The Incomparable Jerry Ellig

Abstract

Jerry Ellig was a unique character and a great economist. He believed in one thing, using economic analysis to help solve problems. He became an expert at Regulatory Impact Analyses and how they helped governments to choose the best option to do just that, all the while recognizing the problems that government has with necessarily much less information than markets. He believed in holding governments to account for achieving results including periodic lookbacks to see what they were doing. What was great about Jerry is that he had fun doing all of this both on the job and at his beloved Tiki bars.

1. Introduction

Jerry Ellig was motivated to solve problems. The big one he took on was the Administrative State, the federal regulations that are smothering productive society. He took on one piece of it at a time, but time ran out on him, and he did not finish. He might have left the hardest part for the next generation – fixing Congress.

Over his career, Jerry studied regulatory agencies and, in time, came to the conclusion that the main problem with those agencies was that they were not accountable for poor results. He left a blueprint for anyone actually interested in how to fix the problem. Unfortunately, he ran out of time on how we can generate interest in holding agencies accountable. (Ellig & Fike, 2014).

I first met Jerry right before I retired from the Food and Drug Administration in 2007. He was one of several scholars I interviewed for a position I was applying for running the Regulatory Studies Group at the Mercatus Center. They were suspicious of what I, a career bureaucrat with more than 30 years in government, could bring to the Center. Jerry asked me, as his incoming supervisor, whether I thought it was necessary to “approve” papers before they went out. My reply was that I thought my job was to offer helpful suggestions, but scholars’ work was their own. He smiled and I knew I had at least one person interested in working with me.

Over time at Mercatus, I came to understand that Jerry’s main issue with federal regulations was that if regulators were not accountable for achieving results, it was unlikely that they would achieve “outcomes” – results that help people. He wanted to push back on those who simply believed in the Wilsonian notion that having a professionally trained bureaucracy of experts would result in outcomes in the interest of the common good, that is, what he called a “Faith-Based” philosophy (a riff on then President George Bush’s “faith-based initiative”), placing faith that regulators will achieve good results, instead of evidence that they do so.

Unfortunately, a Faith-Based system is precisely what we have now. The federal government started assembling the rules (Federal Register) in 1936 and agencies still promulgate huge rules with insufficient evidence that, once in place, they will work and rarely follow up to see whether or not they are working. Along with Jerry authors such as Robert Crandall, Maurice Mctigue, Bruce Yandle, and many others have created a system that, if followed, would go a long way toward remedying those problems. (Ellig, 2013).

This system has four primary steps:

1. 1) Establish Goals – express clearly the promises that regulatory agencies make to the American people in their strategic documents, particularly those contained in the Government Performance and Results Act (GPRA).

2. 2) Analyze Expectations – these are the expected benefits and costs of individual regulations found in Regulatory Impact Analyses (RIAs).

3. 3) Measure Results – the actual outcomes after regulations have been put in place and enforced. They can be found in retrospective reviews and other documents.

4. 4) Ensure Accountability – The actions Congress or other federal oversight bodies take as a result of the differences between goals and results.

Much of the huge volume of work that has Jerry as a sole or co-author falls into one of these four steps.

2. Goals

Jerry worked on goals with Maurice McTigue, a former New Zealand administrator and diplomat. While many of the people he worked with just criticized the government, both Jerry and Maurice worked on making it better. In 2011, Jerry and Maurice, along with OMB employee Henry Wray, published Government Performance and Results: An Evaluation of GPRA’s First Decade. (Ellig et al., 2012) They were interested in making sure government agencies clearly expressed their goals with measurable outcomes. Outcomes are different than inputs or outputs, which are often reported as successes by agencies.

One example of this is children learning to read. For example, the Department of Education could have a goal to get every student within 2 years of their grade level in reading. Buying them books would be an input. An output might be each student spending 1 h per week reading aloud. But the outcome that matters is actual reading skill. Agencies should be expected to list outcome goals and report on progress in achieving them, even if they take years to accomplish. They wrote, “we believe citizens are better off when government management and budget decisions are based on actual outcomes produced by programs and policies” (Ellig et al., 2012).

While they were writing the book, Maurice and Jerry began a program at Mercatus, the Government Accountability Project, in which they measured the success of agencies in listing outcome goals with measurable results and then reporting those results. (McTigue et al., 2009) This would include what they learned each year from successes and failures. The idea was, that once the legislative body that exercises oversight, Congress, had access to results they would adjust their missions and their budgets accordingly. We’ll get back to that.

3. Expectations – regulatory impact analyses

Jerry worked tirelessly to improve regulatory impact analyses (Ellig, 2016). RIAs are expected to succinctly state the problem they are trying to solve, estimate the benefits (outcomes) and costs of several regulatory options, and then choose the best option. Like the risk assessments that often underlie these analyses, there is a great deal of uncertainty in estimating both benefits and costs and, while the regulatory option that has the greatest difference between benefits and costs is the economically preferred solution, decision makers are often guided by other than utilitarian concerns. Nevertheless, they should explain their decision including whether or how they used the RIA to make their decision.

