Savings & Investing

Spending in Retirement…or not?

Feb 15, 2018
By Bruce Wolfe, CFA, Robert Brazier

Who ever thought spending
retirement assets would be so
difficult?…today’s retirees apparently

Something unexpected has been the shared experience for our most recent generation of retirees. The vast majority haven't been spending their retirement savings—leaving nest eggs mostly untouched and living on ready sources of income instead. However, future retirees may be less fortunate.

While on the surface this is indeed good news—and appears to support the argument that fears of a future retirement crisis are overstated, the conditions that supported this spending and savings behavior are unlikely to persist. Future retirees will face a much different retirement landscape and will need to adopt new sets of skills—behavioral and financial—that will help them systematically tap into retirement savings to support future spending.

Financial industry norms and academic theories have always assumed assets accumulated for retirement would be systematically withdrawn—following the “4% rule” or some other rule of thumb or system—by retirees in order to maintain a consistent standard of living. Technically, this is referred to as “consumption smoothing” whereby individuals seek to have consistent spending on par with pre-retirement levels. With concerns that retirement savings for individuals may be dangerously low,1 the fear has been that withdrawals for such smoothing could leave retirees running out of funds well short of their passing away.

This research conducted by the BlackRock Retirement Institute (BRI) in conjunction with the Employee Benefit Research Institute (EBRI) found that on average across all wealth levels, most current retirees still have 80% of their pre-retirement savings after almost two decades in retirement.

Four key takeaways:

  1. These findings begin to challenge industry norms and academic theories about lifecycle consumption especially during the retirement phase
  2. Across all wealth levels measured, more than one third of current retirees grew their assets—leaving considerable potential consumption on the table
  3. Late in life out-of-pocket medical expenses—a major reason to retain assets—do not appear to be warranted except for a very small portion of the population
  4. The financial landscape for future retirees will most likely be more challenging, requiring different savings and spending behaviors

Download summary version

This paper sets to lay the foundation for how retirees have managed their sources of cash—assets and income—against their spending behaviors. The resulting “husbanding” of assets over the past two decades may be due to a host of favorable environmental factors current retirees benefited from during their working, accumulation years. These included beneficial changes to Social Security and Medicare, a relatively high percentage of jobs that offered defined benefit pensions, strong real estate appreciation and an investment market that generally delivered strong returns and high interest rates. Has the confluence of these factors created a situation whereby retirees may not have felt the pressure to draw down principal from retirement savings in order to maintain a reasonable standard of living? Perhaps retirees had other plans for their assets beyond themselves—bequests or charitable donations come to mind. Possibly they would have preferred to spend more freely but lacked the financial confidence or tools to efficiently decumulate their assets or were worried about end-of-life healthcare expenses? Looking further, perhaps there were strong emotional biases at play—with fear of outliving retirement assets at the top of the list.2


The objective of the study was to analyze the “how” and some possible “whys” spending and liquid assets change during retirement, taking into account (non-housing) assets, income, spending, out-of-pocket medical expenses and bequests. Data was collected from the bi-annual Health and Retirement Study (HRS, 1992-2014) and the Consumption and Activities Mail Survey (CAMS, 2005-2015). A sample of 7,148 retiree households provided self-reported asset data and out-of-pocket medical expenditure and a subsample of 1,660 households provided the household expenditure data. Retirees were segmented into three groups based on pre-retirement non-housing retirement assets — $0 to less than $200,000 (lowest wealth), $200,000 to less than $500,000 (medium wealth) and $500,000 and above (highest wealth).

Retirement assets mostly resilient
over time

Looking first at household asset trends through retirement, we measured median non-housing assets right before retirement to a maximum of 17-to-18 years into retirement. Figure 1 illustrates that retirement assets remained remarkably steady across the time period. Yes, the wealthiest group retained the most assets (83%), but the medium and lowest wealth groups also retained a strong percentage of assets over the same period (77% and 80%, respectively).

This supports prior research3 that suggests households tend to preserve retirement assets, with rates of returns on those assets often exceeding withdrawals, resulting in asset balances for many retirees growing through at least 85 years old.

