Getting State Economic Analysis Right

In a recent study, “Principles for Constructing a State Economic Analysis Unit,” James Broughel and Patrick McLaughlin outline how economic analysis can help state regulatory agencies adhere to principles of high-quality regulation, such that rules are more likely to achieve significant benefits for the public at a reasonable cost. Their paper emphasizes that the institutions tasked with producing analysis must be sufficiently insulated from politics, and analysis must be timely for it to be useful and effective.

In this interview, Chad Reese sits down with Mercatus senior research fellow James Broughel to discuss some common questions that have arisen in response to this research.

In your paper, you and Patrick McLaughlin discuss the importance of maintaining independence from politics. Why is an independent economic analysis unit important? Don’t we want oversight from judges and legislators in some cases?

Analysis should be independent of politics, but that’s not the same thing as independent of oversight. Oversight is very important, in fact. Unfortunately, there is no way to completely insulate analysis from politics. Analysts come to the table with their own sets of beliefs and personal biases, just like everyone else. However, some institutions are more likely to achieve greater objectivity and freedom from politics than others.

Where do you think these units ought to be located in order to obtain this independence?

One way to ensure a basic level of objectivity is to separate the role of regulator from the role of analyst. This can be done within an agency by setting up an analysis office that is independent of the program offices that write and implement regulations. The Federal Communications Commission recently set up a structure along these lines.

Alternatively, analysis responsibilities could be removed from agencies altogether. For example, the state of Virginia has tasked the Planning, Evaluation, and Regulation division within the state Department of Planning and Budget with producing economic analysis for rules.

On the other hand, there is no reason why this analysis needs to be produced in the executive branch of state government or even within the government. In fact, removing analysis functions from the executive branch could help insulate analysis from the political influence of the governor. Federal analysis offices like the Congressional Budget Office, the Congressional Research Service, and the Government Accountability Office are examples of legislative-branch agencies that are generally trusted to be objective.

States might also consider the model in Washington State. The Washington State Institute for Public Policy (WSIPP) is a public research group, created by the state legislature, that is periodically charged with producing benefit-cost analysis of existing state programs. The WSIPP maintains a close connection with a public university in the state, Evergreen State College, which highlights another option available to state governments. States can utilize academic experts at state universities. This could be done, for example, if a legislature were to commission evaluations of particular rules or programs from academics (similar to how the WSIPP works), pay for sabbaticals for professors in order for government programs to be analyzed, or tie funding for state universities to conditions requiring economic analysis be produced for the government.

There seems to be concern among some legislatures about the potential cost of setting up a new analysis unit? Is it possible to make the economic analysis unit both effective and inexpensive? And who has the specialized knowledge to run or work for the unit?

In the policy paper Patrick McLaughlin and I wrote together, the word “unit” was used somewhat loosely. Analysis needs to be done by individuals with relevant expertise, but that doesn’t necessarily mean a new “unit” needs to be established. Many states already have institutions where these responsibilities could sensibly be housed without creating a new bureaucracy.

For example, Pennsylvania has an Independent Regulatory Review Commission, as well as an Independent Fiscal Office. Either of these existing entities would be a logical place to house analysis responsibilities.

That said, states generally lack the personnel capable of doing this kind of work, which requires technically trained experts with a background in benefit-cost analysis. Either some new workers will have to be hired, or the work will have to be contracted from experts outside the government.

But it’s important to remember that the benefits of analysis can easily pay for these extra costs. If an analyst finds a way to implement a multi-million dollar rule more inexpensively, the savings could easily pay for the analyst’s entire lifetime salary many times over. When we are dealing with rules that impact a state’s entire economy, small improvements add up quickly. Paying for a few extra workers is cheap in comparison.

Currently, many states have analysis requirements in place, but they often don’t invest resources so that the analysis is done seriously. If states don’t rely on trained experts, the quality of analysis will be poor. Additionally, the more resources invested, the more programs can be analyzed. So there are clear tradeoffs to consider. The decision about how much to invest is going to vary by state depending on resources available.

What powers should the unit have? Should it be allowed to, for example, send regulations back to agencies or turn them over to the legislature for approval? What if the role of the unit is scaled back such that it just reviews regulations and their accompanying analysis but doesn’t independently produce the analysis?

Poor analysis probably should be grounds for a regulation to be returned to the regulator so it can do a better job. Regulatory agencies are delegated the power to write regulations based on their expertise. If they fail to utilize that expertise, then there is no particular advantage to having regulators write law as opposed to legislators, and there are lots of disadvantages since regulators are less accountable to the public than elected officials.

