Regulatory Review for the States
Dudley Square is Boston's busiest bus stop, with 30,000 passengers passing through daily. Eighty thousand people live within a mile of the stop. Yet the area is an entertainment vacuum, with almost no restaurants, clubs, or coffee houses. The lack of local businesses is not just an inconvenience to local residents; it represents a dearth of service-sector jobs that are badly needed in the neighborhood.
When local leaders are asked why businesses are not taking advantage of Dudley Square's constant flow of potential customers, nearly all of them mention the city of Boston's difficult regulatory regime. The state legislature has capped the number of liquor licenses in Boston since the 1930s, and as a result, the price for a license can top $300,000. Any "common victualler" will need to obtain a bevy of licenses and attend a Licensing Board hearing. If a bar wants to install a television set, it will need an entertainment license, and if the television exceeds 27 inches, the bar will have to pay a $180 fee. If it wants to go from two to three televisions, that will require an added trip to the Licensing Board. There are reasons behind all of these rules, of course. But together, they can kill a neighborhood.
Sensible regulatory requirements can reduce illnesses and accidents, protect the environment, and maintain quality of life. But when regulations are onerous and poorly designed, they can cause serious harm — overwhelming small businesses, reducing economic growth, eliminating jobs, squelching innovation, and causing serious hardship. Many studies have found that entrepreneurship is the lifeblood of urban regeneration, meaning that business regulations that stymie entrepreneurship, like the ones in Boston, can have particularly large costs, especially during times of economic difficulty. While regulation is an important tool for governments, it is vital that the authors of such rules strike the right balance, giving careful consideration to the track records of old rules and the likely consequences of any new requirements.
For more than three decades, under both Democratic and Republican presidents, the national government has attempted to strike this balance and discipline the regulatory process by requiring detailed analysis of both the costs and benefits of new regulations. Only if the benefits of a proposed rule justify the costs can regulatory agencies go forward and burden the private sector. Cost-benefit analysis provides an important mechanism for evaluation; it also operates as a safeguard against the power of interest groups, including big businesses, which often try to encourage agencies to regulate with their interests in mind. This kind of analysis is intrinsically neither anti- nor pro-regulation. It is simply a tool that helps policymakers produce good regulations and prevent bad ones from being made law. Whether or not the tool has proved adequate — and at the national level, things could surely be far better than they are — it has certainly made things better than they would otherwise be.
Cost-benefit analysis of proposed regulations has been so valuable that in 2011 the Obama administration, with bipartisan support, called for an ambitious process through which federal agencies would periodically evaluate their existing rules. This "regulatory lookback" attempts to eliminate or streamline regulations based on the results of cost-benefit analysis. This process has produced over 500 reform proposals so far, and it is saving billions of dollars each year.
As the community around Dudley Square knows all too well, however, the federal government is not the only entity that issues and enforces regulations. State and local governments issue their own (sometimes wasteful, onerous, or harmful) regulations, but remarkably, these more local rules are rarely subject to anything like rigorous analysis. In today's economy, the need for regulatory discipline from state and local governments, and for careful cost-benefit analysis, is especially pressing. Such governments are often responsible for requirements that many view as burdensome, unjustified, and harmful. Worse, sub-national governments generally lack an institutional mechanism even to evaluate these complaints.
State and local authorities need regulatory discipline and cost-benefit analysis as much as the federal government does. In fact, because of their relative capacity and incentive to innovate, states have the opportunity to push beyond the tools the federal government has tried, finding creative ways to institutionalize regulatory simplification and freeing up the private sector without jeopardizing public safety, health, or the environment.
For the federal government, the modern era of regulatory review began in 1981 when President Reagan issued Executive Order 12291, imposing two sets of requirements, both effectively binding on the executive branch. The first was procedural; the second was substantive.
Reagan's procedural requirement directed agencies to submit their regulations to the Office of Management and Budget for approval. Created by the 1980 Paperwork Reduction Act and operating within OMB, the Office of Information and Regulatory Affairs (OIRA) gained a great deal of authority over the regulatory process. Subject ultimately to the power of the president himself, OIRA could effectively decline to authorize agencies to impose regulations on the public. It was understood that OIRA would not be acting on its own; its decisions would be reached in close concert with others in the executive branch, including technical specialists of various kinds — an important point to which we will return.
