Autopsy of a Retirement Plan

As I approached age sixty-five after thirty-seven years of university teaching, I took stock of what my retirement income would look like. Many retirement experts claim that at least 70 percent of preretirement income is necessary to maintain one’s standard of living. For example, someone whose final annual income will be $100,000 should have as a goal an income of $70,000 in retirement.

I ran the numbers for my Social Security and my employer’s defined-contribution plan, in which I had participated for thirty-one years. This 401(a) plan, which functions in the same way as a 401(k), had been administered at various times by TIAA, ING, and Prudential.

Together, my projected Social Security and employee retirement plan would amount to just 43.5 percent of my final income. The monthly Social Security check accounted for 19.5 percent; the annuity income option for my defined-contribution plan, 24 percent.

Something had gone terribly wrong. Despite having accumulated almost a half-million dollars, which is much more than the $125,000 average for people approaching retirement, I did not have enough to finance a retirement that would allow me and my family to maintain the middle-class standard of living that my $117,615 final salary as a university professor afforded.

I don’t expect much sympathy from others who earn or have saved less. I had a good job that supported a better standard of living than most, and I was about to lose some of that middle-class privilege.

I’m aware that, as a tenured professor, I had considerable relative privilege within an academic labor force with severe salary inequities. Faculty members holding contingent positions are far less likely to have any retirement plan at all, and if they do have defined-contribution plans, their contributions are likely to be lower—and more of a sacrifice to make—than those for tenured and tenure-track colleagues, whose higher incomes enable higher savings rates with less of an impact on their ability to afford housing, food, health care, and other necessary expenses. Nonetheless, there is a lot for others to learn from my experience. If I had a good income that wasn’t going to turn into a good retirement, then anyone with a defined-contribution retirement plan could be in danger.

But why my experience alone? Aren’t there already studies of these plans? Yes, but they are studies based on projections, assumptions, and modeling or indirect indicators. TIAA, for example, claimed that “on average, participants in TIAA-administered plans are on track to replace over ninety percent of their income in retirement.”

Despite decades of experience with these types of plans and trillions of dollars running through them, there is a dearth of accessible research about actual rather than hypothetical experiences with them. I suspect that the financial companies that administer these plans have performance data they do not share with the public because the resulting retirement incomes are too depressing and contrary to their rosy advertising campaigns.

Retirement-account balances are like poker hands. They are treated by their owners as closely held secrets. Coworkers are unlikely to discuss openly how much they have accumulated in their accounts. It’s like asking people how much they have in their checking accounts: none of your business. The problem, though, is that it is the business of all of us to know what’s going on with defined-contribution plans.

Was I at fault? Had I not saved enough or made poor investing decisions? Or was the game rigged against me and, by implication, anyone else participating in such plans? Was it possible that even if I had saved and invested more responsibly I would have still ended up without enough retirement income?

I was not alone. Increasing numbers of Americans with defined-contribution plans are coming up short for retirement, and this was especially true after the 2008 recession. James Ridgeway wrote a 2009 Rolling Stone article in which he recounted how miserably his own plan was doing and lambasted the whole approach. He started with the bitter faux riddle: “What starts with ‘f,’ ends with ‘k,’ and means ‘screw your workers’? That’s right—401(k).”

For the autopsy I am focusing on my defined-contribution plan. But since Social Security is also a significant part of my retirement income, a few words about my experience with it are in order.

Social Security, unlike a defined-contribution plan, is not a retirement-savings plan. Rather, participants and their employers pay contributions into a social-insurance fund to protect themselves from loss of income in retirement. The contributions go into a collective pool out of which benefits are paid to individuals. They do not go into individual accounts like savings or defined-contribution retirement plan accounts. Participants don’t own their contributions. They accumulate instead guaranteed rights toward income replacement in retirement.

Social Security was designed to replace far less than 70 percent of preretirement income—it replaced just 19.5 percent of mine. The assumption was that employer-sponsored plans, which at the time were mainly in the form of traditional pensions, would be a greater source of income replacement that would be enough to make up the balance. Today, though, Social Security is the biggest source of retirement income for most people. It accounts for more than half the income of 48 percent of married couples and 71 percent of unmarried persons over sixty-five.

Income replacement has two components: the amount of income that is replaced and the period of time that it lasts. With Social Security, the amount replaced varies mostly according to how much you contributed over a working life. The formula for payment is set according to the highest thirty-five years of payments. The payment amount is modified by the program’s being moderately progressive in its distribution. One of the goals of Social Security was to reduce elderly poverty. Hence, low-income workers have more of their preretirement income replaced than do higher-income workers. As a result, Social Security reports tell us, very low-income workers actually have well over 70 percent of their average incomes replaced, while the highest earners have just 29 percent replaced.

