Lane Kenworthy, The Good Society
A century ago, most elderly Americans had little or no economic security. Today, thanks to enhanced access to medical care, public income security programs, pension plans, and asset accumulation, their situation is much improved.
Indeed, by some measures the elderly now enjoy greater economic security than many other groups. Consider the poverty rate. According to the official government data, shown in figure 1, in 1960 roughly 35% of elderly Americans were poor, compared to fewer than 20% of working-aged Americans. Fifty years later the poverty rate among the elderly was below 10%, and it was lower than the rate for the working-aged.1
Have we done enough to ensure economic security for America’s elderly? Or could we — should we — do even better?
For most Americans, income security in old age rests on some combination of five pillars: Social Security and SSI, an employer-based pension, savings, homeownership, and earnings from employment. Let’s consider these one by one.
Social Security has been key to the falling poverty rate among the elderly. Nearly one in three elderly Americans receives 90% or more of their income from Social Security.2 In 1979 the average recipient of Social Security old-age benefits got about $10,000 (in today’s dollars). That average increased steadily over the ensuing three decades, reaching nearly $15,000 as of 2010. During this time the share of elderly Americans receiving Social Security held steady at around 90%.3
Social Security is funded in a “pay as you go” fashion: money paid to current recipients comes directly from the payroll taxes paid by current working Americans.4 As the large baby boom generation retires, the number of elderly recipients per working-age person will rise, putting pressure on this funding arrangement.5 But contrary to frequent assertions by political pundits and journalists, Social Security isn’t in crisis. As figure 2 indicates, the current projection is that its cost will rise by about 1% of GDP between now and 2035 and remain flat thereafter. This can be paid for via some relatively minor changes, such as an increase in the share of earnings subject to the payroll tax (as of 2016, earnings above $118,500 are exempt from the tax), a small increase in the payroll tax rate, a small increase in the retirement age, and/or a reduction in the rate of benefit increase over time.6
Around 5% of elderly Americans receive money from the Supplemental Security Income (SSI) program. SSI provides cash assistance to low-income, low-asset persons who are elderly, disabled, or blind. The program began in 1974. Maximum yearly benefit levels are approximately $8,500 for an individual and $13,000 for a couple. Nearly all of the states supplement these amounts. A significant limitation of SSI for the elderly is its low take-up rate; only slightly more than half of those who are eligible actually receive the benefit.7
The second retirement income security pillar is employment-based pensions. The share of workers who participate in an employer pension plan has remained steady in recent decades at around 45%,8 but the type of plan has changed dramatically. In the early 1980s nearly 90% of Americans with an employer pension plan had a defined-benefit plan with their company. By 2013 that share had shrunk to 30%. These have been replaced by defined-contribution plans such as 401(k)s. Among those with a pension, defined-contribution plans jumped from 38% to 84%.9
Defined-contribution plans have some advantages: they’re portable across employers, the employee has some say in how the money is invested, and a person in financial difficulty prior to retirement age can withdraw some or all of the money, though there is a tax penalty for doing so.10 And pension wealth (as a share of wages and salaries) in defined-contribution plans has increased enough to offset the fall in pension wealth in defined-benefit plans.11 The problem is likely to be in the uneven distribution. Some employees and employers may not contribute enough to defined-contribution plans or keep the money in them long enough.12 If an employee doesn’t know about or understand her firm’s program, or if she feels she needs every dollar of her earnings to pay for current expenses, she may go a long time, perhaps even her entire working career, without putting any money into a defined-contribution plan. Because employer contributions usually take the form of matching funds, with the amount put in by the employer pegged to the amount put in by the employee, no employee contribution often means no employer contribution. Moreover, when a person switches employers she or he has a choice to keep the defined-contribution-plan money as is, roll it over into an individual retirement account (IRA), or withdraw it with a tax penalty subtracted. This creates a strong temptation to withdraw some or all of the money, leaving a lot less, and sometimes nothing at all, for the retirement years.
The third retirement income security pillar is personal savings. It too has weakened. Average household saving as a share of disposable household income fell from 10% in the 1970s to 8% in the 1980s to 5% in the 1990s to 3% in the 2000s.13 The decline probably was even steeper for households on the lower rungs of the income ladder.
