulation that harms the economy or does more harm than good. And virtually no one disputes that regulation often imposes significant costs and burdens.

This suggests that both our debates about regulation and our actual system of regulation would benefit from the development and implementation of a regulatory budget. Such an innovation could help us accept the regulations that are needed while developing ways to limit excessive regulatory costs. And it could transform our political debates about regulation by grounding them firmly in reality. The time has come to develop and implement a regulatory budget, with caps on the costs of regulation.


A lot of what we call regulation is in effect a form of taxing and spending. Costly rules are a kind of "hidden tax" that imposes costs on private parties by forcing them to use their resources for government-mandated purposes. The cost of complying with the regulations for businesses, universities, hospitals, labor unions, non-profits, and even local governments is a kind of fee, though one that is neither transparent nor voted upon by the citizens' representatives in Congress.

At the same time, because these regulations also mandate and control the use to which private resources will be deployed for governmental objectives, they also function as a substitute (or supplement) for direct government spending. For example, the government could provide insurance coverage to workers by paying for it directly — spending its tax revenues or proceeds from public debt — or it could require by regulation that employers provide that same insurance coverage, using the employers' own private funds.

A related but not always understood point is that federal regulations issued by agencies are effectively laws — they are legally binding on everyone who lives and works here. But they are much easier to promulgate than laws passed by Congress, since they are generally made under the control of the president and his presumably like-minded appointees.

There is now an extraordinary number of regulations on the books. Today the Code of Federal Regulations is 175,268 pages in 236 volumes, up from approximately 141,000 pages in 206 volumes back in 2001. (In 1975, it was 71,224 pages in 133 volumes, so it has roughly doubled in the last 40 years.) According to estimates prepared at the Mercatus Center at George Mason University, just to read today's 236 volumes of the Code of Federal Regulations would take an individual person nearly three full years, if that is the only thing he did full-time. The absurdity of this is what Winston Churchill was warning against with his 1949 remark, "If you make 10,000 regulations you destroy all respect for the law."

By now, it is almost cliché to mention how much new red tape is being issued each year. Each new round of major federal legislation, such as the Affordable Care Act of 2010, the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010, and the Food Safety Modernization Act of 2011, provide agency regulators with new authority to issue more new regulations, and the current administration has discerned novel ways to issue once-unthinkable new rules under old laws like the Clean Air Act of 1970.

In hindsight, it is remarkable how much authority Congress has delegated to federal agencies (and the president) over the last century, during various waves of legislative fervor for new forms of government oversight of the private sector, often in reaction to perceived misdeeds or failings. Examples are numerous, beginning more than 100 years ago with the 1906 Federal Food and Drugs Act and the 1914 Federal Trade Commission Act. The 1930s saw the creation of the Securities and Exchange Commission, the National Labor Relations Board, and many others. In the 1960s, Congress created the Equal Employment Opportunity Commission, and the next decade produced the National Highway Traffic Safety Administration, the Environmental Protection Agency, and the Occupational Safety and Health Administration, among others. More recent additions are the Transportation Security Administration, the Consumer Financial Protection Bureau, and various new entities within the Department of Health and Human Services created by the Affordable Care Act. Not only has the government continually expanded for more than a century, but Congress has delegated extremely broad lawmaking power to agencies, with virtually no limitations as to the costs agencies can impose on regulated parties.

The results have been all too predictable. More laws are made by federal agencies than by Congress. In 2014, Congress enacted 223 laws (including more than 50 namings of federal buildings and properties, five medal awards, several Smithsonian regent appointments, and the like). Looking only at "significant" rules that are reviewed by the White House Office of Management and Budget, federal agencies have finalized an average of 352 final new rules per year between 2009 and 2014. If one looks at all new rules issued by federal agencies, during 2014 alone federal agencies issued 3,554 final new rules in the Federal Register.

There are more coming. In 2015, federal agencies proposed 2,342 more new rules. That continues the trend by which the Code of Federal Regulations has grown by nearly 12,000 new pages just since 2009. And a number of these new rules involve record-breaking costs, such as the EPA's most recent plans for revised regulations for mercury and ozone, and its novel new rules for carbon dioxide, all of which involve billions of dollars per year. There are also costly new rules from the Department of Labor, the Department of Transportation, the Department of Energy, and elsewhere. Using the government's own estimates, regulatory analysts at the American Action Forum have compiled data showing that federal agencies' new rules since 2009 have added more than $115 billion per year in annual expenditures to comply with the new requirements. They calculated the total cost of new rules added since 2009 at $710 billion, even before the coming round of new rules.