That statistic immediately raised the question for me of why I was having just 19.5 percent replaced. The answer, I realized, was simple. My 19.5 percent was a percentage of my final salary, whereas Social Security’s statistics are based on thirty-five-year career averages. It doesn’t release statistics based on final salaries. I quickly calculated my career average salary and found the replacement value of it, using Social Security’s methodology, to be closer to 44 percent.

For most, there’s a tremendous difference between replacement of final and average career income. For those who have received raises throughout their work lives, income at age sixty-five is markedly higher than income twenty years earlier. Their incomes at age forty-five would be closer to their career-average income. They would be much better off having 70 percent of their income at age sixty-five. It is also possible to have been forced into lower-income jobs later in life so that final and average incomes are closer. The same is true for those who taper off to part-time work before finally retiring.

It’s understandable why the Social Security Administration releases data on average rather than on final income replacement, given that its retirement income calculation is mostly based on the thirty-five highest years. These figures also avoid the problems of determining how to calculate final incomes when there are career patterns for which replacement ratios for final incomes would be misleading and inadequate in terms of retirement-income needs. But it is nevertheless confusing for those trying to prepare for retirement and predict their income. The Social Security Administration has appropriate data and should use them to calculate final as well as average income replacement.

Financial-service companies have taken advantage of the Social Security practice in making their own claims of likely income replacement. TIAA, the leading provider of retirement plans for academics, has the earlier-cited advertising campaign in which it claims that more than 90 percent of its plan participants are on track to replace more than 90 percent of their income, combined with Social Security, in retirement. This makes it look like participants will clear the 70 percent hurdle with ease. The fact that these figures refer to average, not final, career income is in the fine print. They, like me, will come up far short of 70 percent of final income. Such marketing lulls participants into complacency during their working years. They are in for a rude awakening when they approach retirement with much less than anticipated income for the future  and can do little about it.

Examining my Social Security statement earnings record did reveal some sources of reduced retirement income. For eight years I had public academic employers who did not participate in Social Security—approximately 6 percent of employees in the labor force have employers that are allowed to not participate. Had contributions been made for four of those years, which would have been among my thirty-five highest-earning ones, my retirement income would have been higher. This is similar to the problem caregivers face if they take time out to care for children or other family members. I would have happily paid the contributions during those years but I did not have that choice. My experience raises questions about whether Social Security should allow participants to pay concurrently or retroactively for noncontributing years.

Now we are ready to look at why my employer-sponsored defined-contribution plan failed to provide nearly enough retirement income. I had saved my quarterly statements. I knew how much had been contributed, how the investments had been allocated, and how they had grown. With the help of a spreadsheet I was able to trace the growth of my accumulation and map out alternate scenarios.

The immediate question confronting retirees with defined-contribution plans is what to do with their accumulations so that they can finance their retirement years, which means the rest of their lives. The original idea of the defined-contribution approach was that the accumulations would seamlessly be converted to life annuities to mimic traditional pensions. Life annuities are products sold by life-insurance companies, and I had planned to use my defined-contribution accumulation to purchase a life annuity. In return, the life-insurance company would have paid me a monthly income for the rest of my life, as long or as short as that might be.

Life-insurance companies are crosses between banks and casinos. They are financial firms that make bets on how long annuity purchasers are likely to live on average and then adjust their prices accordingly.

Defined-contribution plans have not turned out as originally anticipated in this respect. Most participants do not use their accumulations to purchase annuities when they retire, but instead manage their money differently—for example, by continuing to invest while making regular withdrawals. Annuity prices are nevertheless useful for knowing what accumulations are worth in terms of potentially providing pension-like incomes for the rest of annuitants’ lives. They form the bases for retirement studies (such as those by the highly regarded Boston College Center for Retirement Research) that calculate potential replacement incomes with given accumulation sizes.

I reached the normal retirement age of sixty-five at the end of 2009. That was a particularly bad time to do so for anyone in a 401(k) or other defined-contribution plan. Because of the 2008 Great Recession, the stock market had tanked, taking with it the prospects of those like me who were approaching retirement. My portfolio lost 19.1 percent of its value between October 2007 and April 2009, seven months before my sixty-fifth birthday. To make matters worse, annuity prices were increasing at the same time. It was a perfect storm for anyone who wanted to retire that year with a defined-contribution plan and purchase a life annuity.

