About a decade ago, my mother’s slow mental decline became too obvious for our family to deny. She continued to live at home with my father under increasingly difficult circumstances until she fell and broke her hip. In the hospital, it became clear that her mental impairment precluded physical rehabilitation and that institutional care was unavoidable.
This is a standard baby boomer’s saga, and what came next was not unusual either. I soon learned that long-term care for institutional residents was a big-ticket item. My parents lived in Connecticut, a high-cost state. The place we chose—a “continuum of care” facility with independent living at the top and a nursing home at the bottom—cost more than $100,000 per year for a semi-private room. (The national average at the time was about $70,000, and it has since risen to $78,000, according to a Metropolitan Life survey.)
Although I had spent much of my life studying public policy, I had no idea how long-term care was financed. I soon learned that Medicare paid for at most 100 days of rehabilitation (useless in my mother’s case) and that Medicaid required beneficiaries to “spend down” nearly all their assets. Private long-term care insurance policies were available, I learned, but my parents—along with most Americans who can afford them—had not purchased one. Fortunately they had lived below their means for decades and had accumulated substantial assets, which proved sufficient to see my mother through nearly five years of full-time care.
I didn’t need to study wealth distribution tables to see that only a tiny fraction of American families could afford to do what my parents had done. The median family could self-finance only a few months of institutional care, after which they would be completely dependent on public resources. But Medicaid is devouring ever-increasing shares of hard-pressed state budgets, and huge federal budget deficits are putting pressure on the decades-old fiscal partnership between the states and the national government, and the pressure will only intensify in the decades ahead.
In this presidential election year, the impact of demographic change—especially the growing weight of immigrants and minorities—commands our attention. But another demographic change—the relentless aging of the U.S. population—will be far more consequential for national policy. Long-term care expenditures accounted for nearly one-third of Medicaid’s total outlays of $389 billion in 2010. As the population ages, the tension within Medicaid between caring for the elderly and the health needs of poor and near-poor families will escalate.
The problem is already acute. According to a recent report from the National Governors Association, Medicaid already constitutes the single largest share of state budgets—24 percent, a figure that rises relentlessly year by year. State spending on the program rose by 20 percent in the most recent reporting year and by even more—23 percent—in the previous year. The report estimated that by the end of fiscal year 2013, total Medicaid enrollment for low-income Americans and the dependent elderly will have risen by 12.5 percent in just three years. Because state revenues are growing much more slowly than Medicaid outlays, other priorities are getting squeezed. In many states, for example, public higher education—key not only to future prosperity and competitiveness but also to opportunity and mobility—is reaching a breaking point.
In short, there’s a looming crisis in long-term care because our current model for funding it is crumbling under the weight of multiple demands and inexorable demographic shifts. But we’re doing almost nothing to respond. It’s time to shift to a new long-term care model that combines personal responsibility and social insurance, the government and the market, in ways that would benefit not only current and future beneficiaries but the rest of society as a whole.
Roots of the Impending Crisis
Let’s start with the fundamentals. As a share of the total population, Americans over age 65 are projected to increase from 13 percent in 2010 to more than 20 percent in 2050. The share of Americans over 85 (the cohort most likely to need long-term care) will increase far more steeply, from 1.8 percent in 2010 to 4.3 percent in 2050. Depending on assumptions about medical advances and lifestyle changes, these projections imply that the number of disabled elderly needing long-term care three or four decades from now will be two to three times today’s total, which stands between 11 and 12 million.
To be sure, not all long-term care involves the elderly in nursing homes. In fact, more than eight in ten recipients live in their communities, and of them, nearly half are mentally impaired or individuals with disabilities under the age of 65. What unites all recipients of long-term care is their need for assistance carrying out some or all of the basic tasks of living, such as getting out of beds and chairs, maintaining personal hygiene, and dressing themselves. With suitable part-time assistance, many of these individuals can remain in their homes. But some cannot. The number of nursing home residents now exceeds 1.5 million, and as the number of Americans aged 85 and over keeps rising, so too will the need for institutional care.
While many people believe that caring for the elderly full time at home would be cheaper than in nursing homes, the facts suggest otherwise. As noted above, the annual cost of care in a nursing facility averages around $78,000 a year—more for a private room. Home care is around twice that: Wages for home health-care aides average $20 per hour, which works out to $480 per day, or roughly $175,000 per year for round-the-clock care. (Remaining at home is less expensive for those who need only a few hours of help every day for a limited range of activities.) Given these stark economic realities, it is easy to understand why well over half of all long-term care—with an economic value calculated at $375 billion in 2007, and more today—comes in the form of unpaid assistance that spouses and other relatives provide.
Despite the fact that the public sector finances more than 70 percent of all long-term care costs in the United States, U.S. public outlays for this purpose are on a par with the rest of the Western world, as are total outlays. Although technological change—a principal driver of cost increases—is not nearly as important in long-term care as in the health-care sector as a whole, expenditures have been growing faster in the former over the past half century and are projected to continue to do so over the next four decades. The most plausible explanation is that the aging of the population increases the demand for long-term care even faster than for acute care. If so, the kinds of measures that are projected to slow health-care cost increases are unlikely to prove effective for long-term care. Aging Americans face the prospect of high and rising long-term care costs as far as the eye can see.