That said, it’s not clear to me that the analysis unit is the best entity to return rules. The Office of Information and Regulatory Affairs at the federal level, known as OIRA, reviews regulations and their accompanying analysis and can return regulations to agencies to be reworked if the analysis is deemed sufficiently poor in quality or if the agency has overlooked certain critical information. However, rules aren’t returned by OIRA very often, even though most regulations have either no analysis or very incomplete analysis, accompanying them.

That’s a real problem, and it suggests that just having the unit review regulations, even while giving it the power to return regulations, may not be sufficient. Analysis production should be independent of rule-writing and enforcement activities, and a better place to house the review function might be in the courts. The courts are already tasked with vacating regulations when certain legal criteria are not met. Furthermore, regulatory agencies are very attuned to the possibility of court challenges, so if they see poor analysis as opening the door to a rule being overturned by a judge, they will take analysis much more seriously. In fact, Mercatus research has found that oversight from the courts corresponds with higher quality analysis.

A meaningful peer review process could also help ensure analysis is high quality, but it might be better for scrutiny to come from experts outside the government—for example, at academic institutions. That way reviewers aren’t overseen by political personnel. A problem with OIRA review at the federal level is that OIRA is overseen by the White House and hence its reviews can be influenced by politics.

Should the unit produce analyses of existing regulations or just new ones that have yet to go into effect? What about emergency regulations implemented on short notice due to safety?

To some extent, this will depend on resources available, but ideally, analysis should occur for both new and existing regulations. Historically, benefit-cost analysis has been applied to new regulations and only very rarely to existing regulations. This is not ideal, of course. Ideally, we want to be able to learn from past experience and apply those lessons as we go about crafting new policies, such that policymaking works in an iterative process.

Additionally, analysis can complement other review efforts, such as periodic reviews, legislative reviews, sunset reviews, or red tape cutting efforts that states have in place. These kinds of efforts are often hindered by the fact that decision makers lack high-quality information about how well rules are working or are likely to work. Without reliable information, risk-averse policymakers often choose to maintain the status quo, and as a result, many regulatory reform efforts don’t work as well as they should. It’s not surprising President Trump has also made analysis an important part of his regulatory reduction efforts.

As far as emergency rules are concerned, these rules, when they are in response to a genuine emergency—a natural disaster, an act of terrorism, an outbreak of disease, a war, and so on—should be allowed to go into effect without analysis. But these rules should not remain in effect long term. Emergency regulations should be temporary and should have to be replaced with permanent rules that go through ordinary rulemaking produces, once the emergency has subsided. Rulemaking procedures establish needed checks and balances, one of which should be the scrutiny of economic analysis.

Would it be beneficial to have the unit only look at regulations that have a certain level of impact? If so, what should that level be? Who determines if the level has been reached?

It will be hard to subject all rules to meaningful benefit-cost analysis. Some system of prioritization is probably needed to determine which rules get scrutiny. For existing rules, one option is to have a legislative committee make individual requests for analysis of specific rules or programs. Such a process might be driven largely by constituent concerns, for example. This approach has an advantage in that it is democratic and is responsive to the concerns of the public. However, it will not always guarantee that all high-impact rules are scrutinized.

Alternatively, a trigger mechanism could be set up to determine which rules get reviewed. Triggers are usually based on economic impact, and several states already have triggers for some regulatory procedures, which could serve as a model.

For example, Wisconsin requires legislative approval for regulations costing $10 million or more over any two-year period. At the federal level, executive branch rules with an annual impact over $100 million require benefit-cost analysis, though presumably the threshold should be set lower for state regulations.

The analysis unit, or a similar body like a fiscal note office, is a logical entity to determine whether a trigger is met. Alternatively, the regulating agency could make an initial determination about whether a rule meets a threshold, but some other entity should probably have to certify the agency’s determination is correct.

What happens after the analyses are produced? How do we prevent agencies/executives/lawmakers from ignoring the findings?

Ultimately, the decision of whether and how to regulate is a political one, not purely an analytic one. So long as the analysis is objective, transparent, and available in a timely manner, then it is doing its job. Legislators and regulators are free to ignore analysis if they so choose, but they should explain their decisions when they do so, and ultimately, if policymakers ignore crucial evidence that is pertinent to a rulemaking, it is the job of voters to hold them accountable.

What’s most important is that the relevant tradeoffs are identified and reported in a transparent manner, so that decision makers and the public are aware of these tradeoffs before regulations go into effect or are allowed to remain in place.

Photo credit: Pennsylvania General Assembly