If I had had a crystal ball, in September 2007 I would have moved all of my stock investments to bonds and money-market funds. Then, when the market began to recover in late 2009, I would have moved them back into stocks. But I did not have a crystal ball. I did have enough sense not to completely panic in 2008, when it became clear that the fall in stock prices was not part of a temporary cycle of ups and downs. I didn’t sell and hoped that the market would bounce back.

It was not at all clear how long it would take for the market to regain its lost values. After the 1929 crash, the Dow Jones took twenty-five years—until 1955—to regain its value. If the recovery after the Great Recession had taken that long, I would have been ninety before my investments regained their value.

Fortunately for me, this time the market recovered from its losses a lot faster. By 2013 the Dow was back to where it had been five years earlier, before the crash. By the middle of 2012 I was back to where I had been in 2007—faster than the Dow because of the proportion of bonds that had not lost value in my accounts. Then I cashed out completely—but more about that later.

Were my problems the fault of timing, the bad luck of approaching retirement as the market was crashing? To test that possibility, I pretended that I had turned sixty-five in 2007, when the market was at its high point. If I had been able to retire then, my sources of retirement income would have added up to a 47.5 replacement rate, somewhat higher than the 43.5 percent rate I was looking at two years later but still well short of the 70 percent needed to maintain my family’s standard of living. It’s worth pointing out that while the 70 percent replacement rate is the most often cited goal, there are other retirement advisers and experts who think it should be much higher, including New School labor economist Teresa Ghilarducci, who thinks it should be 100 percent.

Some investment experts advise gradually reducing stock ownership as you get closer to retirement age to avoid losses like the ones I experienced. They have designed target-date plans that incorporate this strategy. A target-date plan would have helped in my case, but such products achieve predictability at the expense of returns and still wouldn’t have gotten me to the 70 percent replacement rate. Nari Rhee, director of the Retirement Security Program at the University of California at Berkeley, estimates that target-date funds have an average 6.05 percent rate of return. If I had had my retirement savings in such a fund with that rate of return, I would have avoided losing value in the 2008 crash but actually come up shorter—replacing, combined with Social Security income, 40.3 rather than 43.5 percent of preretirement income.  That was because, as we will see below, my actual average rate of return was higher than the 6.05 percent target-date fund assumed average rate of return.

The leading explanations for people who come up short with savings in 401(k)s and other defined-contribution plans is that either they didn’t save enough or they didn’t invest what they saved wisely enough—suggesting that luck isn’t a significant variable. Such explanations are undeniably true up to a point. If you save more, you will accumulate more; if your rates of return are higher, you will accumulate more. But they don’t answer the question of whether it is realistically possible to do enough of either to reach the 70 percent replacement rate.

Was I saving enough? Between my employer’s 8 percent of salary contribution and my 5 percent of salary contribution, I was saving a combined 13 percent of my salary for thirty-one years. That is a savings rate nearly double the average 7 percent of salary saving in 401(k) plans.

If my 13 percent rate of savings had not been enough, how much higher would it have had to be to get me to 70 percent? I went back to my spreadsheet and inserted higher and higher rates of savings until my combined income reached 70 percent. Reaching that level would have required an unrealistically high savings rate of 26.1 percent of salary, just over double my actual savings rate. Such a high rate of savings would have exceeded the legal limits for tax-deferred contributions.

But even if I had been able to put in the excess savings in a fully taxable investment fund outside of my retirement plan, you can defer gratification only so much before you end up lowering your and your family’s preretirement standard of living. The two big expenses of the middle-class family—the home mortgage and paying for college for children—present additional obstacles to saving. Although many of us consume more than we need to, at some point frugality crosses over from cutting out superfluous expenses to going without necessities; well before that point you are cutting out the small pleasures that make life worth living, such as an occasional movie or dinner out, not to mention vacation travel.

If I can’t be blamed for not saving enough, perhaps the problem lay in how I was investing what I was saving. When I started in these plans in 1979 I knew next to nothing about investing. Confronting a decision over how to invest my money through TIAA, I reasoned clumsily that bonds were guaranteed and stocks a gamble. I preferred what was guaranteed, assuming incorrectly that it would work out to more or less the same accumulation in the end.

Eight years later, I read somewhere that stocks had much higher accumulations than bonds in the long run. I then changed my allocation, putting 65 percent in stocks and the rest in bonds and money-market funds, a mixture that remained for the rest of my career. The money accumulated in TIAA bonds had to remain because the company, unlike all others, does not allow participants to move it into stocks except over a protracted period of ten years through a process that is somewhat difficult to set up. Although I was not committed enough to invest the time and energy into that project then, I probably should have made the effort to do so.

