Wednesday, September 7, 2016

Why the Retirement Train Wrecked

I was recently contacted by Northwestern University MBA student, Jared Scharen, about writing a guest post for his eRetirements blog. The following post first appeared there. I hope you will visit his blog.

My first job out of college, with a degree in computer science, was with staid General Electric's Information Systems Division in 1975. I was promptly taken under the wing of a few GE lifers whose first advice was this: “Never change employers. Stick with GE for your entire career and they will pay for an excellent pension for your retirement.”

At the ripe old age of 21, retirement wasn't on my radar. I stuck with GE for three whole years. Since the 1980's, defined benefit pension plans have all but evaporated, replaced by defined contribution plans like 401(k)'s. That's retirement planner-speak for the burden of funding retirement being shifted from your employer, like GE, to you and me.

Whereas the 1975-vintage GE retirement plan guaranteed eventual retirement benefits, new plans define how much you can contribute to the plan with tax advantages and your eventual retirement income depends on how well you invest those contributions. Defined benefit "pension" plans like GE's are nearly extinct, as you can see from the following graph published by the Employee Benefit Research Institute (EBRI).

The terms "defined benefit plan" and "defined contribution plan" can be confusing. A defined benefit plan, sometimes referred to as a pension, says, "Work for us for 30 years and we will pay for your retirement. Let us worry about where the money comes from to pay for it." A defined contribution plan, like an IRA or 401(k), says, "You can save for retirement and we will give you a tax break on those savings. You can use this money to help pay for retirement and the amount available to do that when you retire depends on your personal investment results."

401(k) plans were created during the Reagan administration and quickly began to replace defined benefit pensions. While it is tempting to imagine conspiracies at play to transfer the risk of retirement funding from corporations to employees, it appears that defined contribution plans were a random development that just happened to meet the interests of large companies that wanted to shed the burden of their employee pension plans. Pension plans weren't perfect, either – some became “mobbed up” and others simply made promises they couldn't keep. Several failed.

The notion that we are in a retirement-funding crisis is well documented. John Bogle, the storied founder and retired CEO of Vanguard Investments has referred to our retirement system as a “train wreck” waiting to happen. Labor economist, Teresa Ghilarducci, has testified many times before Congress on the nature of what she refers to as “Our Ridiculous Approach to Retirement.”

As Ghilarducci explains, we need to save 20 times our annual income (in some cases more) in financial assets to fund retirement. Retirement researcher, Wade Pfau, finds that in some cases employees will need to save upwards of 20% of their income throughout their career. These are staggering numbers for families trying to pay the mortgage and put a couple of kids through college. The results are seen in EBRI's Retirement Readiness Rating™ that "finds that nearly one-half (47.2 percent) of the oldest cohort (Early Baby Boomers) are simulated to be 'at risk' of not having sufficient retirement resources to pay for 'basic' retirement expenditures and uninsured health care costs."

Families need to earn enough over a 35-year career not only to support themselves for those three and a half decades, but also to help fund perhaps three more decades in retirement. Most Americans will receive Social Security benefits, but they are meant to replace only about 30% of pre-retirement income.

The “train wreck” we face is the result primarily of four trends. First, life expectancy at birth has increased from 47.3 years in 1900 to 68 years in 1950 to 78.2 years in 2009. Because we essentially pay for retirement by the year (a three-year retirement is far cheaper than a 30-year retirement, right?), these extra years of life for the typical retiree dramatically increase the overall cost.

A second contributor has been the stagnation of middle class incomes over the past 30 years. At the same time we shifted the enormous burden of personally saving enough money to fund retirement from corporations to individuals, after-inflation incomes for all but the highest-paid of the middle class flat-lined, as shown in this chart from the Economic Policy Institute.

(Note: Some economists question this income stagnation. I'll let you do your own research and make your own assessment, but I am personally convinced that it is all too real.)

A third contributor to the train wreck is medical cost inflation. Medical services are, as you might expect, disproportionately used by the elderly and represent a substantial risk to retirement finances. As I mentioned above, retirees are living longer and people who live longer consume more medical services than those who don't. The following chart from Business Insider shows both the stagnation of income for many in the middle class alongside the tremendous increase in medical costs since 1980.

A final contributor to the train wreck of retirement finance has been under-saving by Baby Boomers. (Before you complain about Boomers, note that studies like the  EBRI's Retirement Readiness Rating™ referenced above indicate that Gen X'ers and subsequent cohorts are in even worse shape.) In addition to stagnant incomes and a huge savings burden, we Boomers didn't recognize the importance of the sea-change from company-provided pensions to 401(k) plans quickly enough. Those few of us who did save enough didn't do so because we foresaw the retirement funding problem. We saved in 401(k) plans because we had high incomes and the plans offered substantial tax savings.

