Forty percent are in good shape and know it. One in five are in trouble and know it. Remaining are either not worried enough or too worried, with wealthy in danger of misjudging assets and acting too late.
The National Retirement Risk Index (NRRI) measures the percentage of working-age households that is at risk of being financially unprepared for retirement. The NRRI, developed by the Center for Retirement Research at Boston College during the Great Recession, calculates that even if households work to age 65 and annuitize all their financial assets, including the receipts from reverse mortgages on their homes, roughly half are at risk of being unable to maintain their standard of living.
The NRRI Brief examines whether households have a good sense of their own retirement preparedness – do their expectations match the reality they face? That is, do households at risk know they are at risk? The Brief concludes that almost 60% of self-assessments agree with the NRRI results and that the 40% of households that get it wrong do so for predictable reasons. The issue remains, however, whether unprepared households that recognize their situation are any more likely to take corrective action than those that do not.
The Index is based on the Federal Reserve’s Survey of Consumer Finances (SCF), a triennial survey of a nationally representative sample of U.S. households. The Index calculates, for each SCF household, a replacement rate – projected retirement income as a percentage of pre-retirement earnings – and compares that replacement rate with a target rate derived from a consumption smoothing model. Those who fail to come within 10% of the target are defined as “at risk,” and the Index reports the percentage of all households at risk.
The gap between self-assessed views of adequate preparedness and the NRRI has not always been so large. Prior to 2016, the share of households that self-reported being at risk was relatively consistent with the NRRI and, in fact, slightly higher (see Figure 2). However, in 2016, it dropped substantially.
A potential explanation for this sharp decline is that the SCF question changed in 2016. Prior to 2016, households were asked to assess the adequacy of their retirement income from Social Security and employer pensions, including 401(k)s/IRAs. After 2016, households were asked to consider all sources of retirement income, which could now include housing wealth and other financial assets. After the change, the share of households rating their retirement income as “inadequate” or “totally inadequate” fell from 57 percent in 2013 to 36 percent in 2016 and 34 percent in 2019, with high-income households reporting the steepest drop.
When comparing individual household assessments with the NRRI, 28% think they are not at risk while the NRRI predicts they are (this group is “not worried enough”), and 15% think they will fall short while the model predicts they will have enough (“too worried”).
Results by income show that high-income households – perhaps overreacting to the impact of the strong economy on housing and stock prices – are the most likely to be “not worried enough” and low-income households are the most likely to be “too worried.” The remaining 57% get it right, with 19% correctly rating they are at risk and 38% correctly rating they are not at risk.
What Explains Misperceptions?
The question is what characteristics are associated with a household being “not worried enough” or “too worried,” as opposed to getting it right. The analysis uses regressions to explain the probability of households ending up in a given category using a variety of factors, including: retirement plan participation and account balance, homeownership and housing wealth, risk aversion, self-assessed financial knowledge, education, household type, race/ethnicity, and age. The major reasons for being “not worried enough” or “too worried” are summarized below.
Major Reasons for “Not Worried Enough”
Conceptually, households that were overly optimistic about the economic recovery or overestimated how much income their assets could provide may be more likely to be “not worried enough.” Their overconfidence may lead them to underestimate possible risks. Therefore, it is not surprising that households with higher housing debt-to-asset ratios, relatively low asset balances in 401(k)s and other defined contribution (DC) plans, and two earners but only one saver were more likely to be “not worried enough” (See Figure 4).
Housing debt-to-asset ratio. As the housing market improved, households may have been comforted by the rising value of their asset, without considering how much they still owed. The positive relationship between the housing debt-to-asset ratio and the “not worried enough” group is especially strong for high-income households, who tend to own more expensive homes.
Below median DC balance. Similarly, the danger with DC plan assets is “wealth illusion.” That is, $100,000 looks like a lot of money to many people even though it provides only about $617 per month in retirement income.6 This wealth illusion may have been exacerbated by the strong market performance. Having only a modest DC balance is associated with a higher probability of being “not worried enough” for low and middle-income households.7
Two earners but one saver. Many dual-earner households may not realize they will have to replace both spouses’ earnings to maintain their standard of living in retirement.8 So, not surprisingly, dual-earner households where only one spouse has a retirement plan are more likely to be “not worried enough.” This probability also increases with income because Social Security replaces a smaller share of pre-retirement income for high earners. Black/Hispanic. Black and Hispanic households are also more likely to be “not worried enough,” perhaps due to racial/ethnic gaps in financial literacy.
Major Reasons for “Too Worried” Unlike overly optimistic households, those who are “too worried” are not aware of how much income they will have in retirement and perhaps have less optimism in the asset markets. Characteristics that capture these factors – such as risk aversion, married one-earner households, homeowner, and low self-assessed financial knowledge – predicted households’ likelihood of being “too worried” (see Figure 5).
Risk aversion. Households who are risk averse might be more conservative when judging their financial situation and less likely to be swayed by optimism in the asset markets. Therefore, it is not surprising that risk-averse households are more likely to be “too worried.”
Married one-earner. Single-earner households may not take account of Social Security’s spousal benefit – equal to 50 percent of the benefit of the working spouse – when evaluating their retirement income. Perhaps for this reason, married low-income households with only one earner have a higher probability of being “too worried.”
Homeowner. Homeowners are more likely to be “too worried” because they do not plan to tap their home equity to support general consumption in retirement. The effect is especially large for low-income homeowners because their home represents a much higher portion of their total net worth.
Low self-assessed financial knowledge. Households that rate themselves as having low financial knowledge may be less confident or aware of their financial situation. Interestingly, these households are doing better than they think, as they are more likely to be “too worried.”
Conclusion Overall, the results suggest that households with incorrect perceptions get it wrong for predictable reasons. A little education about the value of various sources of retirement income could reduce the size of the “too worried” group.
Despite research showing households have large gaps in financial knowledge, nearly three out of five have a good gut sense of their financial situation. This share has remained relatively constant despite a 2016 change in the SCF survey. However, classifying households by the accuracy of their perceptions about retirement security does not answer the question of whether they are likely to take remedial action.
Households that are “not worried enough” are the least likely to change their saving or retirement plans. This group accounts for 28% of households, so a significant portion of the population needs to get a better assessment of their retirement income needs. The additional one-fifth of households that do understand their plight may need less convincing to act, but they still must act.