My approach to investing was thus haphazard and far from fully diligent. To that, I plead completely guilty. I didn’t know much—and I still don’t—about investing. But should retirement security depend on investing skill or luck? If retirement incomes are supposed to reward careers of hard work, then they should depend on how many years are worked and the salaries during those years. That is how Social Security and traditional pension plans, neither of which require investing skill, determine retirement incomes. Some would agree that virtuous, provident saving should also determine retirement security. Adding investing skill or luck to that mix builds in an extraneous factor that has nothing to do with working or saving.

Enough of my excuses. How well did I do as an investor over thirty-one years? Using the spreadsheet I entered my final plan accumulation and year-by-year contributions. Through that exercise I was able to determine that my average yearly return on investments had been 7.1 percent.

How good or bad was that? The average rate of return on the Standard & Poor’s 500 for the same years had been 8.1 percent. My rate of return was lower. The average rate of return for the S&P 500, however, is not the same as the average investor’s experience, which is lower because of the costs of retirement company fees, commissions, and other profits. According to Dalbar, a financial research company, the S&P 500 grew by an average of 10.35 percent between 1986 and 2016, but the average investor had a return of only 3.66. Dalbar attributes the gap to investor errors, such as not holding funds long enough to realize full gains, as well as what it refers to euphemistically as fund expenses.

It appears I stumbled into a relatively decent rate of return. My 7.1 percent average rate of return was even higher than the 6.05 percent rate cited above that is assumed for target funds, yet I still came up very much short.

What rate would I have needed to make it to the 70 percent replacement? Instead of my actual 24 percent replacement rate from the defined-contribution plan, I would have needed a 50.5 percent replacement rate. That would have required an 11.4 percent rate of investment return on the amounts of my and my employer’s contributions. Had I known how to beat the market so successfully, I could have quit my job to write a get-rich-through-stock-investing best seller or become an adviser to Warren Buffett.

Altogether, to have reached the 70 percent replacement rate—to yield $82,330 annually from savings and Social Security combined—I would have needed to accumulate $1,046,074 in my defined-contribution plan. That was $550,020 more than the $496,054 that I had—more than twice as much. Since a number of financial planners attempt to estimate accumulations needed as multiples of final salaries, it’s worth noting that this million-dollar-plus figure represents 8.9 times my final salary of $117,615.

Perhaps the problem was that I had not been in the game long enough to reap its rewards. I was thirty-four when I began my first job with a defined-contribution retirement plan. That gave me thirty-one years to save and invest.

To test this possibility, I gathered my Social Security statements from before that first job with a defined-contribution plan. Where there were noncontributory employers, I estimated the incomes.

I first appeared in the Social Security files at age sixteen with a yearly income of seventy-three dollars from a brief busboy job. I then pretended that I had been making 13 percent of salary contributions with a 7.1 percent rate of return from that job at age sixteen and all my subsequent jobs until the time I began my actual retirement plan. That seems to be in line with the financial services message that it is never too early to think about, worry about, and save for retirement.

The result was an accumulation of $10,416.84, which then hypothetically grew for another thirty-one years at the 7.1 percent return until I reached age sixty-five for a grand total extra accumulation of $87,339.28. That gave me nineteen extra years for a total of fifty years to accumulate. I added the results to my actual accumulation. My defined-contribution plan rate of replacement hypothetically went up from 24 to 28.2 percent. That, along with my Social Security income, brought me up to a hypothetical 47.9 percent replacement rate—still far short of 70 percent.

Any way I looked at my situation at age sixty-five, I was in deep trouble. My wife did not have enough income from her Social Security or employer retirement plan to compensate for my shortfall and we still had a dependent daughter living at home. My options were either to take a vow of poverty for my retirement years or to strategize a way to be able to retire with enough financial security at age seventy—or, in the worst case, never retire.

As a result of exceptional circumstances, I was able to escape the trap I was in. Delaying my retirement and switching from my defined-contribution plan to a defined-benefit pension plan provided me with a much higher retirement income.

I was fortunate to be able to delay retirement. I didn’t have a physically strenuous job that had worn me out, my health was good, and my job was secure. In addition, for what it’s worth, I didn’t hate it. That, of course, is not the situation for everyone who comes up short at age sixty-five.

Delaying retirement until age seventy carries significant advantages for Social Security income. Every year past the normal retirement age results in 8 percent higher income. The additional five years of work increased my replacement rate from 19.5 to 29.8 percent for the Social Security portion of my retirement income.