Perhaps most significant, all of these added costs weighing down our economy are essentially "off-budget," meaning that they are a way for the government to require spending on federal-agency objectives without requiring congressional appropriations. Particularly after the Budget Control Act of 2011 placed caps on agency spending, regulations became the easiest means of directing non-Treasury funding to the policy priorities of the executive branch (as well as those of the "independent agencies," whose leaders are appointed by the president).

In recent years, there have been several studies published describing the aggregate, cumulative cost of regulation. The numbers are substantial. In 2010, economists Mark and Nicole Crain prepared a study for the Small Business Administration that estimated the cumulative total annual cost of regulations to be approximately $1.75 trillion in 2008. In 2014, in a study prepared for the National Association of Manufacturers, the same team estimated this cost had grown to just over $2 trillion per year. Separately, in 2015, regulatory analyst Clyde Wayne Crews from the Competitive Enterprise Institute estimated the total annual cost of regulations to be approximately $1.88 trillion per year. In another paper, Crews looked at a 2002 Office of Management and Budget study and, after adjusting its data to 2013 dollars, he reported an estimated cost of $1.25 trillion annually. To put figures like this into perspective, in 2014 the U.S. Treasury collected $320.7 billion in corporate income taxes and approximately $1.4 trillion in individual income taxes. Putting aside some technical aspects about comparing costs measured in different ways, it is not yet widely appreciated that the annual cost of regulations may be roughly as consequential as what the public pays in income taxes.

Looked at another way, apart from automatic entitlement spending, Congress appropriated approximately $1.1 trillion in 2014 for spending by the entire government (that includes the Departments of Defense and Homeland Security, as well as all other domestic agencies). The available studies say that the annual cost of regulation likely exceeded that. And while appropriated federal spending generally did not increase during the four years after Republicans took control of the House of Representatives in 2010, the cost of regulation increased substantially. An analysis of agency estimates by the American Action Forum identified a total of about $85 billion in new annual regulatory costs added by rules issued between 2011 and 2015, with a net present value of nearly $500 billion in total new regulatory costs added since the start of 2011. Spending on everything from national defense to transportation infrastructure to higher education was basically frozen, but the cost of regulation increased by billions of dollars each year. For the people who bear the cost of regulation, it makes little difference whether the pain comes from taxes or from regulations: Their economic burdens are increased either way.

Even if we froze federal spending, would it matter to the economy if regulatory costs increased on businesses, schools, hospitals, non-profits, and others? The relatively few studies of the macroeconomic impact of regulatory costs suggest that it does matter, in much the same way that tax policy does. A World Bank study in 2005 observed that "a heavier regulatory burden — particularly in product and labor markets — reduces growth." More recently, a 2012 study by NERA Economic Consulting for a manufacturing trade group concluded that manufacturing regulations alone reduced U.S. gross domestic product by somewhere between $240 billion and $630 billion, such that an average American household lost between $1,800 and $5,000. And a 2013 academic study by economists John Dawson and John Seater concluded that U.S. regulations that were added between 1949 and 2005 "reduced the aggregate growth rate on average by about two percentage points." In 2014 alone, 2% of GDP came out to approximately $350 billion dollars forgone.

Nonetheless, regulation is often the most convenient way for government to achieve political leaders' policy objectives, and it will not readily be restrained. Indeed, the last three presidents before President Barack Obama each set a record for the number of costly, "economically significant" (involving more than $100 million per year) final rules published in his last year in office, and that was true regardless of his political party.

Moreover, regardless of who is elected next, even the most zealous of limited-government advocates is likely to find himself confronted with a bureaucracy that continues to deliver a persistent number of new rules on its own, and with at least a few cabinet and sub-cabinet officials who will passionately seek to use costly regulations to achieve key policy goals.

So what is to be done? How can government agencies be allowed to accomplish the necessary elements of their missions, without generating a cumulative and aggregate level of regulation that is harmful to economic prosperity and individual liberty?


