If you think about a retirement benefit as an in-kind benefit – e.g., every employee needs enough money to replace 70-80% of her preretirement income – then the cost of providing that benefit has doubled in the last 13 years. Why? Because interest rates have declined by more than 400 basis points since 2000.
The impact of this change isn’t subtle. A recent Employee Benefits Research Institute (EBRI) study found that, compared with historical returns, in a long-term low interest rate environment the number of workers that will “have sufficient retirement resources to cover 100% of the simulated retirement expenses” drops by about 10 percentage points. The impact is greater the longer the worker has to retirement: longer periods of low returns = lower benefits. And the impact is greater as you go up the income scale: (oversimplifying somewhat) Social Security benefits are largely unaffected by interest rate changes.
These results are intuitive. Obviously, they are very assumption-dependent: What is an adequate retirement income (EBRI uses an elaborate, bottom up model)? What are reasonable asset allocation and return assumptions (EBRI uses a 6% equity premium)? How do you determine ‘real’ values (basically, how do you account for inflation effects)? But there’s no question, declining interest rates are squeezing retirement savings, and something has to give.
So, what’s Plan B? In real life, there are a limited set of options. You can save more — but with the cost of health care and education going up as well, it’s hard to see exactly where the average worker is going to get the extra savings. You can decide to live on less. Or you can work longer. There aren’t a lot of other choices.
Does working longer actually solve the problem? Maybe. A recent paper by the Center for Retirement Research found that “over 85% of households would be prepared to retire by age 70. Thus, many individuals will need to work longer than their parents did, but they will still be able to enjoy a reasonable period of retirement, especially as health and longevity continue to improve.” That result also is intuitive: if you work longer you save more and earn more on your savings, and the period over which your savings have to be paid out is shorter.
But, there is a key, emerging element of retirement income adequacy that the CRR paper doesn’t capture. In its paper CRR takes a ‘replacement rate’ approach to retirement income adequacy, generally defining adequacy in retirement as a percentage of preretirement income. This in effect assumes a ‘smooth’ rate of spending in retirement (e.g., the 78% preretirement income replacement target recommended by Aon Consulting and Georgia State University Replacement Ratio Study (2008)).
As an EBRI review of the CRR analysis points out, however, there are retirement expenses that are, in effect, fixed — they don’t go down just because you’re working longer and retiring less. As Americans live longer, end-of-life expenses – health care and, especially, long-term care – are becoming a bigger part of the ‘adequacy equation.’ Working longer doesn’t reduce those costs, unless you work until you die.
Since implicit assumptions about inflation play such an important role in these analyses, it’s worth considering it’s impact on income, wealth and retirement adequacy. For someone living at a subsistence level (e.g., EBRI’s bottom income quartile), current CPI numbers may make sense – they may reflect the actual impact of price changes on their lives. And, certainly, increases in the cost of education also have a direct effect on the wallets of those workers sending their kids to college.
But consider health care. How much of the increase in health care costs is just paying for better quality? Maybe that’s a hard question to answer, but in other areas – most obviously, technology – it’s an easy one. Some things really are getting cheaper. That is to say that what inflation means to you depends on what you’re buying. And, thus, for some Americans, the ‘real’ (adjusted for inflation) rate of interest may actually be much higher than reported, because ‘actual’ inflation is much lower (even, perhaps, negative). Thus, for some, ‘living on less’ may not be all that bad.
It’s conceivable that ‘real’ rates will go back up significantly and for the long term (they have moved up about 1% since 2012). However, historically low rates may be around for a while: our aging population and the world’s demonstrated preference for the dollar vs. other currencies being two big reasons why. For many Americans, working longer will be a no-brainer – they may even enjoy it. Others may dislike work so much they will opt for living on less. Many will have to do both.