Most Boomers, in my experience, reached age 60 with very little idea how their retirements would be funded, except for some vague notion that they had earned Social Security benefits.

Corporations stopped pooling retirement risk for us and left us to handle it on our own. This put a tremendous savings burden on the middle class. Costs went up dramatically while incomes remained flat, so there wasn't much left to save. We enjoyed longer lifespans so retirement costs even more. Most of us had no idea this was transpiring.

Train wreck.

This explains the roots of the retirement finance “train wreck”, or “Our Ridiculous Approach to Retirement”, but it doesn't explain the challenge of retirement planning. In essence, we have about thirty or so years of career that must not only pay for those thirty years of supporting a family but also generate enough additional wealth to fund perhaps thirty or more years of retirement.

Fortunately, we don't have to generate that entire retirement income while we are working because we can invest our retirement savings and hopefully watch them grow, taking advantage of the time value of money. A dollar earned today and invested might be worth $7 or $8 in thirty years. (Of course, halfway through our careers we have only 15 years of job earnings remaining, so a dollar saved then is worth only $2.80, and right before we retire a dollar saved is worth about a dollar.) If our investments are successful both while we are contributing to our 401(k) plans and after we retire, we can generate a lot of the income from earnings. We need only save an adequate nest egg.

Unfortunately, that's far easier said than done.


  1. I have a slightly different perspective on the retirement "train wreck." I think the primary causes are the increase in longevity of people once they get into their 60s (much of the life span increase around the world over the past century has been decreasing infant mortality) and a decrease in private sector unions, particularly for blue collar jobs.

    The legendary pension only applied to a portion of the work force historically and many people were never eligible for a pension in the so-called golden years. Blue collar workers who did have a pension often had poor working conditions, so their survival rates after retirement were short due to disability and disease (coal miners and black lung are a poster child). So improved safety regulations and Medicare mean that many of these workers are now living much longer than previously at the same time they were losing their pensions.

    Unions helped keep workers employed for long periods of time with good wages. Most workers now are likely to work for different employers over their career now, so unless they have a union multi-employer pension plan, they would not have good pension rights because pensions improve greatly as you get to 20 and 30 year milestone dates of employment. Unions were also a primary force in getting wages up.
    Social Security is structured to replace most of the income of the bottom quintile of income. Its replacement percentage decreases as you go up, but financial resources generally increase so the big hole in the current retirement system is in the middle three household income quintiles, mainly because these people are living longer than before and need more financial assets - many of these people didn't have pensions before, but they didn't need them because they didn't live past 70-75.

    I think the problem is not so much that the 401k exists - I think it is actually preferable to the old pension structure for many people because it is portable between jobs with little penalty for switching jobs as long as one participates and does not cash out. Instead, I think the issues are that employers used it as an opportunity to slash the amount of money that they put in period, abdicated management to expensive providers (changing now with lawsuits and fiduciary rules), had generally too long vesting periods, and did not push people (until recently) to contribute.

    A low-cost 401k (similar to what I have now but unlike what I had for the first 30 years), with a 2-3 year vesting window (I don't mind some incentive to keep people for at least a couple of years), auto-enrollment at 4% - 6%, and a corporate contribution of at least 4% would completely change the retirement landscape for the next generations. A fundamental key would be making multi-employer 401ks available to small business owners so that a high percentage of the population can be easily covered. The current system is far too complex with an unbelievable array of retirement account types that all essentially do the same thing but with very different rules and limits. A much simpler system would also hopefully make it more difficult to cash out which many people do now out of sheer confusion about what to do.

    1. I largely agree with your comments, but I suppose that's because, as you state, it's only a "slightly different perspective."

      The only "slight disagreement" I have is that, while I appreciate the portability of 401(k)s, I don't consider them better than a pension because an individual can never replace the pooling benefit of a large pension. Furthermore, annuities purchased individually by retirees have to price in adverse selection, whereas pension participants of all ages are pooled.

      This is largely moot, since pensions aren't coming back.

      Thanks for such a well-thought and comprehensive comment!

    2. I agree that pensions are great things. However, it is only a minority of the population that has ever seen the full benefit of both being eligible for a pension and having enough years in at one employer to get a sizable benefit. According to the BLS, only 38% of private sector workers were eligible for a pension in 1980. it is likely that many of those workers did not have 30 years in to get the full pension benefits.

      So the "good old days" never existed for many workers. It does highlight how critical Social Security is as this is the primary defined benefit that most people have had and currently will have.