Concerns about regulatory excesses are not new, nor are concerns about the costs imposed on the economy by regulation. Poll results in recent years reflect heightened levels of public concern about over-regulation, but the general unease with out-of-control regulation is longstanding.

There have been proposals to reform the way regulatory agencies operate going at least as far back as 1940, when Congress passed the Walter-Logan bill, which aimed to rein in the growth of New Deal agencies, although President Franklin Roosevelt vetoed it. In 1946, Congress tried again, passing the Administrative Procedure Act, which President Harry Truman ultimately signed, despite some of Roosevelt's advisors urging him to veto it. The APA was and remains an important law, but it addressed procedural requirements to limit agency caprice, arbitrariness, and blatant lawlessness; it did not address the costs of regulation. If anything, its provisions on rulemaking (and court interpretations of them) may have unwittingly resulted in making it relatively easier for agencies to issue broad and costly rules.

While calls for further regulatory reform continued for decades (including the work of the Second Hoover Commission in 1955, the Ash Council Report of 1971, the American Bar Association's 1979 report on Federal Regulation, and others), major legislative changes to federal-agency rulemaking were not passed by Congress. Two modest new laws were enacted in 1980, the Paperwork Reduction Act and the Regulatory Flexibility Act. But in the early 1980s, President Ronald Reagan took important steps to manage how federal agencies issued rules, with stronger requirements for cost-benefit analysis of certain rules and review by the Office of Management and Budget. Even during those years, however, Congress did not enact significant new laws to alter and improve the rulemaking process.

In the mid-1990s, when Republicans gained majorities in both houses of Congress for the first time in four decades, new proposals were made to reform agencies' authority over rulemaking. Several became law, four of which bear note in this context.

First, Congress passed and President Bill Clinton signed the Unfunded Mandates Reform Act of 1995. The measure was thoroughly bipartisan, garnering 86 votes in the Senate and 394 in the House. Among other things, that law required that, for new regulations involving more than $100 million annually, agencies would be required to conduct a regulatory analysis (generally of costs and benefits). However, this law is not subject to judicial review with regard to the content of those analyses, so it is up to the agencies and the Office of Management and Budget to ensure the studies are actually undertaken.

Second, Congress passed and President Clinton signed the Congressional Review Act of 1996. (This law likewise received 328 bipartisan votes in the House and unanimous consent in the Senate.) This law created a process by which Congress can vote on resolutions of disapproval of regulations, subject to veto by the president, and thus in effect usually requiring two-thirds of both houses of Congress to eliminate a rule promulgated by a federal agency. Consequently, only one resolution of disapproval has ever been passed by Congress and signed by the president. It involved an OSHA ergonomics rule issued during the Clinton administration, the resolution of disapproval of which was signed by President George W. Bush in 2001. On four other occasions, all in 2015 and 2016, Congress sent resolutions of disapproval to President Obama, who vetoed them and thereby preserved the regulations: They involved an NLRB union elections rule, two EPA greenhouse-gas regulations, and an EPA regulation concerning waters of the United States. While Congress can consider cost (or any other factors that members wish) in voting to disapprove individual regulations, the necessity for a two-thirds vote to override the president's support for such regulations has limited a more robust use of this process.

Third, Congress passed and President Clinton signed the Information Quality Act of 2000, which required that agencies issue guidelines for maximizing the "quality, objectivity, utility, and integrity" of information disseminated by agencies, and that they develop mechanisms for affected parties to seek correction of information that does not meet the standards. But the effectiveness of this legislation has been limited by courts holding that the law cannot be enforced by judges and questioning whether it even applies to rulemakings.

Fourth, Congress passed and President Clinton signed the Regulatory Right-to-Know Act of 2000, to require that the Office of Management and Budget annually report to Congress on the cost and benefits of regulations from the preceding year, including the aggregate totals and the effect on economic growth. However, in practice, these reports have been limited to a listing of selected rules for which agencies conducted cost-benefit studies, and they have excluded some agencies, so the reports have not met their intended purpose, and in recent years have not been issued in a timely way.