It was interesting to work at the Mercatus Center at George Mason University because, along with Jerry and me, many of the economists working there had studied under Nobel Prize Winner James Buchanan. Buchanan pointed out that people that worked for the government were not Saints, they were ordinary people with the same motivations as the rest of us. He called it Politics without Romance (1979). A lot of the GMU economists did not like benefit–cost analysis because it involves quantifying people’s preferences, through what they will pay for and what they say they will pay for but true preferences are nearly impossible to know for many government actions. While Jerry acknowledged the difficulty, he also argued that even given how uncertain such estimates were, without the attempt to understand impacts, regulations were based only on the preferences of government decision makers and those whose lobbying efforts were most effective. He also showed that regulations supported by higher-quality RIAs were also more likely to survive litigation (Bull & Ellig, 2018). Again, he did not just criticize the administrative state, he understood that as long as we are going to have one, it is important to make it better.

Realizing that creating better RIAs is a crucial step toward getting better regulations. Jerry was responsible for starting the Mercatus Centers’ Regulatory Report Card (2008 to 2013) (Ellig, 2016). Scholars working on the Report Card analyzed individual regulations from a list of criteria that Jerry created. For example, going back to GPRA, each RIA had to discuss the problem it was attempting to solve, such as a failure in the market.

Grades were assigned to each element of the RIA including whether it contained a sufficiently broad number of alternatives to solve the problem, how much of the problem, such as a reduction in risk each alternative would obtain and the costs of doing so, and a full explanation of the assumptions and the data in the analysis. All these requirements had been spelled out in various administrative documents, including Executive Orders, since 1980 (see, for example, Reagan 1981; Clinton, 1993). Jerry went further, however. He insisted on grading the agencies on whether they had explained how they had used the analysis to make a decision. He also wanted grades on whether they included measures by which they could track the regulations’ effectiveness in the future.

As Jerry and I began the Report Card, we met with OIRA staff and, although they seemed to like the Report Card, they were very concerned about what was missing. Because so many federal regulations are based on science (or engineering), agency economists must rely on experts within their agency to provide them with that information. Unfortunately, the science provided is often biased, such as by cherry-picking supportive studies, to support agency proposed rules (Williams, 2016). Of course, it’s not just agency scientists. Some agency economists adjust benefits and costs to support the preferred option, either out of belief in the program or because it is necessary for their career paths. Jerry hoped that the Report Card would help to force them to do a better job.

4. Results

Agencies are charged with looking back on past regulations and discovering what is working, what is not, what can or should be dropped, or what should be changed. ¹ It is called Retrospective Review. Numerous scholars have examined the lack of retrospective reviews and come to the conclusion that there is almost no incentive to perform them (Lutter, 2012). First, agencies are too busy putting out new regulations to worry about ones that are already on the books. Second, most incumbent firms are not interested in reviewing existing rules because they worry about having an existing rule eliminated and creating a brand new one to replace it with new start-up costs. They also like a lot of regulations that put smaller firms and new entrants at a disadvantage – that’s why they supported them in the first place. There are various places where agency results are examined but they are far from sufficient.

The final step is where the information in the Government Accountability Project and the Regulatory Report Card is supposed to come together – accountability.

5. Accountability

It happened at an annual Mercatus “retreat” for Congressional Chiefs of Staff. This was an annual event where all Chiefs of Staff were invited to hear from Mercatus scholars, both in-house and associated scholars. Jerry had given a talk on the Government Accountability Project, taking the legislators through Congressional responsibilities for overseeing regulatory agencies.

After his talk, one of the Chiefs came up to us and told us the entire enterprise was a waste of time. He said, “You guys don’t get it, no one cares. They don’t care about whether the agencies achieve any outcomes or not. Once they have passed a law, they view their job as being done.”

We ended the Government Accountability Project, but the Report Card stayed on for a while. The sad thing is, even without judging the often-poor science, the average grade for RIAs was an F. There were lots of problems that Jerry wrote about including the fact that the economists were writing their RIAs after decisions had already been made (Williams, 2008). In addition, they were usually working directly for the managers who made the decisions. In a sense, they were evaluating the quality of their boss’s decisions. That meant their career paths depended on making their boss happy and saying that costs exceed benefits or that another regulatory option was more cost-effective, like doing nothing, was going to hurt them.

A few years ago, William Yeatman reported, “Today, apathy (or worse) is the defining characteristic of legislative oversight” (Yeatman, 2020). Congress simply does not care. If Jerry were still with us, I think he would have prioritized finding a way to make Congress care about the trillions of dollars citizens are forced to spend complying with federal regulations and his next step and would have been to find colleagues to work with him and come up with creative solutions.

Of course, he would have maintained his balance in life. He took pleasure in things like his train set, which you would see the moment his front door opened. He also introduced me, and many others, to the tiki life. His favorite place was the very dark Frankie’s Tiki Hut in a concrete building squatting in the parking lot of a strip mall in Las Vegas. We would walk in listening to jukebox music from outcast 1950s beatnik and surf genre and wait for our eyes to adjust to the low blowfish lights and statues of tiki Gods. Sipping on extremely alcoholic and colorful tiki drinks was a great way to forget about the Administrative State for a while.

He also never failed to talk about his wife and daughter to whom he was devoted and endlessly praised. He left us all too early but his character and his work live on. It will fall to the next generation of scholars, perhaps economists, psychologists, attorneys, or political scientists to figure out a way to make Congress care enough about the Administrative State to carry on this work.