Five more years of accumulations in my defined-contribution plan at my average rate of return would have lifted that portion of my replacement rate from 24 to 36 percent. I would still be short at 65.8 percent. It would take me until age seventy-three to accumulate enough to retire at a 70 percent replacement rate—all the time keeping my fingers crossed that there wouldn’t be another stock market crash. I would have to keep up an aggressive investment strategy rather than tapering off into the security of bonds—which many experts recommend—as I approached my new delayed retirement age. The expert advice works only if you have enough, not if you have to make a mad rush to get enough.

Instead, I was able to take advantage of an opportunity to roll my defined-contribution money into my employer’s defined-benefit pension plan and essentially purchase an annuity at a much better rate than those available on the commercial market. To do that I had to spend four years organizing a rank-and-file union campaign with my fellow defined-contribution plan participants, including AAUP members. It was not easy. It required educating and mobilizing members of a number of different state-employee unions. In the end, though, we won when our public employer, the state of Connecticut, allowed us on a voluntary basis to transfer to the state pension plan and use our accumulations in the defined-contribution plan to purchase credit for time served. I later wrote a book about this process that drew some attention. The rollover raised my employer-plan replacement rate from 24 to 46.4 percent, much better than if I had left it in the defined-contribution plan. The total replacement rate went from 49.6 to 76.2 percent, and I was able to retire.

The possibility of that radical solution exists for any employee in a defined-contribution plan whose employer also has a defined-benefit plan—the situation in many public universities. That is true despite many employees’ being forced out of defined-benefit pension plans and into 401(k)s since the early 1980s. In my case I had been given a supposedly irrevocable choice in 1986, when I had first taken a position at Eastern Connecticut State University, between continuing the TIAA plan I had from previous universities and joining the state pension plan. I assumed (wrongly) that the two would have roughly the same benefits; since I did not yet have tenure, I appreciated that the TIAA balance could be rolled over to another position if necessary. That was the biggest financial mistake of my life. It was a loaded choice that formed part of the basis for a union grievance that we filed and won, paving the path to the solution of being able to change to the much better pension system. It was well worth the effort in terms of retirement-income payoff. In addition to the victory in Connecticut,  public employees, including university faculty and school teachers, in Massachusetts, Florida, and West Virginia have fought for and won similar opportunities to transfer from defined-contribution to defined-benefit plans.

A less ambitious and effective reform would be the creation of a public bank or life-insurance company, perhaps attached to Social Security, that would allow people to transfer their defined-contribution accumulations into nonprofit annuities—a public option in the annuities market. Such an approach was envisioned when Social Security was created in 1935 but abandoned under pressure from financial firms that didn’t want the competition.

The current average commercial annuity payout rate that includes a 3 percent annual cost-of-living increase has decreased from the 5.68 percent rate in existence when I turned sixty-five to 4.5 percent today. At the same time, there are cash-balance retirement plans and a small number of traditional defined-benefit plans that issue annuities from their own internal members’ accumulations. Because these plans are not burdened with the profit needs and other overhead expenses of commercial life-insurance companies, they are able to offer much higher payout rates—generally around 7 percent, which is one-third higher than the commercial rate.

It is legally possible for any employer with both defined-benefit and defined-contribution plans to sell at-cost annuities from the former to participants in the latter, though few currently do.

Making nonprofit annuities available to all defined-contribution plan participants—along with expanding Social Security—would help to reverse some of the damage caused by the wholesale substitution since the early 1980s of those plans for traditional pensions. Current commercial providers of annuities would surely oppose such competition. However, since relatively few retirees with defined-contribution plans annuitize because of the high costs, undermining that market would have relatively little impact on the overall financial market while delivering a much better option for retirement income for 401(k) and other defined-contribution plan participants.

In the end, the defined-contribution approach to retirement provision is a particularly inefficient one compared to defined-benefit plans. A dollar invested in a defined-benefit plan will deliver much more retirement income than one invested in a defined-contribution plan. Financial-service company fees, commissions, profits, and other overhead expenses siphon off much of the potential retirement income from defined-contribution plans. Even more potential income is lost because defined-contribution plans do not have the risk-sharing advantages of social insurance defined-benefit plans.

Until there are national reforms to correct the problems of defined-contribution plans, their participants should at least be aware of the high probability of receiving far less income from them than needed—the situation I was facing—and not be seduced by the deceptive advertising claims of the companies that administer and profit from them.

James W. Russell is university professor emeritus of sociology at Eastern Connecticut State University, an adjunct professor of political science at Portland State University, and the author of eight books, including Social Insecurity: 401(k)s and the Retirement Crisis. He consults with faculty and other employee groups seeking to transfer from defined-contribution to defined-benefit pension plans. His email address is [email protected].