As the flow of costly regulations has continued and increased in recent years, it is unsurprising that regulatory analysts — and many members of Congress — have concluded that the reform efforts of the 1990s have not proven adequate to the problem. Since 2010, a number of regulatory-process reforms have been proposed, several of which have passed the House of Representatives in the current Congress (and some passed the House in the prior Congress as well). Examples include the Regulations from the Executive in Need of Scrutiny Act (REINS Act), the Regulatory Accountability Act, and the Sunshine for Regulations and Regulatory Decrees and Settlements Act. Some of these, such as the Regulatory Accountability Act, would make a significant and positive difference in the way regulations are made. They deserve to be considered on their merits as ways to improve the functioning of the administrative process.

Perhaps surprisingly, however, none of the bills passed by the House so far directly addresses the problem of ever-increasing regulatory costs, or the incentive to use rulemaking as an alternative to spending without the budget caps that apply to federal spending.


What this experience reveals is that the time has plainly come to develop and apply a regulatory-cost budget, akin to the fiscal budget, that would provide both overall and agency-by-agency limits on the total amount of costs imposed by agency rules. This would not end or preclude necessary and appropriate regulation, but it would manage its costs and burdens in a more rational way.

In principle, this idea is not new. Over the years, there have been several different proposals related to regulatory budgeting. Regulatory scholars and practitioners such as Christopher DeMuth, John Morrall, Clyde Wayne Crews, James Tozzi, Fred Thompson, Jason Fichtner, Patrick McLaughlin, and others have written about the concept during the last 35 years. Moreover, the origins of the regulatory-budget concept were bipartisan.

In 1978, the Joint Economic Committee, chaired by Democratic senator Lloyd Bentsen of Texas (later Treasury secretary under President Clinton), held a hearing about regulatory costs, and that same year he introduced the first legislation to create a regulatory budget. Senator Bentsen's stated aim was to "force agencies to choose the least costly way of achieving regulatory goals."

Bentsen's Federal Regulatory Budget Act would have amended the Congressional Budget Act to establish a joint legislative-executive budget process governing regulatory costs. First, in November of each year, each regulatory-agency head would submit a report to Congress and the president detailing the actual costs of complying with all regulations under the agency's purview in the preceding fiscal year, as well as the expected costs of compliance with all new or existing agency regulations in the upcoming fiscal year.

Second, the president would include with his annual fiscal budget submission a "regulatory budget for each agency," containing "recommendations for the maximum costs of compliance." If the president's budget called for a cut in the regulatory costs projected by an agency's November report, the president would recommend specific actions to achieve those cuts.

Third, based on the agency reports and the president's budget, the congressional budget committees would develop and Congress would vote on a concurrent resolution "to establish a regulatory budget for each agency which sets the maximum costs of compliance" for all rules in effect for the upcoming fiscal year. The Bentsen bill created no hard sanction for exceeding authorized regulatory costs. Instead, an agency's November report would simply include a comparison of the actual costs of compliance for the preceding year with the budgeted costs and provide a "full explanation for any costs of compliance which exceeded the regulatory budget for such fiscal year."

Two years later, President Jimmy Carter's 1980 Economic Report of the President briefly discussed the regulatory budget, noting the practical difficulties of measuring costs, but concluding that "tools like the regulatory budget may have to be developed" to improve regulatory priority-setting and cost-effectiveness. That same year, the White House Office of Management and Budget circulated among agencies a proposed "Regulatory Cost Accounting Act" that would have created a "regulatory cost accounting system" resembling a regulatory budget. The proposal would have required each agency to submit to OMB an annual report on the costs of its regulations for the upcoming fiscal year. The president would then submit to Congress a plan to adjust those costs by revising regulatory statutes. But the Carter administration's draft legislation was never actually submitted to Congress.

In the three decades since then, policymakers and regulatory scholars alike have at different times debated various proposals to create a regulatory budget. The idea follows from the core principle that an agency's authority to require private expenditures to comply with new regulations should be limited and allocated much like the authority to spend public tax dollars. As such, the call for a regulatory budget is premised on the view that the transfer of private resources by regulation is no less a cost imposed by government than the collection and consumption of private resources through the tax and spending powers. But while government expenditures are constrained by the ability to tax and borrow, regulatory costs are subject to no built-in limitations. By creating a systematic limitation on regulatory costs at the agency level and across all agencies combined, a regulatory-cost budget would counteract the tendency to treat private resources as a "free good" and counteract the political impulse, in an age of increasingly untenable levels of federal debt, for political actors to rely increasingly on regulation to "fund" major government initiatives.

