“Last” Chances

By Nevin Adams, EBRI

While many Americans seem to lack a definitive sense of what living in retirement will be like, how long it will last, or how much it will cost, their sense of when it will begin has been trending older.  The 2012 Retirement Confidence Survey (RCS) noted that, whereas in 1991, just 11 percent of workers expected to retire after age 65, in 2012, more than three times as many (37 percent) report they expect to wait until after age 65 to retire—and most of those indicated an expected retirement age of 70 or older.1

Those expecting to delay retirement perhaps found solace in a recent report by the Center for Retirement Research (CRR) at Boston College which concluded that by postponing retirement until age 70, the vast majority of households (86 percent) were “…projected to be prepared for retirement.”

That sounds good – but what about the assumptions underlying that conclusion?

In 2003, the Employee Benefit Research Institute (EBRI) constructed the EBRI-ERF Retirement Security Projection Model® (RSPM)—the first nationally representative, micro-simulation model based on actual 401(k) participant behavior and a stochastic decumulation model.  And though we have explicitly recognized that many individuals were retiring at earlier ages, a retirement age of 65 was chosen for baseline results, based upon the assumption that most workers would have the flexibility to work until that age, if they so chose.

Last year we modified the RSPM to determine whether just “working a few more years after age 65” would indeed be a feasible financial solution for those determined to be “at risk.”  Unfortunately, for those counting on that as a retirement savings “solution”, the answer is not always “yes.”

Indeed, results from the EBRI modeling indicated that the lowest pre-retirement income quartile would need to defer retirement to age 84 before 90 percent of the households would have even a break-even (50‒50) chance of success.

Working longer does help, of course.  A recent EBRI Notes article titled “Is Working to Age 70 Really the Answer for Retirement Income Adequacy?2” finds that 23 percent of those who would have been at risk of running out money in retirement if they retired at age 65 would be “ready to retire” if they kept working to 70.  Better still, if those individuals are assumed not only to delay retirement, but also to keep participating in a defined contribution plan, a full third of those who would have been at risk of running short of money if they retired at age 65 would be “ready” to retire at age 70.3

What accounts for the difference in the projections?  For a household to be classified as “ready for retirement” under the CRR method, a projected replacement rate is simply compared with a benchmark rate, while the RSPM uses a fully developed stochastic decumulation process to determine whether a family will run short of money in retirement (and, if so, at what age) under each of a thousand alternative, simulated retirement paths.  Unlike the CRR model, EBRI’s RSPM model simultaneously considers the impact of longevity risk, investment risk, and the risk of potentially catastrophic health care costs (such as prolonged stays in a nursing home).5

Which, as it so happens, are the same things that those trying to make sure they have enough money to last through retirement—and those trying to help them do so—need to consider.


1 see “Is Working to Age 70 Really the Answer for Retirement Income Adequacy?

2 Also from the above article, “It’s worth nothing that a significant portion of the improvement in readiness takes place in the first four years after age 65, but that tends to level off in the early 70s before picking up in the late 70s and early 80s.  Higher-income households would be in a much better situation: 90 percent of the highest-income quartile would already have a 50 percent probability of success by age 65, while those in the next-highest income quartile would need to wait until age 72 for 90 percent of their group to have a 50 percent probability. Those in the second-lowest income quartile would need to wait until age 81 before 90 percent of their group had a 50 percent probability of success. 

3 At the same time, the percentage of workers expecting to retire before age 65 has decreased from 50 percent in 1991 to 24 percent (see this EBRI analysis, online here).  A sizable proportion of retirees report each year that they retired sooner than they had planned (50 percent in 2012). Those who retire early often do so for negative reasons, such as a health problem or disability (51 percent) or company downsizing or closure (21 percent).  The 2011 RCS found that the poor economy (36 percent), lack of faith in Social Security or the government (16 percent) and a change in employment situation (15 percent) were the most frequently cited reasons for postponing retirement.