The substantive requirement directed agencies to produce careful analyses of the costs and benefits of regulations, and to proceed only if the benefits outweighed the costs (subject to legal mandates issued by Congress). Under this approach, an agency could not finalize a regulation protecting worker safety if the costs were, say, $700 million and the benefits were $20 million. The importance of this constraint cannot easily be overstated. It was designed to ensure that agencies would make decisions after evaluating the actual consequences of the rules they proposed — instead of being swayed by interest groups, dogmas, and gripping anecdotes.
As part of the requirement to balance costs and benefits, the executive order also directed agencies to "maximize the net benefits." This meant that even if a proposed approach — involving, say, environmental protection — would have net benefits of $100 million, agencies could not proceed if another approach, whether less or more stringent, had net benefits of $200 million. (The task of monetizing costs and benefits presents its own challenges, and we will return to this important point below.)
In imposing these requirements, Reagan was motivated by three distinct concerns. The first involved purely economic considerations. In his view, many regulations had done a great deal of harm, and it was important to ensure that, if the government was imposing significant costs, it was creating significant benefits as well. A clear goal was to scale back the production of new regulatory requirements by ensuring that they had strong, evidence-based justifications.
The second and very different motivation involved political control. During the early days of the Reagan administration, there was a widespread view that the sheer number of regulatory agencies created a serious problem for presidential management. Even if the heads of the agencies were presidential appointees, some of their regulations might not reflect the principles or priorities of the president. OIRA review, undertaken with close reference to those principles and priorities, could supply a corrective. A central goal, then, was to ensure hierarchical control of the operations of regulatory agencies with close reference to the principles and commitments of elected officials, especially those of the president.
The third concern involved inter-agency coordination. No single agency can easily think through all the possible consequences of a proposed regulation, and every agency could benefit from the expertise of others in making regulatory decisions. In this respect, the OIRA process was understood to have a coordinating function, ensuring that multiple actors, with their own areas of specialization, could bring their knowledge to bear. For example, the Small Business Administration knows a great deal about the effects of regulatory requirements on small business, and OIRA could help provide a mechanism by which the views of the SBA would be taken seriously as new regulations were considered.
During the Reagan and George H. W. Bush administrations, OIRA review proved to be highly controversial. Many policymakers and analysts, especially on the left, were enthusiastic proponents of regulation, which they thought an important means of protecting public health and the environment. In their view, regulatory review tended to operate as an unfortunate obstacle to desirable initiatives. It was therefore important, and indeed historic, that President Bill Clinton essentially approved the Reagan approach with his own executive order in 1993.
President Clinton did inaugurate a few important changes with Executive Order 12866. For example, he reduced the number of rules subject to OIRA review, generally limiting it to those with a significant economic impact ($100 million annually) or raising novel questions of law or policy. He changed the language of the substantive requirement so that the benefits of a policy must "justify" rather than "outweigh" the costs, thus allowing room for consideration of nonquantifiable values (such as protection of privacy). He imposed a set of disclosure requirements on OIRA and made it clear that, in the event of conflict, the vice president (and ultimately the president) would make the final decision. Despite these changes, however, the larger message was that Clinton embraced, essentially wholeheartedly, Reagan's reforms and accepted many of the concerns that had led Reagan to act 12 years earlier.
Subsequent presidents have agreed. Indeed, President Obama has taken particularly ambitious steps to ensure careful balancing of costs and benefits, not least by calling for quantification of both (wherever feasible) and by insisting on scientific integrity — and hence a clear separation between political judgments and judgments about economic values. As we have noted, he went well beyond his predecessors in calling for retrospective as well as prospective analysis. As a result, the effort to reassess existing requirements, and to simplify or remove them if necessary, is now an established part of the structure of national regulation.