A regulatory budget would also yield a number of benefits beyond this critical realignment of fiscal and regulatory incentives. First, it would improve priority-setting and the economic efficiency of regulation across agencies and programs. In 1980, the Economic Report of the President noted that "the regulatory process as yet lacks any mechanism analogous to the expenditure budget for comparing and integrating priorities among different program areas." That is still true. A regulatory budget would correct this deficiency by enabling comparisons across regulatory programs and facilitating the coordination of costs and benefits of various initiatives, as is done in the fiscal budget. Doing so would allow government decision-makers to consider regulatory costs in the more realistic context of a hierarchy of competing public interests, rather than a series of isolated regulatory initiatives that are never compared to one another.

This broader view would improve the economic efficiency of regulation by permitting resources to be transferred from low-yielding regulatory efforts to those with more "bang for the buck." In addition, a regulatory budget would also allow regulators to consider similar costs imposed on the same group of regulated industries by other federal agencies. It would help to avoid the "overloading" of one sector of the economy with new costs in a very short timeframe. Thus, by requiring judgments about an integrated regulatory strategy, a regulatory budget would force trade-offs among competing priorities and broaden the "tunnel vision" that sometimes plagues agencies that are focused only on their own missions and objectives.

Second, a regulatory budget would for the first time give agencies an internal incentive to limit the compliance costs of their regulations, and to achieve a greater level of cost-effectiveness. Under the current regulatory regime, the only pressure on agencies to minimize costs is largely external — from stakeholders, from OMB, or in some cases from the courts. By giving an agency a fixed sum of regulatory costs it can "spend" each year, a regulatory budget would align the agency's interests with the goal of cost control. An excessively costly regulation would come at an opportunity cost to the agency, because it would require the agency to forgo other regulatory initiatives.

Third, a regulatory budget would likewise for the first time give agencies an internal incentive to revisit and repeal their own old rules that are ineffective, outdated, or unnecessary, because repealing such rules would create budget "savings" that could be used for more effective purposes. Requirements for "retrospective review" of regulations have existed for decades, with sadly little to show for them. Perhaps that is because so-called "command and control" requirements are relatively less effective than incentives for improvement. A regulatory budget would for the first time align agency incentives with those of the public in favor of jettisoning archaic or useless rules.

Fourth, a regulatory budget would strengthen political accountability for rulemaking. Rather than a series of discrete regulatory actions under authority delegated by past Congresses, regulations under a budget constraint would be the product of a yearly, comprehensive review by politically accountable elected officials. Such officials would become responsible for making policy decisions regarding the allocation of resources to regulatory goals and programs. In the words of former OMB official Christopher DeMuth, a regulatory budget "would acknowledge explicitly the political nature of regulatory benefits and permit the President and Congress to make political judgments in light of more thorough information about economic costs."

Fifth, the additional transparency of a regulatory budget would be valuable for its own sake, as it would provide policymakers and the public with a more complete picture of the economic footprint of regulation. The Senate Joint Economic Committee argued in a 1979 report that "with the rapid growth of new regulatory agencies....the Federal budget no longer conveys a complete picture of the Government's economic impact." Moreover, the Carter administration's 1980 Economic Report of the President stated that, "as more goals are pursued through rules and regulations mandating private outlays rather than through direct governmental expenditures, the Federal budget is an increasingly inadequate measure of the resources directed by government toward social ends." A budget would provide the public and Congress with better information to evaluate regulation and, as University of Chicago law professor Julie Roin pointed out in a 2002 working paper, "more information about the strength of interest groups." The economic impact of spending programs, taxes, and so-called "tax expenditures" are reported in detail every year to Congress and the public. That same transparency for regulatory costs would by itself be a desirable feature, whether or not it is combined with enforceable cost caps.

Sixth, and relatedly, having a regulatory budget would enable Congress and other policymakers to begin to consider the macroeconomic impacts of regulatory-policy decisions, in a manner somewhat comparable to what they do now regarding tax policy. While it would not be simple or easy to adjust to desired levels and consequences, the effects of cumulative regulatory costs on macroeconomic performance have been wholly ignored or discounted for too long. As Professor Richard Epstein of New York University recently wrote in these pages, "[h]igher regulation and taxation reduce gains from commerce and stifle innovation. People in high places who ought to know better seem to have a very hard time learning this basic lesson."