      I think a fundamental societal goal should be to improve the conditions for the majority of the population that were previously largely left out in the cold instead of optimizing it for the top third. Defined contribution plans, if structured properly, should be able to successfully cover a much higher percentage of the work force than the old pension structures did and current DC plans do. That would be a very good thing. Even if they only end up saving a total of 6% or so a year (employer contributions included), it would be a vast improvement for many middle-class people. I think the biggest barrier to accomplishing this is the complexity and opaqueness of the system. Talking to people, it is clear that they are often simply paralyzed by confusion and indecision. The recent changes of auto-enrollment, target date funds, and increased fiduciary standards are clearly improving things over the past few years. Even auto-enrollment with payroll deductions into a Traditional/Roth IRA with the $5500 max would be an improvement over the current system for half the population.

  2. I would add to your increasing longevity and medical care costs, the increase in the median and average house sizes while house hold size is declining. This has been a major impetus in cost of living increases and increasing mortgage size. Declining interest rates have helped allow it, but you still have to pay off the mortgage amount at some point. Unlike what the realtors and bankers tell potential buyers, the typical house only increases in value with inflation. So if you have good luck in the housing market, you have a relatively illiquid asset with significant potential for loss of value at some point. If unlucky, then you have a place to live but no equity.

    An additional factor is that college and university educations have increasingly been critical in getting decent jobs and those costs have been shooting up as well. an increasing number of people are no carrying significant student loan debt. I am not at all convinced that the past statistics when a much smaller percentage of the population attended college (usually only the top students in a high school class) are a good predictor of how well the people from lower down in the high school graduating class are going to do in the job market. so once again, much of this may be marketing hype selling an expensive product without the promised benefits.

    1. My last few posts about reverse mortgages were aimed at the liquidity problem. The Economist issue to which I linked explains the huge degree of government subsidization of U.S. housing, which is also a part of your concern.

      The college/jobs issue will no doubt affect future retiring cohorts, but college served us Boomers well and we fell to the retirement crisis first, so I don't think that's a cause for the derailment. It is, I agree, an ongoing concern.

      Great points. Thanks for writing!

    2. I think the rapidly rising college costs over the past 20 years or so have been very important to the Boomers because of the rising percentage of children that have been going to college. There has been much discussion about the student debt taken on by Millenials but much less discussion about the cash expenditures and debt taken on by the parents who are now retiring. Since financial aid grants are often made to students from lower financial tier households, much of the increase in the college costs has been borne by Boomer parents in the top half of the income brackets.

      My guess is that much of the "missing" 401k contributions in Boomer households over the past couple of decades can be found in the extra 1,000 sf of house and increased tuition payments for two kids in college. We can only hope that the past tuition payments will turn into future increased GDP and household incomes in decades to come to generate future payroll taxes to support SS and Medicare. That is after all what the implicit promise has been made to the increased percentage of the population who have attended college over the past 20 years.

      There have been a number of unparalleled simultaneous confluences of housing, education, longevity, and retirement changes over the past three decades that are just now starting to reveal their consequences. I think there will be a lot of lessons absorbed by Millenials watching their parents go through this that will have profound impacts on saving and spending trends over the coming decades, especially on the big ticket costs of housing and education.

  3. In the third paragraph, just above the EBRI chart, you meant to say "Defined benefit . . .", but said "Defined contribution . . ."

    1. Nice catch. Thanks!

      I have corrected that and asked Jared to change it on his blog, as well.

  4. College is important for obtaining quality employment - the type that actually offers a 401k retirement plan. It is not required but it helps a lot. When I went to school you could get a good education for minimal dollar outflow. I find it sad that today's boomers are unwilling to ensure that a public college education is still affordable.
    As to 401Ks vs traditional retirement, I enjoyed a 403b plan. Working and saving diligently from 1985-2000 (lucky years to be working and saving) allowed me to retire after a long 14 year career - something that would probably not be available to me in a defined benefit plan.
    I ported my 403b through three employers. I think portability is hugely important.
    Two changes I would like to see are A. mandatory contributions, and B. no early withdrawal.

  5. Nice post Dirk ... kind of a downer theme though. One of the elements I would add is the change from the extended family that used to exist and tight knit communities that took care of the elderly, to the dispersed family where people need to care for themselves. This trend is all over the world where the rural are moving into the suburban and urban where extended family living isn't like it used to be.

    1. Thanks, Larry! (I find reality depressing sometimes, too.) :-)

  6. Dick, I really enjoy your posts. It's easier to identify the cause of the train wreck, but another matter to avoid it. For many seniors their only real asset is the equity in their home. Seniors have to let go of the retirement dreams their parents had (leaving the home as a legacy to their kids) and instead, use their equity to help them make it through the next 30 years.

  7. Interesting reading Dick. Although UK has some different systems and certainly health costs are not such a big concern as in the USA, the move from defined benefits to defined contributions has been the same here and the "train wreck" analogy is worryingly appropriate here too.