Figure one - Median non-housing household assets4

 Chart : Median non-housing household assets

Source: Employee Benefit Research Institute estimates based on Health & Retirement Study (HRS, 1992-2014) Consumption and Activities Mail Survey (CAMS, 2005-2015)

Figure two - Percent of assets remaining after 17-to-18 years of retirement5

 Chart : Percent of assets remaining after 17-to-18 years of retirement

Source: Employee Benefit Research Institute estimates based on Health & Retirement Study (HRS, 1992-2014) Consumption and Activities Mail Survey (CAMS, 2005-2015)

However, medians don’t mean everyone is not spending down, or adding assets, and Figure 2 provides a fuller picture behind the averages. We separated our three asset level groups further in terms of percentage of assets remaining after 17-to-18 years in retirement— those with less than 20%, 20 to 50%, 50 to 80%, 80 to 100% and more than 100%. Surprisingly, over one third of households across each wealth group had more assets after 17-to-18 years in retirement than initially. At the other end of the spectrum, those who spent down to less than 20% of initial assets was approximately 16% and 12%, for the medium and highest wealth groups, respectively. However, for the lowest wealth group, there was approximately an equal set of retirees with less than 20% of assets as those with over 100% of initial assets. So while most—particularly the two wealthiest segments— are doing well enough to grow or minimally dip into savings principal, a smaller group across the wealth spectrum are spending down. For them, this spending down could represent a steadier drawdown consistent with systematic decumulation of assets. For others this spending down may have been unplanned and ad-hoc after suffering one, or more, financial shocks or unexpected expenses, ranging from a death of a spouse, divorce, home repair, family or medical emergency.6

Retirement income: the sources and differences across wealth groups

If most current retirees aren’t aggressively drawing down assets, then what are their other main income sources? Outside of retirement savings principal, income in retirement is generally derived from four sources: labor, capital7 , pensions, and Social Security, as illustrated in Figure 3. For the lowest wealth group, Social Security contributes by far the largest percentage of household income, followed by pensions, with labor and capital contributing a negligible amount.

Figure three - Sources of household retirement income

Chart : Sources of household retirement income

Source: Employee Benefit Research Institute estimates based on Health & Retirement Study (HRS, 1992-2014)

What is not captured here are the various government food and housing subsidies often available to this group. For the medium wealth group, Social Security also contributes the largest percentage (although less so compared to the lowest wealth group), with income from pensions playing a more substantial role (and the largest percentage for the three groups) followed by income thrown off by capital and a very small percentage from labor. For the wealthiest group, Social Security again is the largest contributor with income from capital second (the largest percentage for the three groups) and pension income third. Pensions also provide the opportunity for payouts at retirement in the form of lump sums, a strategy more likely credited to retirees in the higher asset wealth levels (potentially adding to their capital income bucket).

Figure 4 illustrates the gross, total household income over time for the three wealth groups, including the four sources previously mentioned as well government transfers, alimony, insurance payouts and inheritance. After an expected drop at retirement, income remains mostly steady for the three groups over the course of the 18 years measured. In terms of after tax replacement ratios—net income in retirement divided by net income preretirement—estimated ratios for the three wealth groups are slightly lower8 but basically in line with the 60 to 70% replacement ratio conventional wisdom states is needed for most people to maintain their standard of living in retirement. Being that our research is based on self-reported data, the replacement rates we observed may actually be higher. Other research9 suggests that retirement income is oftentimes underreported in government surveys and actual replacement rates are healthier than widely thought. Taken as a whole, these historical income patterns appear to align with a low need to tap retirement savings principal in order to sustain reasonable spending levels.

Figure four - Median household gross income—just before and during retirement (Measured in 2015 dollars)

 Chart : Median household gross income—just before and during retirement

Source: Employee Benefit Research Institute estimates based on Health & Retirement Study (HRS, 1992-2014) Consumption and Activities Mail Survey (CAMS, 2005-2015)

Most retirees are spending in line
with income

Looking at spending and income together, overall spending slowly and steadily declined with income, and the highest wealth group showed the largest spending drop over time (Figure 5). Retirees in the lowest wealth group displayed a more bumpy ride, with an initial increase in spending in years one and two— possibly from health expenses and/or an earlier than expected retirement, and again at years seven and eight—possibly when Required Minimum Distributions (RMDs) kicked in. Other research into changes in household spending during retirement suggest that such steady declines are fairly common,10 with certain spending categories such as healthcare often increasing, while others such as transportation and entertainment tending to go down. This slow decrease in spending matched what would be expected with an ageing lifestyle. Most retirees likely didn’t require major adjustments nor additional income from retirement assets to meet more modest lifestyle needs.

Few retirees ran household
budget deficits

For the minority of current retirees whose spending had exceeded income sources, the magnitude of the deficit might shed light on the ability for these retirees to still refrain—or not—from dipping into principal.