A RARE CONSENSUS, WITH CHALLENGES
Notwithstanding the generally bipartisan support for the approach that Reagan created and that Obama has continued, no one believes that the current set of requirements has produced the best of all possible worlds. Many people think that federal regulatory requirements are excessive, even wildly excessive, and that the OIRA process has not done nearly enough to constrain them. Many other people believe that OIRA has been too aggressive and that it has slowed down or stopped a range of regulations that would have reduced illnesses, accidents, and deaths. But the general consensus is that the process of regulatory review, with its insistent focus on costs and benefits, has made the situation a lot better than it would have been otherwise.
That consensus stems from an appreciation of the procedural and substantive innovations that Reagan inaugurated. If, for instance, the Environmental Protection Agency is issuing a rule that affects the transportation sector, or that has an impact on farmers, it is important and even indispensable for the Department of Transportation, or the Department of Agriculture, to offer its views. Inter-agency coordination that aggregates diverse information is vital, and the OIRA process helps to ensure that it occurs. This point suggests that, in the review process, OIRA itself is often a convener far more than it is the central actor. It will elicit information and take steps to ensure that dispersed knowledge is brought to bear on regulatory choices.
The process of information aggregation includes the involvement of the public as well. Since 1946, the Administrative Procedure Act has generally required a public-comment period for federal regulations. It is not always the case, however, that individual agencies ensure a robust opportunity for public comment, or that they pay close attention to the comments that they receive. To make sure the public's views are taken into account, the OIRA process is very much focused on both receiving public comments and giving them careful attention. Members of the public often have information that agencies lack. For that reason, the public-comment process serves as a valuable means of reducing the risk of error. In many cases, it helps reveal that proposed rules have overreached and that less costly approaches would be better.
The OIRA process also promotes managerial goals. It helps to ensure that agencies' regulatory policies are consistent with the views of national leaders, especially those of the president. If an agency wants to pursue a path that conflicts with high-level priorities, the process will require the conflict to come to the agency's attention. Suppose, for example, that economic growth is the president's priority, or homeland security, or implementation of the Affordable Care Act. It is not only legitimate but also extremely important to ensure that the president can manage the operations of his own team, and that he has the means to make sure that executive-branch agencies are not working at cross purposes.
One of the most important functions of the OIRA process involves economic considerations. In any period, it is valuable and even necessary to ensure that when agencies impose significant costs, they do so for good reason. In periods of economic difficulty, when stringent regulation can take an especially serious toll, the case for ensuring that benefits justify costs is even stronger. There is a detailed literature on the relationship between regulation and job loss, and while any particular assessment requires a great deal of technical work, there is no question that some regulations cost jobs. And whenever regulations impose high costs, they can force price increases, straining consumer budgets and reducing consumer purchases.
As we have noted, the calculation of costs and benefits can create serious challenges, but economists have well-established technical methods for evaluation that have long been used at the national level. An assessment of costs, or at least a range of likely costs, is often feasible, as regulators work with those who are likely to be affected (while keeping in mind that their assessments may be self-interested and therefore exaggerated). And when a regulation has purely economic benefits — for example, energy-efficiency requirements that reduce the costs of appliances — the calculation is often a fairly straightforward empirical challenge that economists are well equipped to meet.
The benefits calculation becomes more complicated when a regulation reduces risks to life and health. Questions about how many lives could be saved with a particular regulation can be difficult and controversial to answer. If the air is a bit cleaner, what are the mortality and morbidity benefits? Fortunately, scientists and economists have tools for gaining traction on such questions. Questions about how much those lives are worth are more fraught. How does one put a price tag on human lives? This is an uncomfortable and highly controversial question, but one that must be answered if cost-benefit analysis is to work. Someone must make a judgment about whether the benefits justify the costs, and this depends on an implicit assignment of monetary value: If you are willing to issue a regulation costing $200 million to save 200 people, but not to save two, you are making such an implicit assignment. In this context, there are technical methods, based on decades of research, available to help monetize risks to life and health. And where such methods are not available, OIRA and agencies have worked together to develop strategies to help decide how to resolve these seemingly intractable dilemmas.
