Kevin Drum poses a reasonable question about the existence of a retirement crisis in a recent blog post. He notes that retirement income projections from the Social Security Administration's MINT model show income for older households rising from 1971 to the present, while incomes for those in the age 35 to 44 were nearly stagnant. The model also shows income for older households continuing to rise over the next three decades. Kevin's conclusion is that we are wrong to spend a lot of time worrying about retirees, and would be wrong to consider increasing Social Security taxes on the working population to maintain scheduled benefits for Social Security recipients.
While the story of rising income for retirees is correct, there are several points to keep in mind. First, the main reason that income for the over 65 group has risen is that the real value of Social Security benefits has risen. Social Security benefits are tied to average wages, not median wages. This is important. Most of the upward redistribution of the last three decades has been to higher end wage earners like doctors, Wall Street types, and CEOs, not to profits. Since the average wage includes these high end earners, benefits will rise through time, pushing up retiree incomes. For the median household over age 65, Social Security benefits are more than 70 percent of their income, so the story of rising income is largely a story of rising Social Security benefits.
However, even with this increase in Social Security benefits, replacement rates at age 67 are projected to fall relative to lifetime wages (on a wage-adjusted basis) from 98 percent for the World War II babies to 89 percent for early baby boomers, 86 percent for later baby boomers and 84 percent for GenXers. There are several reasons for this drop. The most important is the rise in the normal retirement age from 65 for people who turned 62 before 2002 to 67 for people who turn 62 after 2022. This amounts to roughly a 12 percent cut in scheduled benefits. The other reason for the drop is the decline in non-Social Security income. This is primarily due to the fact that defined benefit pensions are rapidly disappearing and defined contribution pensions are not coming close to filling the gap.
It is also important that the over 65 population on average has a considerably longer life expectancy today and in the future than was the case in 1971. In 1971 someone turning age 65 could expect to live roughly 16 more years; today their life expectancy would be over 20 more years. This is a good thing of course, but it means that when we use the same age cutoff today as we did 40 plus years ago we are looking at a population that is much healthier, and therefore also more likely to be working, and further from death. If we adjusted our view to focus on the population that was within 16 years of hitting the end of their life expectancy, the story would not be as positive.
The data from the MINT model may also be somewhat misleading because it includes owner equivalent rent (OER) as income. While not having to pay rent is clearly an important savings to an older couple or individual that has paid off their mortgage, it can give an inaccurate picture of their income. There are many older couples or single individuals that live in large houses in which they raised their families. The imputed rent on such a house can be quite large relative to their income as retirees. (Imputed rent is almost one quarter of total consumer expenditures even though only two-thirds of families are homeowners.) There are undoubtedly many retirees who live in homes that would rent for an amount that is larger than their cash income, which will be primarily their Social Security check.
In principle it might be desirable for such people to move to smaller less expensive homes or apartments, but this is often not easy to do. Government policy that hugely subsidizes homeownership and denigrates renting is also not helpful in this respect.
The other part of the income picture overlooked is that almost all middle income retirees will be paying for Medicare Part B, the premium for which is taking up a large and growing share of their cash income. That premium has risen from roughly $250 a year (in 2013 dollars) to more than $1,200 a year at present. This difference would be equal to almost 5 percent of the income (excluding OER) of the typical senior. That means that if we took a measure of income that subtracted Medicare premiums (not co-pays and deductibles) it would show a considerably smaller increase than the MINT data. The higher costs faced by seniors for health care and other expenditures is the reason that the Census Bureau's supplemental poverty measures shows a much higher poverty rate than the official measure.
Finally, there is the need to focus on the question of how well seniors are doing. Seniors income has been rising relative to the income of the typical working household because the typical working household is seeing their income redistributed to the Wall Street crew, CEOs, doctors and other members of the one percent. However, even with the relative gains for seniors, their income is still well below that of the working age population. The median person income for people over age 65 was $20,380 in 2012 compared to a median person income of $36,800 for someone between the ages of 35 to 44. Now we can point to the fact that incomes have been rising considerably faster for the over 65 group, but this would be like saying that we should be annoyed because women's wages have been rising more rapidly than men's wages. Women still earn much less for their work and seniors still get by on much less money than the working age population.
The bottom line is that it takes some pretty strange glasses to see the senior population as doing well either now or in the near future based on current economic conditions. We can argue about whether young people or old people have a tougher time, but it's clear that the division between winners and losers is not aged based, but rather class based.