The fourth pillar of retirement income security is homeownership. In an ideal scenario, a person or couple will purchase a home by age 35, taking out a 30-year mortgage loan, and by age 65 the loan will be paid off. Having no mortgage or rent payment sharply reduces expenses during the retirement years. Or the house can be sold (or borrowed against via a “reverse mortgage”) to provide an additional source of income. At age 65, 80% of Americans live in a home they own, though not all have fully paid off their mortgage. At age 85, 70% still do. Median net equity among elderly homeowners is $100,000.14
The sharp increase in home prices during the housing bubble from the mid-1990s to the mid-2000s boosted the assets of many middle-class homeowners in or near retirement. But with the bursting of the bubble, all or much or that gain has disappeared for many of them.15
The fifth and final retirement income security pillar is earnings from employment. Currently, 19% of Americans age 65 and over are employed, up from 14% a decade ago. Many work part-time.16
What’s the bottom line on retirement income security? The best information on current and future income trends is from the Social Security Administration’s MINT (Modeling Income in the Near Term) projections.17 Figure 3 shows projected income for retirement-age households in the bottom fifth of incomes (“Q1”) and in the next fifth (“Q2”). Among households in the bottom fifth, median income for those age 65 in 2007 was $11,000. This is projected to rise to $14,000 in 2037. Among households in the lower-middle fifth, the projection is for a rise from $24,000 in 2007 to $28,000 in 2037. So the best available estimates suggest that despite falling savings and the shift from defined-benefit to defined-contribution employer pensions, the incomes of elderly Americans aren’t likely to decline in absolute terms. Instead, they will increase.18
On the other hand, figure 3 also shows that these income levels are projected to rise less rapidly than GDP per capita.19 Retirement incomes for many ordinary Americans will, in other words, fall farther and farther behind growth of the economy. Over the long run, that’s a potentially worrisome trend.
How can we do better? Social Security benefits could be increased.20 A modest, gradual rise over time is appropriate as the economy grows, and it surely is affordable. But this is only a partial solution. As figure 4 indicates, with Social Security alone, the income replacement rate for a typical American retiree is fairly modest by the standards of the world’s rich countries. We need one or more of the other retirement income security pillars to increase as well.
We could try to encourage more saving. But previous attempts, such as offering tax advantages (IRAs), have had little impact on the savings behavior of most ordinary Americans.21
We need to shore up employment-based pensions. There is no going back to the defined-benefit past; for most Americans, a pension in the future will be a defined-contribution one. Rather than allow Americans to contribute to defined-contribution plans if they have a steady job and if their employer offers a plan and if they know about it and if they feel they can afford to put some of their earnings in it, we could make contributing the default option and make it available to everyone. Employers that have an existing plan could continue it, but they would have to automatically enroll all employees and deduct a portion of earnings unless the employee elects to opt out. Employees who lack access to an employer plan would be automatically enrolled in a new universal retirement fund, and those who lack an employer match would be eligible for matching contributions from the government.22
In the absence of federal government action along these lines, a few states have created their own programs. California’s Secure Choice Retirement Savings Program, the largest of these, requires firms with five or more employees to enroll them if it doesn’t offer a company-sponsored pension program. The program contributes 3% of each paycheck, unless an employee chooses to opt out. But there is no matching contribution from the employer or the state.23
We also could facilitate greater employment among the elderly. Figure 5 shows the trend in the employment rate among Americans age 65 and over. The rate dropped steadily in the 1960s, 1970s, and 1980s. Since the late 1990s it has slowly but steadily risen.24 Later retirement isn’t a good option for everyone, of course, especially those who have spent most of their working lives in stressful or physically taxing jobs. But for those who can manage it, it is doubly beneficial: it provides an additional source of income, and it allows people to delay receipt of Social Security, which in turn increases the benefit level they receive.