4. For more on how this modeling works, see “Single Best Answer.”

5 For an explanation of four things that are sometimes overlooked by retirement-needs projection models, see “Generation ‘Gaps,’” online here.

Halfway Honed

By Nevin Adams, EBRI


This week we published(1) the results of an update of EBRI’s Retirement Readiness Rating from the Retirement Security Projection Model® (RSPM). That model, which has been modified over the years to take into account certain structural and market changes,(2) projects that more than half (56 percent) of Boomers and Gen Xers will be able to retire with enough money to cover the cost of basic retirement needs as well as uninsured health care costs, including stochastic expenses from nursing home and home health care.(3)

On the other hand, that same model projects that about 44 percent won’t have “enough” to cover those expenses.

It’s worth noting that the trends are positive. Even after the toll of the 2008 financial crisis, the 2012 number of those at risk of running short is some 5−8 percentage points “better” than what was found in 2003. Moreover, the analysis is able to point to some important trends; eligibility for a workplace retirement plan remains a significant factor in reducing the risk of running short,(4) while the more recent broad-based advent of automatic enrollment plan designs makes it ever more likely that those eligible to participate—particularly lower-income workers—do so.

The research does point out that lower-income households are much more likely to be at risk for insufficient retirement income,(5) even though basic retirement expenses are modeled as a function of the household’s expected retirement income. In fact, the 2012 baseline ratings for Early Boomers range from a projection that 87 percent of the simulated lifepaths for the lowest-income households are at risk in retirement to only 13 percent of retired highest-income households.

Obviously, the reality of more than 4 in 10 Americans not having sufficient post-retirement wealth is of concern, though I find that many are pleasantly surprised at how high a percentage is expected to have sufficient assets. Indeed, whether one draws comfort from that finding likely depends on your expectations (and perhaps on which side of that line you think you might find yourself post-retirement).

Regardless of those expectations—and whether you find the picture to be one of a glass half full or half empty—the data give us all something to work with, and to work toward.


(1) EBRI Notes May 2012, “Retirement Income Adequacy for Boomers and Gen Xers: Evidence from the 2012 EBRI Retirement Security Projection Model,®” online here. 

(2) The Retirement Security Projection Model® (RSPM) was developed in 2003, and in 2010 it was updated it to incorporate several significant changes, including the impacts of defined benefit plan freezes, automatic enrollment provisions for 401(k) plans and the recent crises in the financial and housing markets. EBRI has recently updated RSPM for changes in financial and real estate market conditions as well as underlying demographic changes and changes in 401(k) participant behavior since January 1, 2010.

(3) A household’s simulated lifepath in retirement is considered to be at risk in the baseline version of the model if its aggregate resources in retirement are not sufficient to meet aggregate minimum retirement expenditures, defined as a combination of deterministic expenses from the Consumer Expenditure Survey (as a function of income) as well as some health insurance and out‐of‐pocket health‐related expenses, plus stochastic expenses from nursing home and home health care (at least until the point such expenses are picked up by Medicaid). The resources in retirement are assumed to consist of Social Security (status quo benefits for the baseline version of the simulation); account balances from defined contribution plans; individual retirement accounts (IRAs) and/or cash balance plans; annuities or lump-sum distributions from defined benefit plans; and net housing equity (in the form of a lump‐sum distribution at the point that other financial resources are exhausted). This version of the model is constructed to simulate “basic” retirement income adequacy; however, alternative versions of the model allow similar analysis for replacement rates and other thresholds.

(4) For an idea of just much of an impact plan eligibility makes, consider that, according to the simulation results, Gen Xers with no future years of eligibility would run short of money in retirement 60.7 percent of the time, whereas fewer than 1 in 5 (18.2 percent) of those with 20 or more years of future eligibility would run this risk.

(5) In addition to underlining the importance of automatic enrollment for the lowest income, this also underlines the importance of Social Security as a post-retirement income source for this group.