Tag Archives: regulatory process

Does Anyone Know the Net Benefits of Regulation?

In early August, I was invited to testify before the Senate Judiciary subcommittee on Oversight, Federal Rights and Agency Action, which is chaired by Sen. Richard Blumenthal (D-Conn.).  The topic of the panel was the amount of time it takes to finalize a regulation.  Specifically, some were concerned that new regulations were being deliberately or needlessly held up in the regulatory process, and as a result, the realization of the benefits of those regulations was delayed (hence the dramatic title of the panel: “Justice Delayed: The Human Cost of Regulatory Paralysis.”)

In my testimony, I took the position that economic and scientific analysis of regulations is important.  Careful consideration of regulatory options can help minimize the costs and unintended consequences that regulations necessarily incur. If additional time can improve regulations—meaning both improving individual regulations’ quality and having the optimal quantity—then additional time should be taken.  My logic behind taking this position was buttressed by three main points:

  1. The accumulation of regulations stifles innovation and entrepreneurship and reduces efficiency. This slows economic growth, and over time, the decreased economic growth attributable to regulatory accumulation has significantly reduced real household income.
  2. The unintended consequences of regulations are particularly detrimental to low-income households— resulting in costs to precisely the same group that has the fewest resources to deal with them.
  3. The quality of regulations matters. The incentive structure of regulatory agencies, coupled with occasional pressure from external forces such as Congress, can cause regulations to favor particular stakeholder groups or to create regulations for which the costs exceed the benefits. In some cases, because of statutory deadlines and other pressures, agencies may rush regulations through the crafting process. That can lead to poor execution: rushed regulations are, on average, more poorly considered, which can lead to greater costs and unintended consequences. Even worse, the regulation’s intended benefits may not be achieved despite incurring very real human costs.

At the same time, I told the members of the subcommittee that if “political shenanigans” are the reason some rules take a long time to finalize, then they should use their bully pulpits to draw attention to such actions.  The influence of politics on regulation and the rulemaking process is an unfortunate reality, but not one that should be accepted.

I actually left that panel with some small amount of hope that, going forward, there might be room for an honest discussion about regulatory reform.  It seemed to me that no one in the room was happy with the current regulatory process – a good starting point if you want real change.  Chairman Blumenthal seemed to feel the same way, stating in his closing remarks that he saw plenty of common ground.  I sent a follow-up letter to Chairman Blumenthal stating as much. I wrote to the Chairman in August:

I share your guarded optimism that there may exist substantial agreement that the regulatory process needs to be improved. My research indicates that any changes to regulatory process should include provisions for improved analysis because better analysis can lead to better outcomes. Similarly, poor analysis can lead to rules that cost more human lives than they needed to in order to accomplish their goals.

A recent op-ed penned by Sen. Blumenthal in The Hill shows me that at least one person is still thinking about the topic of that hearing.  The final sentence of his op-ed said that “we should work together to make rule-making better, more responsive and even more effective at protecting Americans.” I agree. But I disagree with the idea that we know that, as the Senator wrote, “by any metric, these rules are worth [their cost].”  The op-ed goes on to say:

The latest report from the Office of Information and Regulatory Affairs shows federal regulations promulgated between 2002 and 2012 produced up to $800 billion in benefits, with just $84 billion in costs.

Sen. Blumenthal’s op-ed would make sense if his facts were correct.  However, the report to Congress from OIRA that his op-ed referred to actually estimates the costs and benefits of only a handful of regulations.  It’s simple enough to open that report and quote the very first bullet point in the executive summary, which reads:

The estimated annual benefits of major Federal regulations reviewed by OMB from October 1, 2002, to September 30, 2012, for which agencies estimated and monetized both benefits and costs, are in the aggregate between $193 billion and $800 billion, while the estimated annual costs are in the aggregate between $57 billion and $84 billion. These ranges are reported in 2001 dollars and reflect uncertainty in the benefits and costs of each rule at the time that it was evaluated.

But you have to actually dig a little farther into the report to realize that this characterization of the costs and benefits of regulations represents only the view of agency economists (think about their incentive for a moment – they work for the regulatory agencies) and for only 115 regulations out of 37,786 created from October 1, 2002, to September 30, 2012.  As the report that Sen. Blumenthal refers to actually says:

The estimates are therefore not a complete accounting of all the benefits and costs of all regulations issued by the Federal Government during this period.

Furthermore, as an economist who used to work in a regulatory agency and produce these economic analyses of regulations, I find it heartening that the OMB report emphasizes that the estimates it relies on to produce the report are “neither precise nor complete.”  Here’s another point of emphasis from the OMB report:

Individual regulatory impact analyses vary in rigor and may rely on different assumptions, including baseline scenarios, methods, and data. To take just one example, all agencies draw on the existing economic literature for valuation of reductions in mortality and morbidity, but the technical literature has not converged on uniform figures, and consistent with the lack of uniformity in that literature, such valuations vary somewhat (though not dramatically) across agencies. Summing across estimates involves the aggregation of analytical results that are not strictly comparable.