Medical care is another key determinant of economic security in old age. Health insurance is a must, as medical care is too expensive to be purchased out of pocket — particularly for the elderly, who tend to need more expensive care and have less income than working-aged adults. Nearly all elderly Americans are enrolled in Medicare.25 Most also have additional supplementary insurance for medical expenses that Medicare doesn’t cover.26 For most, the cost of Medicare enrollment is about $1,200 per year.27
For elderly Americans, Medicare pays, on average, about 65% of total medical expenses. Medicaid and supplemental health insurance plans pay about 20%. The remaining 15% comes from out-of-pocket payments.28 The elderly spend an average of 15% of their income on medical expenses.29
Medicare is a very popular program. And it has proven effective at helping to ensure economic security among older Americans. Indeed, it has largely eliminated financial hardship due to the cost of medical treatment among the elderly. However, Medicare also is expensive. Its cost has risen from 0.5% of GDP in 1970 to 3.5% in 2012, and as figure 6 shows, the cost is projected to continue increasing rapidly, reaching 6% of GDP in 2060. About half of this rise is a product of the same demographic development — the baby boom generation reaching retirement age — that is responsible for the coming increase in Social Security expenditures. The other half of the projected rise is a function of the increasing cost of healthcare in general.30
In all likelihood, the projected increase in Medicare costs is affordable. Continued reduction in the cost of food and manufactured goods will free up money we as a society can use to pay for medical care.31
But there also are grounds for optimism about our ability to reduce the rate of growth of Medicare expenditures, and to do so in a way that maintains or improves economic security for recipients. (Proposals for Medicare reform that don’t meet this criterion include increasing the age of Medicare eligibility and shifting to “premium support.”32) Healthcare experts largely agree that the problem owes to excessive treatment and overpricing by medical care providers, and we have begun to address this.33 Since the late 1990s Medicare costs per recipient have risen less rapidly than costs for those with private insurance. And the 2010 Patient Protection and Affordable Care Act (PPACA) created a number of mechanisms to slow the growth of Medicare costs. It reduces Medicare “overpayments” (certain types of payments to providers and subsidies to private insurers that offer Medicare Advantage Plans). It encourages a switch in payment from “fee-for-service” to “accountable care,” whereby physicians and hospitals are paid according to their effectiveness in delivering successful outcomes while controlling costs. It penalizes hospitals if too many of their patients end up being readmitted soon after release. Early results are encouraging: expenditures per Medicare beneficiary have been flat for the past three years. And there is more to come: the PPACA authorized creation of an Independent Medicare Advisory Board (IPAB), which will be able to limit the rate of growth in Medicare spending without needing approval by Congress.34
The economic security of America’s elderly has improved immensely since the 1930s. In some respects that improvement has continued in recent decades. But two significant challenges loom: ensuring that the incomes of elderly Americans in the bottom half don’t fall farther and farther behind the growth of the economy, and slowing the rise in Medicare costs.
- The official poverty rate measure has an assortment of deficiencies. But it’s not a bad measure, and it has the significant advantage of being available over a long stretch of time. ↩
- Center on Budget and Policy Priorities, “Top Ten Facts about Social Security,” 2016 ↩
- Annual Statistical Supplement to the Social Security Bulletin, tables 3.C4 and 3.C5. ↩
- Half the payroll tax is paid by employers. ↩
- Today there is one OASDI beneficiary per three employed persons. By 2030 this will rise to nearly one for every two workers. Source: 2013 OASDI Trustees Report, “Table IV.B2. Covered Workers and Beneficiaries, Calendar Years 1945-2090.” ↩
- Peter A. Diamond and Peter Orszag, Saving Social Security: A Balanced Approach, Brookings Institution Press, 2004; Kathy A. Ruffing, “What the 2011 Trustees’ Report Shows About Social Security,” Center on Budget and Policy Priorities, 2011; Henry Aaron, “Progressives and the Safety Net,” Democracy, 2013. ↩
- Kathleen McGarry, “The Social Safety Net for the Elderly,” in The Legacy of the War on Poverty, edited by Martha Bailey and Sheldon Danziger, Russell Sage Foundation, 2013. ↩
- Center for Retirement Research, “Pension Participation of All Workers, by Type of Plan, 1989-2013.” ↩
- Center for Retirement Research, “Workers with Pension Coverage by Type of Plan, 1983, 1992, 2001, and 2013,” 2009, using data from the Survey of Consumer Finances. The numbers sum to more than 100% because some people are enrolled in both types of plan. ↩