Figure five - Median annual pre-tax income and spending11

Chart : Median annual pre-tax income and spending1 - Lowest wealth

Chart : Median annual pre-tax income and spending1 - Medium wealth

Chart : Median annual pre-tax income and spending - Highest wealth

*Measured in 2015 dollars

Figure six - Percentage of households over the age of 65 whose spending exceeded income

Chart : Percentage of households over the age of 65 whose spending exceeded income

Source: Employee Benefit Research Institute estimates from Health and Retirement Study (HRS) and Consumption and Activities Mail Survey (CAMS)

Figure 6 illustrates that the percentage of retirees outspending their income was relatively low, with less than 18% of retirees across the three wealth levels outspending their household income over the time period (This figure also shows a strong uptick in outspending for the two wealthier groups around the time of the 2008 - 2009 financial crisis—most likely due to a drop in capital income). Figure 7 shows that for those who did spend above their income, their spending gap as a percentage of initial nonhousing assets was reasonably low. It should be noted that the lowest wealth group generally receives subsidies, which helps bridge their larger spending gap. One possible relief valve for over-spenders in the highest wealth group and to a lesser degree the medium group could have been drawing appreciated capital from their investment portfolio beyond what we are seeing in the median numbers.

Figure seven - Spending gap for households over the age of 65*

Chart : Spending gap for households over the age of 65

Source: Employee Benefit Research Institute estimates from Health and Retirement Study (HRS) and Consumption and Activities Mail Survey (CAMS).
*As a percentage of initial non-housing assets

During the 2005-2013 period, investment returns for a conservative 20% equity-80% fixed income portfolio and a more aggressive 60%-40% portfolio delivered annualized returns of 5.3-6.4% respectively. Looking at the time period for the entire study (1992-2015) similar portfolios delivered annualized returns of 6.6-8.0%, respectively.12

Accessing these appreciated assets could meaningfully contribute to any shortfall without reducing principal. Further analysis here is needed, but it’s within reason to suggest that even for these “over-spenders,” the need to dip into principal to fund a deficit was minimal—if at all.

Why are current retirees spending perhaps less than anticipated?

It would appear that for most retirees, keeping up with the day-to-day expenses of retirement isn’t requiring them to dip into retirement capital. While this sounds like good news for those worried that we might already be mired in a retirement crisis, why then aren’t retirees loosening the purse strings more on their retirement assets to fund additional discretionary spending? We looked at two common reasons often attributed to holding on to retirement savings.

Concerns about end-of-life care expenditures

One of the greatest financial fears for people in retirement can be the cost of long-term care associated with a major medical procedure, sharply declining health or treatment for cognitive disorders—particularly in the last year or two of life. Looking at Figure 8, our research suggests however, that out-of-pocket medical expenses were quite low for the vast majority of retirees during this period.

Figure eight - Out-of-pocket medical expenses in the last 1 to 24 months before death, by age at death13

Chart: Out-of-pocket medical
expenses in the last 1 to
24 months before death,
by age at death

It’s not until the 95th percentiles of those surveyed that out-of-pocket expenses jumped to more significant levels. Even then, it could be argued that for someone facing such an acute medical situation late in life, they would unlikely be spending money on much else. For them, most if not all of their income—social security, pensions, investment income—would be diverted to full-time medical care, potentially lessening the need to dip deeply into retirement assets.

The motivation to leave money behind

Another potential reason why current retirees may not be spending down their retirement assets is the high percentages of households— across all wealth groups—that appear to be interested in leaving a bequest. However, one study14 has found that leaving an inheritance (beyond money to support a spouse) is not a strong motive. Only 18% of people age 68-80 with investible assets of at least $200,000 think leaving a bequest is important to them and 73% state that their bequest will be whatever is left over at time of death. According to Figure 9, which measures the dollar amount of bequests including homes retirees are expecting to leave behind, the “whatever is left” is proving to be a significant amount of retirement assets,15 particularly among the medium and highest wealth households.

Figure nine - Average self-reported probabilities of leaving inheritance among households above the age of 65

Chart: Average self-reported probabilities of leaving inheritance

Source: Employee Benefit Research Institute estimates from Health and Retirement Study (HRS)

Looking back: most didn’t need to or didn’t want to spend savings

Most retirees in our study appear to have coped and managed pretty well in retirement. Many could have afforded to withdraw a little and, in some cases, a lot more from their retirement accounts but chose not to, potentially leaving in some cases large amounts of hard-earned savings unspent.