I don’t doubt Sen. Blumenthal’s sincerity in believing that the net benefits of regulation are reflected in the first bullet point of the OMB Report to Congress.  But this shows one of the problems facing regulatory reform today: People on both sides of the debate continue to believe that they know the facts, but in reality we know a lot less about the net effects of regulation than we often pretend to know.  Only recently have economists even begun to understand the drag that regulatory accumulation has on economic growth, and that says nothing about what benefits regulation create in exchange.

All members of Congress need to understand the limitations of our knowledge of the total effects of regulation.  We tend to rely on prospective analyses – analyses that state the costs and benefits of a regulation before they come to fruition.  What we need are more retrospective analyses, with which we can learn what has really worked and what hasn’t, and more comparative studies – studies that have control and experiment groups and see if regulations affect those groups differently.  In the meantime, the best we can do is try to ensure that the people engaged in creating new regulations follow a path of basic problem-solving: First, identify whether there is a problem that actually needs to be solved.  Second, examine several alternative ways of addressing that problem.  Then consider what the costs and benefits of the various alternatives are before choosing one. 

Do We Need Greater Congressional Oversight of Agency Rulemaking?

Katherine McFate, president of the Center for Effective Government, writes in the Hill that all regulations are based on congressional law, implying that efforts aimed at greater oversight of agency rulemaking are unnecessary. Technically, she is correct – agencies cannot regulate unless they are authorized to do so by congressional statutes. But her assertion is highly misleading. In fact, agencies have considerable discretion to determine policy and to publish rules that fit their as opposed to Congress’s agenda. Thus, Congress is fully justified in its efforts to push for greater agency accountability.

Scholars have long realized that the traditional rulemaking model (or the “transmission belt” model as Richard Stewart called in his seminal article) in which Congress determined policy through legislation and agencies simply filled in the details was far from reality. While constrained by congressional statutes, agencies nonetheless can substantively shape the policies within their jurisdiction.

Agencies have two sources of power in the rulemaking process: the first mover advantage and expertise. Congress over time delegated considerable policymaking powers to agencies through broad open-ended statutes, which meant that agencies did not need to seek congressional approval in order to regulate. In many cases, they can point to existing statutes as a source of their authority. Rather than initiate policy, Congress ends up reacting to the bureaucracy’s regulatory actions. Yet, given how difficult it is for Congress to agree on any legislation, it may be an uphill battle for Congress to overturn a regulation, letting agency decision stand by default.

Expertise is the second source of agency power. Congress defers to agency expertise on many complex regulatory issues. However, agencies engage in what Wendy Wagner called “the science charade” – masking policy decisions as matters of science. As she explains in her article and an edited volume, scientific analysis often drives policy decisions. Through selective use of evidence or assumptions, agencies can push the scientific analysis towards the answer that would yield their preferred policy alternative.

The EPA’s and DOE’s use of social cost of carbon (SCC) in their rulemaking estimates demonstrates the point. The SCC is an estimate of economic damages caused by greenhouse gas emissions. Agencies use the SCC to estimate the benefits of rules aimed at reducing greenhouse gas emissions and consequently to decide whether the rules’ benefits justify the costs. Higher SCC estimate would justify more expansive and costly regulations.

Even though the agencies claim that they derived the SCC through scientific analysis, critics point to policy choices embedded within the analysis that pushed the estimated cost higher. For example, the agencies chose to estimate global rather than domestic impacts of carbon. Similarly, they omitted from their analysis the recent scientific literature, which pointed to a lower impact of greenhouse gases on climate. These and other choices pushed the SCC higher (almost double the previous estimate), allowing agencies to push for more stringent and costly regulations.

Despite the major policy impact of the SCC’s use in rulemaking, the agencies did not have to consult Congress. They could chose to use the SCC estimates under the powers already delegated to them, even in the face of stiff opposition from Congress. In the meantime, congressional efforts to reassert its authority on the major environmental policy issue have stalled. The GOP-led House passed a bill that would prevent the EPA from factoring in the SCC in its economic analysis. Yet, the measure’s fate in the Senate is uncertain and it would still face the presidential veto.

Contrary to McFate’s assertion, agencies do not simply implement congressional policy. As the SCC example demonstrates, agencies can drive major policy decisions without congressional approval. Thus, Congress needs better tools for more effective oversight of agency regulations.

Why Regulations Fail

Last week, David Fahrenthold wrote a great article in the Washington Post, in which he described the sheer absurdity of a USDA regulation mandating a small town magician to develop a disaster evacuation plan for his rabbit (the rabbit was an indispensible part of trick that also involved a hat). The article provides a good example of the federal regulatory process’ flaws that can derail even the best-intentioned regulations. I list a few of these flaws below.

