Inequality and Retirement: A Conversation with Researcher Alicia Munnell

For years, you’ve been a leading voice on this wonky, complex issue. Why study retirement policy — and public pensions in particular?

After graduating from college, I moved to Washington, D.C., to work at the Brookings Institution and served as a research assistant on a book on the Social Security program. Then I went to Harvard to get a PhD and wrote my thesis on Social Security and saving. Social Security was never enough for private employees, so I started looking at private pensions. In the 1970s and ​’80s, I did a little bit of work on public plans. But then I moved on with my career until I started the Center for Retirement Research at Boston College in the late ​’90s and had an opportunity to go back to studying public plans because they were a neglected area of research. I started collecting data, and that’s really what brought us to where we are today.

What was the impetus for creating the Center for Retirement Research?

A Request for Proposals on retirement research came across my desk from the Social Security Administration, and I thought, ​“I know something about that.” I had never written a grant application before. I had this wonderful senior who was graduating. So, she and I filled out the grant application and scooped up people from MIT, the Urban Institute, and the Brookings Institution, and we won. Quite amazing.

I remember someone asking me at the time — it was a five-year grant from the Social Security Administration — won’t you run out of topics after five years? I knew that it was a field where we were just beginning to build new datasets that follow people over time, and that there were huge unanswered questions. So I was confident that it would keep me busy for at least five years. It has actually kept me busy for 20 years.

There’s certainly been a boom in research on pensions during that time, thanks in part to your work. But a lot of the information out there is conflicting. How would you characterize the financial health of these systems?

I think it’s really not useful to discuss public plans on average. Anyone who says, ​“All public plans are in trouble,” or, ​“All public plans are fine,” is wrong because they differ quite a lot. We find it very helpful to think of three groups. One is well-funded plans that are operating in a sensible fashion and have relatively high funded ratios, and that accounts for about a third. And then you have a second group — I characterize them in funded ratios, but you can characterize them in a lot of other things, too — about 60 percent to 80 percent funded. This group of plans could do better; they really need to pay their full required contribution. If they do that, they should be fine in the long run. And then we have those that are less than 60 percent funded. I would basically characterize those as basket cases where they have serious challenges. Nothing could be solved by modest changes, and really people need to sit down around a table and come to some solutions.

Researchers at the Brookings Institution recently released a report arguing that we should be looking at whether pension debt is growing faster than GDP, rather than funded ratios, to gauge how well pension plans are doing. What’s your response?

I’ve looked at the report from Louise Sheiner, and I think what she’s done is very innovative. This is an area where fresh views are very much appreciated. I don’t buy into her conclusions completely, but to think innovatively is really valuable.

I’ll tell you how I think about it. In terms of public plans, I think that the sponsor should put aside an amount each year equal to the normal cost — properly calculated — of the benefits accrued that year. If you have a police officer, you should pay his wages, and you should pay the amount that’s required to fund one year of pension benefits. That should be done for efficiency, so people will see the full cost of the services that they are receiving.

But that puts aside the question of unfunded liability caused by payments that were not made in the past. Our innovation is to ask how much of that cost is due to overestimating how much you’re going to earn on assets — a miscalculation which you should pay off over a relatively short period of time — and how much of that is due to the initial debt burden that the plan had when it started funding. For example, Massachusetts has been promising benefits since the 1920s. The state didn’t start funding its plans until 1990 — it simply paid benefits directly from its existing revenues. If I had my druthers, I would separate that unfunded liability out and pay for it as it comes due through sales tax or some type of bonding. But there’s no reason to associate that with today’s Massachusetts employees. The reason is not that we’re putting so much aside for today’s workers — because they actually pay a large part through their own contributions — but because we have a huge unfunded liability from the past. It really distorts perceptions about how expensive today’s workers are.

I’m guessing a lot of these insights have come from the massive dataset, known as the Public Plans Database, which you and your colleagues have built. Could you tell me about who uses that tool?

Reporters are using it. They go there for the information, and the Public Plans Database is cited often in the ​“Wall Street Journal.” Actuaries use it. One of them comes once a month to pull data and put their performance in perspective. People use it for academic research, and we have a lot of examples of that. We have people who are in the policy world using it. And actually people who manage public plans come to compare their funded ratios to other plans. I think we’ve gained a lot of credibility in terms of the accuracy of our database, and the breadth of it is increasing. 

In your research, you’ve not only looked at systems but also at individuals. Some of your most recent studies focus on whether women and people of color, who often earn less during their careers, will have enough to retire. Tell me about that work.

We’re very concerned about retirement security for low-income people, and that often is women and people of color. We generally talk about those issues through the lens of Social Security. We’ve put out a lot of material on financing Social Security, but we’ve also released some studies on how it could be modernized by taking into account the pattern of women taking time out of the labor force to care for children, and how to do something about the minimum benefit for those who have a sporadic work pattern. And we recently did some work on retirement wealth by race that noted how Social Security helps offset inequality among people entering retirement.

So, Social Security is key to addressing the ripple effects of pay gaps on retirement earnings. But since many state and local workers aren’t enrolled in the program, what are the takeaways for governments?

We just did a really exhaustive study on state and local workers who are not enrolled in Social Security and the extent to which their benefits are comparable. Basically, the deal is that states and localities don’t have to enroll their workers in Social Security as long as they’re providing benefits of comparable value. What we found was, yes, states and localities were mostly satisfying the letter of the law in that you got sort of a comparable benefit at age 65. But if you try to look at it over people’s lifetime with the inflation adjustment and everything else Social Security offers, they were falling short. There was also this issue of how do you think about beneficiaries in plans in places like Chicago, where they’re actually going to run out of money. The answer here to me is very simple — you could solve the problem in a very complicated way with a lot of legal provisions, or we could just extend coverage to state and local workers.

That is a simple solution. It might be harder to implement politically, but it’s simple to conceptualize. It also leads me to my last question, which is: If you were a legislator, what other things would you do to ensure people are able to retire with financial security?

In the public sector, legacy debt is very important because it clarifies the level of benefits and doesn’t put an artificial burden on the current generation to pay everything off. But I would also think hard about retirement ages. In Rhode Island when then-treasurer and now Gov. Gina Raimondo was trying to reform the system, her position was that she wanted her public employees to have a secure retirement. The key question is when does that retirement start? Retirement ages in the public sector are early, and given that we’re all living longer, the answer is not so much thinking about cutting the amount of the benefits, but rather to think hard about when those benefits should start. That’s an issue worthy of consideration.