We have noted that no one thinks that the current national regulatory system is ideal, but most informed observers agree that we are far better off with the OIRA process than we would be without it. But as we have also noted, the federal government is not the only source of costly, burdensome, unjustified requirements. State governments issue their own regulations, but because of our federal system, these are not required to undergo the same OIRA process. The impetus for creating state and local analogues must come from state and local governments.
THE STATE OF STATE AND LOCAL REGULATIONS
Although most academic and public attention has been devoted to federal regulation, there is no question that state and local controls have a significant impact on businesses and local economies — and have since at least the 19th century. As William Novak reminded us in The People's Welfare, even during the allegedly laissez-faire years of the Gilded Age, American states and cities imposed a wide array of rules on their citizens. Indeed, the federal government — in the form of the Supreme Court — often limited local attempts at regulating industry, as it did in the famous (or infamous) 1905 case Lochner v. New York. The number and scope of federal regulations rose dramatically over the 20th century, but those rules did not eliminate the tendency of localities to regulate businesses and households as well.
The strict regulatory regime that helps keep Boston's Dudley Square from becoming a bustling hub of economic activity is, unfortunately, not unique. For example, any New York grocery or restaurant that plans to sell 3,000 pounds of milk monthly must get a separate milk-dealer license. Blue laws that restrict business activity on Sunday are of ancient provenance (Emperor Constantine promulgated one in 321), and they remain in force in many parts of the country.
Transportation is highly regulated. Some states, but not others, require costly annual inspections, even for newer cars that carry little risk. In some cases, state rules on automobiles can have a national impact. California's car market is so large that its emission standards influence the national production decisions of car manufacturers.
Local businesses are also regulated because of their potential to create congestion or other nuisances. For example, installing a gas stove in New York City requires a separate fire-department inspection; using liquefied carbon dioxide or petroleum gases requires separate permits; and commercial vehicles face a separate city tax. Even when it comes to cooking at home, there are state and local rules that directly dictate behavior. Backyard barbecues are regulated in many communities.
We agree that state and local governments do vital work in tailoring laws that are specific to the needs of their local communities. It makes perfect sense that a densely populated area like New York City would need some different regulatory codes from a rural township. The problem is that state- and local-government regulations have real and sometimes far-reaching consequences that have not been foreseen because state and local regulators are usually not required to undergo any kind of review process, or coordinate with any other entity, before issuing a new rule. The lack of cost-benefit analysis and of a central review office through which to aggregate important information about potential regulations means that state and local requirements often have costly unintended effects. Local governments are often just as assiduous about regulating new housing as they are about regulating new businesses and automobiles, and these rules offer a good example of how unchecked regulation can have far-reaching effects.
Housing regulations are ancient and often motivated by obvious risks. Boston banned thatched roofs in 1630, the year of the city's founding, in the wake of an urban fire. New York's Tenement Acts first focused on urban fire risk in 1867 and then moved to mandate better ventilation and light in the interest of reducing contagious disease. Single-use, or Euclidean, zoning was also originally justified, in part, as a tool to keep noxious industrial activities away from residential dwellings.
Structural safety continues to be one reason that states and localities restrict new construction with building codes, but over the years, new rationales emerged for limiting new development. New York City's landmark zoning code of 1916 mandated setbacks to ensure the flow of light, in reaction to the massive shadow cast by the Equitable Building. After the destruction of New York's old Penn Station, a new era of landmarks preservation emerged where individual structures or even entire neighborhoods are placed under heavy supervision to limit change.
Perhaps the most potent form of current land-use regulation is the restriction of minimum lot sizes, which can range from small slivers of land to 60 acres in places like Marin County, California. The motivation for these laws varies, ranging from desires to reduce congestion or local expenditures to a wish to preserve a traditional rural character. But there is little doubt that mandating a minimum amount of acreage ensures less construction in high-demand areas, and, if the laws of supply and demand still operate, this reduced supply will also raise prices. Evidence from several studies suggests that these restrictions do, in fact, make places less affordable and thus adversely affect prospective home owners. Consequently, rules that limit construction can effectively zone poorer residents out of certain areas, creating segregation via regulation.