  1. Bad regulations often start with bad congressional statutes. The Animal Welfare Act of 1966, the statute authorizing the regulation, was meant to prevent medical labs from using lost pets for experiments. Over time, the statute expanded to include all warm-blooded animals (pet lizards apparently did not merit congressional protection) and to apply to zoos and circuses in addition to labs (pet stores, dog and cat shows, and several other venues for exhibiting animals were exempt).The statute’s spotty coverage resulted from political bargaining rather than the general public interest in animal welfare. The USDA rule makes the statute’s arbitrariness immediately apparent. Why would a disaster plan benefit circus animals but not the animals in pet stores or farms? (A colleague of mine jokingly suggested eating the rabbit as part of an evacuation plan, since rabbits raised for meat are exempt from the regulation’s requirements).
  2. Regulations face little oversight. When media reported on the regulation’s absurdity, the USDA Secretary Tom Vilsack ordered the regulation to be reviewed. It seems that even the agency’s head was caught off guard by the actions of his agency’s regulators. Beyond internal supervision, only a fraction of regulations face external oversight. Of over 2600 regulations issued in 2012, less than 200 were subject to the OMB review (data from GAO and OMB). Interestingly, the OMB did review the USDA rule but offered only minor revisions.
  3. Agencies often fail to examine the need for regulation. In typical Washington fashion, the agency decided to regulate in response to a crisis – Hurricane Katrina in this case. In fact, the USDA offered little more than Katrina’s example to justify the regulation. It offered little evidence that the lack of disaster evacuation plans was a widespread problem that required the federal government to step in. In this, the USDA is not alone. According to the Mercatus Center’s Regulatory Report Card, which evaluates agencies’ economic analysis, few agencies offer substantial evidence justifying the need for promulgated regulations.
  4. Agencies often fail to examine the regulation’s effectiveness. The USDA’s plan to save animals in case of a disaster was to require owners to draw up an evacuation plan. It offered little evidence that having a plan would in fact save the animals. For example, the magician’s evacuation plan called for shoving the rabbit into a plastic bag and getting out. In the USDA’s view, the magician would not have thought of doing the same had he not drawn up the evacuation plan beforehand.
  5. The public has little influence in the process. By law, agencies are required to ask the public for input on proposed regulations. Yet, small businesses and individual consumers rarely have time or resources to keep an eye on federal agencies. In general, organized interests dominate the commenting process. The article describes the magician’s surprise to learn that he was required to have a license and a disaster evacuation plan his rabbit, even though the regulation was in the works for a few years and was open for public comments for several months. Most small businesses, much like this magician, learn about regulations only after they have passed.
  6. Public comments are generally ignored. Most public comments that the USDA received argued against the rule. They pointed out that it would impose substantial costs on smaller businesses. The agency dismissed the comments with little justification. This case is not unique. Research indicates that agencies rarely make substantial changes to regulations in response to public comments.

Do “Indirect Effects” of Regulation Matter to Real People?

Congressional regulatory reformers recently caught criticism from advocacy groups for introducing legislation that would require federal regulatory agencies to analyze the “indirect effects” of proposed regulations. The only thing I’d criticize the reformers for is poor word choice.

The very term “indirect effects” suggests that they’re talking about something theoretical, inconsequential, and unimportant to the average citizen. But to economists, the indirect effects of a regulation are often the effects that touch the average citizen most directly.

Consider airport security, for example. The Department of Homeland Security (DHS) recently sought public comment on its decision to deploy Advanced Imaging Technology scanners instead of metal detectors at airports. The direct costs of this decision are the extra cost of the new machines, the electricity to run them, and the personnel to staff them – which airline passengers pay for via the taxes and fees on airline tickets. Those are pretty obvious costs, and DHS dutifully toted up these costs in its analysis of its proposed rule.

Less obvious but potentially more important are the other, indirect costs associated with airport security. Passengers who decline to walk through the new machines will receive additional pat-downs. This involves a cost in terms of time (which DHS acknowledges) and potentially diminished privacy and human dignity (which DHS does not discuss). The now-classic phrase “Don’t touch my junk” aptly summarizes one passenger’s reaction to an indirect effect of security regulation that touches passengers quite directly.

But that does not exhaust the list of significant, predictable, indirect effects associated with airport security regulation. The increased delays associated with enhanced, post-9/11 security measures prompted some travelers to substitute driving for flying on short trips. An article by Garrick Blalock, Vrinda Kadiyali, and Daniel H. Simon published in the November 2007 Journal of Law & Economics estimates that post-9/11 security measures cost the airline industry $1.1 billion in lost revenue in the fourth quarter of 2002. Driving is also riskier than flying. Blalock et. al. estimated that the security measures were associated with 129 additional highway deaths in the fourth quarter of 2002.

I’m all for making air travel as safe as possible, but I’d like to see it done smartly, with a minimum of hassle and a maximum of respect for the flying public who pays the bills. A full accounting of the indirect effects of airport security might just prompt policymakers to consider whether they are pursuing regulatory goals in the most sensible way possible.

Unfortunately, airport security is not an isolated example. Data from the Mercatus Center’s Regulatory Report Card reveal that for about 40 percent of the major regulations proposed by executive branch agencies between 2008 and 2012, the agencies failed to conduct any substantial analysis of costs that stem from the proposed regulation’s effects on prices or on human behavior – two classic types of indirect effects.

This won’t do. Telling federal agencies they do not need to understand the indirect effects of regulation is telling them they should proceed in willful ignorance of the effects of their decisions on real people. The reformers have a good idea here – even if it has a misleadingly boring name.