Finally, as emphasized earlier, a regulatory budget would counteract the unhealthy and undemocratic incentives for unelected agency officials to use regulation as a way to fund their preferred policies without having to obtain appropriations from Congress. Fiscal measures and regulation are often interchangeable tools of policy. A budget is the tool to constrain them both, and ultimately to subject them to the wishes of the American people when they vote for Congress and the president.


In the near term, the key to implementing a regulatory-budget system would be to focus on new, incremental costs of new regulations only. For major new rules, federal agencies have already been required for several years to prepare cost-benefit studies, from which the cost data could be extracted for budget purposes. To commence the process of budgeting, allocations and caps can be set on prospective new rules alone for an initial period of time.

The reason to proceed prospectively with regard only to new regulations is that implementation challenges are one of the key reasons that a regulatory-budget process has not already been adopted, even when policymakers with concerns about regulatory costs have been in the White House or in the majority party in Congress. At this juncture, however, the challenges could be managed if budgeting is applied to the costs of new rules.

Here is how it would work: A regulatory-budgeting process would involve multiple steps, analogous to the fiscal-budget process. First, agencies would identify their proposed plans for the upcoming year, with estimated costs for every new regulation and overall. Second, prioritization would take place, with caps being imposed both as allocated to each agency and with a "top line" cumulative limit across all agencies. (Some reserve authority or waiver process for natural disasters or other emergencies would likely be included.) With regard to the "top line," the president (or OMB) would have some limited authority to transfer some amount of cost authority between agencies, in order to use regulatory-cost allowances where higher benefits might be obtained. Third, agencies would be expected to comply with the caps, and to report on their compliance after the conclusion of the year.

Who would set the regulatory-budget caps and allocations? Ultimately, that should be Congress, as with the fiscal budget. The House and Senate Budget Committees, or some other committee given centralized cost-budgeting authority, would need to have the responsibility for providing the overall regulatory-cost budget, as is done with the fiscal budget. But it could be beneficial if the president initiated a prototype of this system by executive order and imposed caps unilaterally initially. That would allow a period of time in which the executive branch could develop necessary systems, identify challenges, and iron out any wrinkles. It would also enable some time for agencies to focus on improving the quality of their economic analyses and expanding processes for public input, to ensure that the cost estimates for their regulations are as accurate as reasonably possible. (Unfortunately, it would also mean some budget exceptions might be necessary for rules required by Congress in prior statutes with upcoming statutory deadlines, unless relief could be obtained by legislation. That is another reason why, ultimately, it will be necessary for Congress to enact regulatory budgeting by legislation.)

With regard to the setting of budget allocations and caps, it is important that the budget caps apply to costs — not to benefits — just as with the fiscal budget. When Congress appropriates fiscal expenditures, the budget puts a cap only on the cost. One of the main considerations in budgeting is that resources are finite and not unlimited, regardless of how many good and legitimate potential uses might exist for the money. Choices have to be made among them. By contrast, a cost-benefit evaluation is an essential tool for discerning whether proposals do more good than harm and therefore are rational to pursue at all. The cost-benefit evaluation can also identify which proposals have the greatest net benefits, doing the most good at the least cost. That can help select among various proposals within the budget cap. But the cap needs to apply to costs alone, without regard to benefits. That would be true in all events, but is especially true in the real world of regulation, where estimates of benefits are far more speculative and subject to controversy than estimates of cost.

In addition, whether done voluntarily or imposed by legislation, it would be beneficial for the next president to submit to Congress planned regulatory "spending" at the same time he submits the fiscal budget, and to report the following year on all the rules that were finalized and whether the final cost estimates fell within the planned caps.

When the concept of a regulatory budget was first developed more than 30 years ago, the Code of Federal Regulations was smaller, and there were numerous other proposed tools being considered as ways to improve the regulatory process. Decades later, we now see that other reforms have not functioned as successfully as we initially hoped, and the Code of Federal Regulations has nearly doubled. But experience with some of the reform tools, such as the use of cost-benefit analysis across the terms of five presidents of both parties, has provided a foundation of experience with estimating regulatory costs, which would be used to move toward an incremental regulatory budget.