The crafting of local building regulations is also complicated by the web of environmentally oriented local regulations that is shaped by diverse geographies and government actors. For example, Massachusetts's rules protecting wetlands originate in state legislation, but the commissioner of the Department of Environmental Protection has significant leeway in administering and interpreting the law. Moreover, many communities in eastern Massachusetts have gone beyond the state standard and have enacted more stringent local definitions of wetlands.
Yet it is not obvious that environmentally motivated rules are actually good for the environment. Edward Glaeser and Matthew Kahn documented in a 2010 study that carbon emissions are far lower in coastal California than elsewhere in the U.S. because the climate is so mild. When California's environmentalists enact laws that prevent new local building, they are not stopping development everywhere. They are ensuring only that the development that would have been near the coast will happen elsewhere, probably in some place where commutes will be longer and energy use in air conditioning or heating will be higher. Local environmental-impact reviews do not account for this displacement because they evaluate only the impact on the local environment if the project proceeds locally, not the impact on the global environment if the project is displaced to another area.
Inadvertent environmental side effects are only one reason why critics of local land-use regulation suggest that they may go too far. Critics hold that localities are motivated by extremely parochial interests — "NIMBYism" — and consider only the interests of local property owners, not the interests of would-be buyers who are priced out because of restricted supply. Indeed, Glaeser and Bryce Ward produced evidence in a 2009 study suggesting that, even if the goal is maximizing total land values, localities are too restrictive, perhaps because of a strong bias toward the status quo. In a study published in 2005, Glaeser, Joseph Gyourko, and Raven Saks attempted to estimate the adverse social effects of building up in Manhattan and found few costs that can justify the regulatory limits placed on new construction.
If the potential downside of regulating housing is that housing becomes too expensive, one potential downside of regulating businesses is that wages will be too low. Regulating businesses that cater directly to new consumers, such as Walmart, may also end up causing local consumers to pay more for locally purchased commodities, either in the form of higher prices or time spent getting to the grocery store. Euclidean zoning rules that restrict the location of businesses represent one constraint on business formation, but localities have many business-related regulations that are less inherently spatial.
Local business regulation has been the subject of much debate, and many of the great legal battles of the late 19th and early 20th century followed the decision of states — and occasionally of lower jurisdictions — to regulate working conditions for reasons other than public safety. The first battles were over physical working conditions, the most famous being Samuel Gompers's fight against tenement-house cigar-making in New York City, which was banned by the New York State Legislature in 1883 with enthusiastic support from a young Theodore Roosevelt.
Yet while there is an extensive literature on the impact of land-use regulations, the literature on the effects of business regulations is far less developed. There is currently no domestic equivalent of the World Bank's Doing Business report, which measures the cost and number of permits required to open a firm in cities throughout the world (the U.S. Chamber of Commerce Foundation is developing such a measure; the question of objectivity is of course important). Many studies of state-level regulation have focused on the correlation between some aggregate measure of business-friendliness and economic growth, but while these results can be suggestive, it is hard to draw anything definitive from such tenuously correlated data.
More compelling evidence has come from studies that examine particular rules, such as branch-banking regulation, where it does appear that some regulations are capable of restricting economic success. One study by economist Thomas Holmes compared neighboring counties across state lines. After the passage of the Taft-Hartley Act in 1947, many states passed right-to-work laws prohibiting firms from signing "union shop" agreements, which require them to fire workers who refuse to pay at least some portion of their union dues. Holmes found that manufacturing increased substantially after 1947 on the right-to-work, pro-business side of state lines.
None of this is to suggest that any or all of these state and local rules have costs that outweigh benefits — only that they are under-evaluated. The wide disparities that exist across states suggest that either the stringent states or the more laissez-faire areas (or both) have made costly mistakes. Moreover, these rules have accreted over centuries, generally without serious cost-benefit analysis whatsoever. In many cases, rules may be sufficiently minor so that serious review isn't worth the effort, but in other cases, it is surely appropriate to weigh measurable costs against benefits — both for new legislation and for existing rules. Based on the success of the OIRA process at the federal level, there may be plausible structures that could be put in place at the local and state level that would help solve these problems.
