‘Cajoling participants’ not enough to boost retirement savings rates
Alicia H. Munnell, has been a leading figure in government and academic circles focusing on retirement issues for several decades. Since 1997, she has directed the Center for Retirement Research at Boston College, where she also serves as the Peter F. Drucker Professor of Management Sciences. She previously served on the President’s Council of Economic Advisors (1995-97), as Assistant Secretary of the Treasury for Economic Policy (1993-1995), and Senior VP of Research for the Federal Reserve Bank of Boston. She holds a Ph.D. from Harvard University. She recently shared her insights on retirement issues with EBN.
EBN: What conclusions of studies that the Center has conducted over the years have surprised you the most?
Munnell: One is that how your retirement investments are allocated between stocks and bonds plays a minor role in determining retirement security. What makes a huge difference is how long you work. The key to this is that most people have so little in the way of assets that asset allocation doesn’t make that much difference, whereas retiring as 70 instead of 62 and claiming your Social Security benefit eight years later gives you a benefit that’s 76% higher, plus all your assets have the opportunity, over that extended period, to grow. It’s obvious when you think about it, but people spend a lot of time talking about asset allocation, and that’s not really the big lever.
EBN: What else?
Munnell: Another one is that the efforts to protect older workers from employment discrimination by making it difficult to fire them, may actually hurt them as a group. That’s because employers are reluctant to hire older workers because if they make a mistake and get a lemon, it’s very hard to get rid of that lemon.
A third surprising observation is that the overall amount of money being saved for retirement hasn’t changed during the transition from the defined benefit plan world to the defined contribution one. I still think that you get a worse result, because it’s harder to squeeze income out of defined contribution assets, and I think that people have to work about a year longer to end up with the same amount of income. I also think that defined contribution wealth is probably more skewed towards higher-income people than defined benefit wealth. We’re looking into that hypothesis.
EBN: Has the overall personal savings rate – not just retirement – declined over time? And if so, is there a cultural shift away from thrift involved?
Munnell: If you look at the household in the middle of the income distribution, their average total non-retirement savings is around $15,000. So people save through either defined benefit pension plans, defined contribution plans, or through paying off their mortgage. Since the 1980s, the amount of money going into savings plans has changed little as a percent of wages. That surprises me, considering the shift to defined contribution plans, which require people to do more.
I don’t think there has been any shift in the culture against saving, but middle class incomes really have not grown for 40 years. So every decade that they don’t grow, the harder it gets for people to save. There is a real headwind there pushing people back, but not because they are less virtuous.
EBN: I hear that fewer people who own homes expect that they will ever pay off their mortgages. If true, how does that affect things?
Munnell: I am distressed at the fact that people are increasingly entering retirement with significant debt. But we are a consumer society. The question is, has it really gotten a little worse, or much worse? I’m not convinced it’s gotten much worse.
EBN: Are you optimistic that defined contribution features like auto-enrollment will turn things around in terms of boosting retirement income security?
Munnell: From the data I look at, less than half of people covered by these 401(k) plans are being auto-enrolled; this idea has not caught on to the extent that you would have thought. But it’s the only way to keep participation up. Also, auto-enrollment without auto-escalation starting from a reasonable deferral rate like 6% could be pernicious if you enroll people at low rates like 3%, and participants just stay there.
EBN: Should auto-enrollment with auto-escalation be federally mandated?
Munnell: Honestly, I don’t understand why Congress can’t pass a law requiring it. It seems we’ve gone about as far as we can go by cajoling participants. It’s true that some companies worry about the added cost to the company if they have to make more matching contributions if everybody’s suddenly in the plan. But it also means that your workers can actually retire when they get to retirement age, and that’s what you would like to have happen for, certainly for workers whose productivity declines.
EBN: If traditional retirement age workers don’t retire “on schedule” en masse, will that limit job opportunities for younger employees?
Munnell: It might look that way if you’re at a small company where one person’s decision not to retire from a senior position keeps someone else from getting that job. But I’m not very sympathetic to that argument; it’s linked in with the “lump of labor” theory that says there is only a fixed amount of work to go around. That idea has played a prominent role in Europe, and has lead to restrictive labor policies that have not worked out too well. Market economies are dynamic; if older people work longer, they have more buying power, fueling demand, stimulating the economy leading to the hiring of more people.
EBN: But don’t you also favor having a mandatory retirement age?
Munnell: Yes, and it surprises people because of my views on related topics. But it actually helps older workers, because when employers know they won’t be put in a position of having to worry about an age discrimination charge, they will be much less reluctant to hire workers who are closer to that mandatory retirement age.
EBN: Do you think Congress would ever go along with that idea?
Munnell: Unfortunately, I don’t; I think it would be seen as a civil rights issue, although I believe it comes under a different heading. If you’re born female you’re female for your whole life, and if you’re born black you’re black for your whole life, but you’re not 65 for your whole life. I mean we all are 65 at some point, or 66 or 67, whatever you want to make that age at which it gives the employer the right to let someone go on the basis of age alone.
EBN: What about Congressional action on less controversial retirement issues?
Munnell: I already mentioned requiring automatic enrollment and deferral rate escalation. That can be bipartisan. I don’t want to rip apart the 401(k) system and replace it with anything else; it’s here to stay.
Another area is how to handle plan “leakage,” people taking money out of their 401(k)s before retirement. There is much too much of that going on. I think the continued focus on plan fees is good. I just want the current system to work well, and Congress should want it to work well, too. But we also need to do something about the people who don’t have access to a 401(k) plan. Some kind of auto-IRA. That seems like a no-brainer to me.
EBN: Why not just expand Social Security?
Munnell: There are all [kinds of] calls for expanding it, so I’m a little less nervous than I used to be about its future. It is the backbone of our retirement system, and the other components really aren’t working that well. But it definitely needs money, there is no manna from heaven here and so the question is how do you fix it going forward? You need to put more money into the system, because people could not survive further cuts in benefits. Benefits relative to earnings are already going to go down, because the full retirement age is moving up, from 65 to 66 where it is now, to 67.
EBN: Do you think it’s possible, for their planning purposes, to give workers any conservative assumptions about how much they should expect a balanced retirement portfolio to grow?
Munnell: People used to think 8% was the right number for stocks, based on historic returns, but I don’t think it’s reasonable to sort of use historic rates going forward. I can’t say any more than that, and I know that other people really don’t really have an answer, either.
EBN: Another wild card is the medical costs people will face in retirement, those that won’t be covered by Medicare. Where do those predictions fit into the equation?
Munnell: We’ve done these studies, and so have other people, that show that medical costs for a couple over their entire life are going to be $200,000 or so. They’re really high, and they’re rising. It is a big risk for people. We’re going to enter an era in which people are relying primarily on 401(k) plans, and so their well-being will depend on how they draw down that money. If they’re really scared of big end-of-life medical costs, then they’re just going to husband those resources, and deprive themselves of things they really need. And so it’s not only the direct effect of how much they’re going to cost, but how it makes people behave during retirement.
EBN: Do you believe that hybrid defined benefit plan designs that shift some of the liability for pension portfolio investment performance to the participant, will stave off the extinction of the DB plan model?
Munnell: So not so much in the private sector; it seems like that opportunity sort of came and went with cash balance plans. But maybe in the public sector you will see some types of risk sharing. For example Wisconsin has some type of arrangement where you don’t get a cost-of-living adjustment unless the plan achieves its assumed return. But I think in the private sector, we’re pretty much just going on this trajectory to end up totally with 401(k) plans.