While many might find this puzzling, research suggests16 that people would rather not touch their savings and instead adjust their lifestyle, making cutbacks where necessary and shifting to “needs” over “wants.” Others may feel the need to hold on to wealth as a form of self-insurance instead of acquiring an annuity to deal with a number of life’s uncertainties, such as longevity risk. Retirees may also hold back due to deeper behavioral biases or tendencies. After being told to “save, save, save,” for decades, the idea of shifting to “spend, spend, spend” underweights the power of inertia and the comfort associated with the status quo.17 The common framing of decumulation as a time to withdraw or remove assets rather than say gaining new experiences faces a strong “loss aversion” bias as well.18 Even the uncertainty or ambiguity regarding longevity itself can lead people to select more certain, but possibly sub-optimal decisions.19 These biases can be exacerbated given the person’s risk tolerances, experience (or lack of) with the investment industry and investing and the overall saving/spending relationship (often family influenced). Retirement planning and financial advice that acknowledges and incorporates solutions to these types of biases could help mitigate behaviors getting in the way of retirees spending a bit more on themselves and using those assets saved over decades.

Looking forward: need to spend down retirement assets may
only increase

Many of the retirees captured in this research were fortunate to be able to maintain a reasonable standard of living without significantly tapping into their retirement savings principal. Future retirees may not be so lucky. Many will likely retire into an environment with multiple headwinds and face growing pressure to save more and maximize the value of their entire retirement savings—principal and all—unless they are willing/able to make dramatic cuts in their retirement lifestyle. Several major challenges rise to the surface as we look ahead (Figure 10):

  • Pension benefits—on average, 42% of the retirees tracked in the research received income from a defined benefit (DB) pension: few, if any, of those retiring over the next 10- 20 years can expect income from a DB plan.
  • Social Security—income from Social Security is the largest component in the retirement income mix for all retirees, but pressure on Social Security finances could lead to a future drop in benefits.
  • Tax implications—most retirement assets for those working in the 1970’s and 1980’s were post-tax, i.e., before tax-qualified vehicles emerged. Most future retirement savings are in tax-qualified vehicles and would need to be tax-adjusted upon distribution, further reducing income.
  • Rates of return—over the past 35-plus years asset classes have delivered robust returns in the form of asset appreciation and interest income; few asset classes are expected to perform at the same levels into the near future.
  • Savings behavior—on a more qualitative level, deeply entrenched saving habits can impede retirees from getting comfortable with the notion of depleting their “nest egg.” Future retirees will need to save more and be more confident around drawing down retirement assets—or else be prepared for potentially significant belt tightening.
  • Longer life span—people are living longer and will need to have their retirement assets last longer, in some cases much longer. Investment portfolios should be re-assessed in light of this longer time horizon, and consider further diversification into less liquid, higher risk premia assets.

Figure ten - Mounting obstacles to retirement success

Chart: Mounting obstacles to retirement success


Shifting demographics and a more challenging market environment will only elevate the complexity and importance of helping retirees maximize the value of retirement savings. Future retirees will face obstacles not seen by prior generations and many of the apparent behavioral biases possibly holding back current retirees from spending will be at play among future retirees as well. Whether they can gain the confidence to spend retirement assets if and when needed—or not and potentially see major adjustments to their lifestyle instead—remains to be seen.

But the good news is that with improved savings behavior, steady and consistent investing, and sound guidance on retirement income, future retirees can take the steps necessary towards a comfortable standard of living. Such guidance can come from a financial advisor who may need to expand their role as a fiduciary to include prodding systematic withdrawal of assets by their retirees. Defined contribution platforms can also be a familiar source for guidance and deliver products designed for both accumulation and decumulation—helping retirees maintain consistent spending in retirement, while providing a seamless transition from the savings phase. While the intent of this paper was not to further fan the flames of the retirement crisis debate, the research and analysis of this study is a step towards better understanding an important gap in knowledge about the financial behavior of American retirees. Further analysis and study into the underlying motivations behind the numbers could be the next step towards closing that gap.

Why BlackRock

BlackRock helps people around the world, as well as the world’s largest institutions and governments, pursue their investing goals. We offer:

We work only for our clients, who have entrusted us with managing $5.97 trillion*, earning BlackRock the distinction of being trusted to manage more money than any other investment firm in the world.

Download now

Bruce Wolfe, CFA
Executive Director, BlackRock Retirement Institute
Bruce Wolfe, CFA, Managing Director, is a member of the US & Canada Defined Contribution (USDC) Group. Bruce is the Executive Director of the BlackRock ...
Director, BlackRock Retirement Institute
Robert Brazier is to the Content Developer and Manager for the BlackRock Retirement Institute where he supports the team's thought leadership efforts.