SCOPE, DESIGN, AND STRATEGY
Given these circumstances, states could well benefit from some version of OIRA, charged with the goals of ensuring that regulations can survive some kind of cost-benefit test, promoting inter-agency coordination, and increasing accountability. But there are several important issues of scope and design that must be considered. First is whether regulatory review should occur at the state or local level; second, whether such a review process should focus on regulation, legislation, or both; and third, whether the approach of an oversight process should be minimalist or maximalist. Moreover, having a regulatory-review office at the state or local level would present distinctive challenges that require thoughtful strategies. For example, these agencies would have to figure out how to approach state-specific challenges when the federal model does not offer an answer, and how to deal with risk and uncertainty, including the quantification of difficult-to-measure costs and benefits.
While regulation occurs at both the local and state level, several factors counsel against establishing the primary centers of regulatory analysis at the local level. Few localities in the country have the depth of personnel needed to create an office of regulatory review. New York City surely can, and perhaps Chicago and Los Angeles, but most smaller areas could not easily support such an administrative task. Moreover, the similarities among smaller localities mean that highly localized offices would end up duplicating tasks, rather than building up a common bank of expertise. In addition, localized offices would necessarily tend to focus on very local costs and benefits, ignoring those that occur beyond the area's borders. Finally, since all localities face an overlapping of state and local regulations, it might be hard for a purely local office to have the perspective to consider how all these regulations fit together.
For these reasons, the natural home for any office of regulatory evaluation would be the state and not the locality. States have greater resources and a wider perspective. Moreover, since city governments all fundamentally operate under state law, there would be no inherent legal problem with requiring local regulations to follow the guidance of a state-wide officer. For example, almost all of the local zoning in the U.S. was made possible only because states enabled it through legislation, such as Massachusetts Chapter 40A. State law must of course be consulted, but it is generally within the state's power to require changes in local codes to adhere to state norms of cost-benefit analysis.
There will, of course, be many cases in which the state itself may be too small rather than too large an entity to evaluate new regulations. Regulations in New York City affect northern New Jersey and vice versa. Restricting housing supply in coastal California may affect the entire nation. Yet, even though states are surely imperfect implementers of such analysis, they are the institutions at hand, and state-level analysis is surely better than any other option.
The second question of scope is whether a state review office should have jurisdiction over regulation, legislation, or both. At the national level, OIRA's powers are limited to reviewing executive-branch regulations; it does not analyze or constrain legislative enactments. We believe that it would be deeply regrettable for any state evaluation office to limit itself to overseeing only regulations. We envision a structure in which the relevant office would have OIRA-like authority over regulations, but an advisory and consultative role in connection with legislative proposals.
Though limited, OIRA's role allows it to do a great deal, because the federal government has a large number of agencies — such as the Department of Transportation, the Department of Health and Human Services, and the Environmental Protection Agency — that have been given broad regulatory powers. To be sure, there do exist many local or state counterparts. We have already discussed the definitional control over wetlands that has been ceded to the Massachusetts Department of Environmental Protection. And at the local level, zoning administration is often the responsibility of agencies such as the New York City Department of City Planning or the Boston Redevelopment Agency. Nevertheless, much of state and local regulation comes directly from state law or local ordinance. Limiting oversight to the state executive branch would leave most local regulations unevaluated.
Given all this, the most effective structure for state regulatory review would involve a single entity, charged by state law to make assessments of executive-branch rulings, state laws, and local legislative actions. It is easy to imagine state legislation that essentially forbid executive and local regulations that failed to be approved by this agency, but the legislature is unlikely to pass a law binding itself. A feasible structure would be one in which the agency could offer a nonbinding analysis of proposed legislation. Even if it were non-binding, the analysis would undoubtedly play a significant role in legislative deliberations.