- Another potential advantage is that a defined-contribution plan isn’t vulnerable to a firm going out of business or declaring bankruptcy. Such an occurrence can leave those in a traditional company pension plan out in the cold, though most such pensions are guaranteed by the federal government-backed Pension Guarantee Benefit Corporation. ↩
- Alicia H. Munnell, Jean-Pierre Aubry, and Caroline V. Crawford, “How Has Has Shift to Defined Contribution Plans Affected Saving?,” Center for Retirement Research, 2015, figure 7. ↩
- “The Trouble with Pensions,” The Economist, 2008; Teresa Ghilarducci, When I’m Sixty-Four, Princeton University Press, 2008; Edward N. Wolff, The Transformation of the American Pension System, Upjohn Institute, 2011; Michael A. Fletcher, “401(k) Breaches Undermining Retirement Security for Millions,” Washington Post, 2013. ↩
- These decade averages are my calculations using OECD data. ↩
- Sudipto Banerjee, “Own-to-Rent Transitions and Changes in Housing Equity for Older Americans,” Employee Benefit Research Institute, 2012, using data from the University of Michigan’s Health and Retirement Study. ↩
- Edward N. Wolff, “The Asset Price Meltdown and the Wealth of the Middle Class,” New York University, 2012. ↩
- The average weekly hours of employment is 30. See Steven Greenhouse, “At Leisure, or Still at Work,” New York Times, 2013, using data from the American Time Use Survey. ↩
- Barbara A. Butrica, Karen E. Smith, and Howard M. Iams, “This Is Not Your Parents’ Retirement: Comparing Income Across Generations,” Social Security Bulletin, 2012. ↩
- See also Kevin Drum, “My Social Security Reform Plan,” Mother Jones, 2016. ↩
- Between 2007 and 2037, according to these projections, GDP per capita will increase by 60%, bottom-fifth household incomes for age-65ers will increase by 27%, and lower-middle-fifth household incomes will increase by 16%. ↩
- Thomas Geoghegan, “Get Radical: Raise Social Security,” New York Times, 2011; Michael Lind, Steven Hill, Robert Hiltonsmith, and Joshua Freedman, “Expanded Social Security,” New America Foundation, 2013. ↩
- Ghilarducci, When I’m Sixty-Four. ↩
- Michael J. Graetz and Jerry L. Mashaw. True Security: Rethinking American Social Insurance, Yale University Press, 1999; Alicia H. Munnell, “Bigger and Better: Redesigning Our Retirement System in the Wake of the Financial Collapse,” in Shared Responsibility, Shared Risk, edited by Jacob S. Hacker and Ann O’Leary, Oxford University Press, 2012; U.S. Senate Committee on Health, Education, Labor, and Pensions, “The Retirement Crisis and a Plan to Solve It,” 2012. ↩
- Chris Farrell, “No 401(k)? Your State May Come to the Rescue,” NextAvenue, 2016; Mary Williams Walsh, “California Aims Retirement Plan at Those Whose Jobs Offer None,” New York Times, 2016. ↩
- The United States has one of the highest elderly employment rates among affluent nations. ↩
- Medicare Part A covers hospital costs. Part B covers other medical costs, including outpatient medical services at hospitals (including emergency-room visits), observed-status services at hospitals even if confined to the hospital overnight or longer, surgical center procedures, doctor services (including most doctor services performed during the admitted hospital stays covered under Part A as well as office visits), a limited number of drugs, and some durable medical equipment. Part C is “Medicare Advantage,” private plans run through Medicare. Part D covers prescription drugs. ↩
- Medicare has no annual or lifetime limit on out-of-pocket costs, there is a 20% co-pay for most Medicare Part B services, and there is a deductible of approximately $1200 for a hospital stay. Sources of supplementary insurance vary: some have employer health insurance that continues in retirement, others have Medicaid, and others purchase private supplemental insurance (“Medigap”). ↩
- Most elderly receive Medicare Part A coverage at no cost and pay about $100 per month for Medicare Part B. ↩
- Centers for Medicare and Medicaid Services, Medicare Current Beneficiary Survey, 2008, “Table 4.1. Personal Health Care Expenditures for Medicare Beneficiaries, by Source of Payment and Type of Medical Service.” ↩
- This compares to 5% for people under 65. See Paul Starr, “The Medicare Bind,” The American Prospect, 2011. ↩
- 2013 Medicare Trustees Report. ↩
- William J. Baumol, The Cost Disease, Yale University Press, 2012. ↩
- See Ezekiel J. Emanuel, “For Medicare, We Must Cut Costs, Not Shift Them,” New York Times: Opinionator, 2011; Paul Starr, “The Medicare Bind.” ↩
- Jonathan Cohn, “What Really Ails Medicare,” The American Prospect, 2008; Emanuel, “For Medicare, We Must Cut Costs, Not Shift Them”; Sarah Kliff, “Should the Doc Fix Get Fixed?,” Wonkblog: Washington Post, 2011; Starr, “The Medicare Bind”; Peter Orszag, “Healthcare Goal Should Be Quality, Not Quantity,” Bloomberg, 2012; Michael Levine and Melinda Buntin, “Why Has Growth in Spending for Fee-for-Service Medicare Slowed?,” Working Paper 2013-06, Congressional Budget Office, 2013; Eduardo Porter, “Medicare Needs Fixing, but Not Right Now,” New York Times, 2013. ↩
- Congress can overrule IPAB decisions, but in order to do so it must come up with alternative mechanisms to achieve the same amount of cost control. ↩