Now that we have more than 30 years of experience with other forms of regulatory analysis, it is clear that there are two primary continuing challenges. One is the measurement of costs, as a consistent method for accounting for estimated costs would need to be deployed. In the beginning, it likely would be feasible to harmonize somewhat different methodologies and apply a certain adjustment "cushion" to the estimates or to the application of the caps. But over time it will be necessary, as with the fiscal budget, to develop greater consistency across the government.

The second key challenge is ensuring the reliability and integrity of the costs — and enforcing the application of the budget caps to them. On the fiscal side, this is sometimes described as the challenge of "scorekeeping" and enforcement. It will be at least as important on the regulatory side, and raises the question of what entity will be the final authority on the cost numbers and the enforcement of limits. Because of the constitutional separation of powers (as construed by the U.S. Supreme Court in the Bowsher v. Synar decision in 1986), budget enforcement can only be implemented by the executive branch and not Congress or its agencies such as GAO or CBO. Hence, in the beginning it would certainly be advantageous to have a president who is supportive of the regulatory budget and will implement it faithfully. Over time, however, it is likely that Congress would want to have at least its own watchdog monitoring the regulatory budget, just as CBO and GAO do with the fiscal budget that is administered authoritatively by OMB.

Fortunately, with regard to both of these challenges, there is recent, real-world experience to draw upon from forms of regulatory budgeting already in effect in the United Kingdom, Canada, and Australia. Each of these countries has adopted a regulatory-budget system of "one in, one out" that is roughly similar to our fiscal pay-as-you-go approach to requiring offsets to the costs of any new regulations. (It is in essence a budget with the incremental new cost cap set at zero, so all new costs have to be offset with the elimination of pre-existing regulatory costs.) Given the strong indicators of success with those systems in other countries, in June 2015 the Senate Budget Committee and the Senate Homeland Security and Government Affairs Committee held a rare joint hearing, at which the president of Canada's Treasury Board described his country's successful experience with this particular form of regulatory budgeting. But adopting a regulatory budget in the United States will require careful planning, and will inherently involve a period of adaptation to the accounting and other new requirements that would be involved.

Moreover, in the decades since Senator Bentsen's Federal Regulatory Budget Act was introduced, the global economy has undergone seismic shifts that make the recognition and management of regulatory costs even more important. Because the United States competes in international markets, there are consequences to consider when U.S. regulations increase relative to other countries' and thus decrease American competitiveness. The key role of technological innovation in economic growth has to be considered as well, as regulators struggle with the application of old regulatory regimes to innovations in transportation (e.g., drones), communications (e.g., internet speed), health care (e.g., genomics), finance (e.g., bitcoin), and other realms. Concerns like these present fundamental questions about the dangers of allowing regulatory costs to expand without limits.


In the long run, regulatory reform will need to go even further than this prospective budgeting system. The enormous volume of existing rules — and the costs they already impose — will also need to be considered. There will need to be an effort to estimate the costs of existing rules in a more systematic way, both to facilitate the identification of rules that might be revised or repealed and to create over time a full "baseline" of the total cost of regulation. This would allow us to know whether the cumulative, aggregate cost is increasing, declining, or remaining constant, so as to inform the process of setting the regulatory-cost caps.

The importance and the difficulty of this task can hardly be overstated. It will take time, and some serious resources. But thankfully, this bigger and more daunting task need not hold up the more immediate effort to put regulators on a budget. A regulatory budget for new incremental costs could be launched far more quickly and would yield some enormous benefits while paving the way for further steps to come.

The perfect should not be the enemy of the very good when it comes to regulatory reform. And because the best time to start the process of implementing a regulatory budget would be near the start of a new administration, the time to adopt this idea should be 2017. Doing so would facilitate its adoption at a time when a brief regulatory "time out" is usually in place already, and when a new president's priorities are being set and disseminated, and new personnel are being appointed to implement them.

The basic structure that a regulatory budget should take is clear enough. The mechanism to put it into place has already been worked out over decades. The rationale for it has never been stronger. And next year would be the perfect time to start.

Jeff Rosen is an attorney in Washington, D.C., and chair of the American Bar Association's section on administrative law and regulatory practice.


from the


A weekly newsletter with free essays from past issues of National Affairs and The Public Interest that shed light on the week's pressing issues.


to your National Affairs subscriber account.

Already a subscriber? Activate your account.


Unlimited access to intelligent essays on the nation’s affairs.

Subscribe to National Affairs.