If the office were to evaluate legislative proposals, as we suggest, then not only OIRA but also the Congressional Budget Office would become a model for its operations. Over the last 40 years, the CBO has maintained a reputation for impartial integrity and sound analysis through Democratic and Republican administrations. Several features of the CBO leadership are worth highlighting: Both branches of the legislature must approve the CBO director, and either house can dismiss him unilaterally. In addition, the position is meant to be filled, at least nominally, without attention to political party. These characteristics might be considered in a state agency as well.
The CBO differs from OIRA along a major dimension that is directly relevant to our proposal. Its goal is to evaluate the budgetary impact of different forms of legislation, not overall costs and benefits, and its evaluations are constrained by very rigid rules. For example, the budget projections often assume no further changes in tax laws, even when those changes are highly likely. Cost-benefit analysis, as opposed to projecting budgets, is typically more difficult and more holistic. We suspect that CBO personnel may indeed be quite helpful in providing support and assistance for nascent state-level evaluation groups, but it needs to be recognized that the task is somewhat different.
As for determining the third question of scope and design, state governments are very diverse in terms of their current capacities and needs, so each state will need to decide for itself how much a regulatory-review office should do and how active it should be. As a starting point, states would gain a great deal by adapting some version of the original Reagan model, with its two simple components: an oversight process and a substantive requirement of cost-benefit balancing. For both components, minimalist and maximalist versions can easily be imagined.
A minimalist version would involve the creation of a small group or office whose task would be to scrutinize rules and work with the agencies, or others within the executive branch, to make some kind of judgment about whether the rules make economic sense. Under the minimalist approach, the small office, with a commitment to some cost-benefit analysis, could help to deter unjustified requirements and ensure that officials are focused on what really matters, which is the human consequences of their actions.
A maximalist version would be more formal, and would involve something closer to what the national government now does. States and cities could create their own offices of regulatory affairs, consisting of people trained in economics and able to produce, or at least to scrutinize, careful analysis of costs and benefits. For certain rules — perhaps those anticipated to cost more than $25 million annually — relevant officials might be required to produce a full-scale regulatory-impact analysis, akin to what the federal government now produces. The length and depth of the analysis would be a function of the cost and importance of the particular regulation.
We much prefer the maximalist model, because it is most likely to prevent harmful outcomes, but we recognize that some states will lack both the will and the capacity for it — and a degree of experimentation is highly desirable. Some states might do well to try the minimalist model as a "pilot" program and then build on it later if it works well.
The strategies used by state review agencies should also draw heavily from the federal model. For many situations, the approach would be essentially identical to that used at the federal level and should be based on the same tools and techniques. OMB Circular A-4, supplemented by various guidance documents (including a simple checklist), outlines those tools and techniques. Air-pollution controls, for example, might borrow directly on what is now done by the EPA, and highway-safety regulation could draw on the longstanding work of the Department of Transportation. But some domains of state and local regulation present special challenges.
Consider a decision to raise the minimum lot size in a community for new construction from one to two acres. The potential beneficiaries from this rule would be the community members who benefit from lower levels of congestion, possibly a nice view when walking or driving, and possibly a lower financial burden on their local public services. These local benefits could be estimated directly or inferred indirectly through land values. A long intellectual tradition argues that local amenities, and the associated tax obligations, are captured in land values, so an assessment of the link between new construction and property values provides one means of assessing the upside of the regulations.
While community members may benefit, there are significant downsides. The new regulation will take options away from owners who would like to subdivide their property, will potentially reduce tax revenues, and will potentially add to traffic congestion if restricting development close to the city center pushes building farther away. There are also the potential local environmental benefits of lower density that need to be offset against the environmental costs of pushing building projects farther away from where people work. These calculations are not simple, but they are doable, and once an agency has developed a template for these calculations, they can easily be repeated in multiple jurisdictions.
A state-wide cost-benefit agency will not have the resources to examine every regulation. It makes sense, therefore, to have a minimum economic threshold required for evaluation, based on a preliminary investigation of the regulation, just as there is at the national level, with the threshold of $100 million in economic impact. Unlike OIRA, however, the state threshold should also be based on repetition. One community's change in zoning laws might not meet a minimum lower threshold on costs, but when repeated hundreds of times in different areas, the threshold would certainly be reached. It makes more sense to have a minimum threshold based on the total number of cases for which the developed methodology is useful.
The last challenge to be addressed is how to deal with risk and uncertainty. States will have their own preferences about these issues, and there are ways to design rules around those preferences. For example, a state that wished to err on the side of elected local officials might well have a rule that allowed regulations to go ahead as long as estimated costs were no more than 10% higher than benefits. A state that was particularly fearful of regulation might set the bar at parity between costs and benefits or even require benefits to exceed costs by 10%.
These statements all concern the average or expected levels of costs and benefits, and in many cases there will be some risk that real costs or benefits will differ widely from the projection. It is certainly possible to adopt approaches that account for risk, perhaps by giving more weight to particularly bad outcomes than to particularly good outcomes.
There are other complex questions to handle as well, such as how a state agency should deal with hard-to-quantify costs and benefits, like environmental benefits that flow to animals or the protection of religious liberty allegedly associated with blue laws. The national government has a great deal of experience with non-quantifiable benefits. The standard approach involves "break-even analysis," by which agencies ask: How much would the non-quantified benefits have to be to justify the costs?
Suppose, for example, that the EPA is considering a water-pollution rule that would cost $1 billion and deliver $200 million in quantifiable benefits (say, in terms of human health and improved recreation), but it would also produce non-quantifiable benefits (in terms of, say, ecological benefits). Relevant questions would include: How many bodies of water? What kinds of improvements? What would those improvements actually achieve? Would they help human beings, and if so, how? Unless these questions have compelling answers, an expenditure of $1 billion would not be easy to defend, with the quantifiable benefits in the range of merely $200 million.
Suppose, however, that, once we investigate the details, we find that the rule would achieve a great deal. Perhaps it would, for example, protect a very large number of water bodies and do a great deal for them, with a wide range of aesthetic and ecological benefits (including the protection of fish and wildlife). Once these questions are explored, we may have enough to justify a serious discussion. If, by contrast, the number of water bodies is relatively small, and the benefits for them would not be great, such a significant expenditure would not be easy to justify. A similar analysis could be applied to a wide range of state regulations with benefits that are difficult or impossible to quantify.
In some cases, the non-quantifiable becomes quantifiable if the state has alternative means of achieving the same ends. For instance, a specific wildlife goal might be achievable either by restricting development in a denser and more economically active section of a state or by increasing the size of a state park in a less dense area. The price of expanding the state park provides a reasonable means of evaluating the environmental benefits of the regulation.
To be sure, some problems present exceedingly difficult challenges, for which economic analysis cannot have the final word. But the federal experience suggests that these challenges should not be overstated, and that quantitative analysis can operate as an indispensable check on decisions that might otherwise be based on anecdotes, dogmas, intuitions, or interest-group pressures.
TOWARD REGULATORY REVIEW
No one should deny that regulation can do a great deal of good, not least by lengthening human lives and improving the environment. But if it is poorly designed, regulation can jeopardize a wide range of social goals by reducing economic growth, harming small businesses, and squelching innovation. For all its limitations, cost-benefit analysis is the most disciplined way to test whether regulatory requirements are a good idea. At the national level, its track record establishes its frequent ability to separate the wheat from the chaff — to make the case for sensible mandates and to decrease the likelihood that regulatory initiatives will be misdirected or have unforeseen consequences.
It is remarkable but true that nearly 35 years after President Reagan established what has proved to be an enduring national commitment to cost-benefit analysis, state and local governments have shown inadequate and at most sporadic efforts in following the federal government's lead. Those inadequate efforts are especially surprising in light of the fact that state and local governments pass laws and ordinances that are so often costly and burdensome for their local economies — and have consequences that can reach beyond city limits and state lines. The federal government has provided an imperfect but proven model for regulatory review. Given the harm that ill-conceived regulations can inflict on a struggling economy, there is no reason for state and local governments to wait any longer to set up their own review agencies.