If most of us could anticipate that we were going to face extended periods of near-total dependence, then only the wealthy could bear the costs, because only they would have the money to set aside. The rest of us would have to throw ourselves on the mercy of our children, private charity, or high-income taxpayers. But that is not the case. Even with today’s medical advances, which enable a higher percentage of the population to live into frail old age, most elderly do not experience extended stays in nursing homes or extended periods of dependence on professional home health services. Still, according to the much-cited estimate of health policy scholars Peter Kemper, Harriet L. Komisar, and Lisa Alecxih, 35 percent of all Americans who turned 65 in 2005 will end up needing institutional care at some point. Economists Jeffrey Brown and Amy Finkelstein estimate that between 10 and 20 percent of those who use a nursing home will live there for five years or more. Anthony Webb and Natalia Zhivan of the Center for Retirement Research at Boston College estimate that couples turning 65 face a 5 percent risk of incurring long-term care costs exceeding $260,000. There is, then, what Brown and Finkelstein call a “considerable right tail” in the distribution of nursing-home expenditures.
This structural fact suggests that long-term care is a classic insurable event—that is, an area of life in which each of us faces a small chance of a catastrophe with which we could not cope on our own. We do not buy insurance for the expenses—such as painting the walls and replacing the roof—that every home-owner incurs. We buy it, rather, to guard against disasters such as fires. Because relatively few of us will experience such disasters, we can use insurance to spread the risk among large groups of homeowners, making the protection we want far more affordable than it would otherwise be.
The Response Thus Far
There are many reasons why our long-term care policy should shift dramatically toward forward-funded insurance based on individual contributions. As more people use such insurance, the pressures on public finances, especially at the state level, would decline. Such a shift would alleviate harsh trade-offs states confront between health care and education, and would allow states to avoid choosing between Medicaid funds for health care for poor children and nursing-home operators. While it’s difficult to estimate the effect on overall health-care costs, rebalancing public and personal financing of long-term care would certainly lower the baseline costs of health-care entitlement programs—a must if we are to address our largest long-term fiscal challenge. Not least, moving toward such insurance would reflect a morally as well as fiscally sustainable balance between individual and social responsibility.
For various reasons, the private market for long-term care insurance has been slow to develop. One issue is information: Many people mistakenly believe that Medicare or their Medi-gap supplemental insurance policies will cover long stays in nursing homes. Time horizon is another impediment: The evidence suggests that consumers exhibit weak demand for other insurance products that must be purchased years or even decades before expected use. (If individuals were reasonably sure that they would incur at most limited health-care expenses prior to retirement, how much demand for health insurance would there be?) By the time most middle-income adults begin to consider purchasing long-term care insurance, they have reached the age at which annual premiums are very high, pricing many of them out of the market. (The average individual purchaser is 59 years old!) And most individuals tend to underestimate the odds that they will need long-term care services, in part because contemplating the prospect of frailty and dependence is so distasteful.
The health economist Richard Frank cites a number of additional factors reducing demand for long-term care insurance. Policies tend to be complex and call for a daunting number of design choices. They require potential purchasers to make sophisticated financial calculations and to assess multiple risks over multiple decades. And the time horizon of these policies raises doubts about the competence and reliability of the insurance companies who issue them. Many companies have been forced to raise premiums abruptly in response to overly optimistic estimates, while others have withdrawn from the market altogether.
There are also factors that tend to depress the supply of long-term care insurance, including the well-known problems of adverse selection and moral hazard. Individuals have information about their personal circumstances that they can use to time their purchase of insurance. Providers respond with forms and tests designed to screen out people whose health profiles suggest elevated risks. Because individuals who are aging and becoming frailer have increasing incentives to claim benefits, providers institute demanding procedures to determine actual eligibility for benefits. And finally, insurers have no choice but to limit their exposure to changes in demography and medicine that can affect entire populations over time, but these limits eliminate a range of policies that some individuals would otherwise wish to buy.
Medicaid also works, often in surprising ways, to decrease demand for private long-term care insurance. As a means-tested program, Medicaid imposes strict income and asset requirements for participation. One might imagine, then, that the program’s appeal would be restricted to low- and moderate-income elderly without substantial assets to preserve. The reality is more complicated, however, because individuals with assets can and do engage in Medicaid planning—those with assets above the threshold simply “spend down,” in the parlance, to qualify for coverage. According to Brown and Finkelstein, favored strategies include giving gifts to children and grandchildren, establishing trusts, and purchasing items that are excluded from the asset test. While Medicaid rules do allow states to examine such disbursements during a specified “look back” period, enforcement appears to be spotty. In practice, then, Medicaid does allow many middle-class Americans to sequester their assets, increasing its attractiveness relative to private insurance.
Some models suggest that this crowding-out phenomenon influences even upper-income households holding very substantial assets. While some analysts are skeptical of these findings, the fact remains that barely one quarter of individuals in the top income quintile own long-term care insurance policies. Given the “long tail” risk of extended nursing-home stays that would exhaust assets worth many hundreds of thousands of dollars, this unprotected exposure to highly negative outcomes is hard to explain. After all, wealthy individuals are disproportionately likely to have access to the information and advice they need to make rational plans. Something else is going on—myopia, or denial, or procrastination, or the mistaken belief that the safety-net programs will allow them to shelter large estates while receiving public benefits.
There’s another possible explanation. Brown and Finkelstein suggest that Medicaid makes private insurance appear less attractive in two ways. Owning private insurance makes it harder to qualify for the program. And because Medicaid is by law a secondary insurer, private insurance is required to pay benefits first, even if the public program otherwise would have covered some or all of those costs. This overlap, which reduces the net value of private insurance coverage, further discourages its purchase.
So, given all this, just reform Medicaid, right? It turns out, however, that stiffening Medicaid eligibility rules wouldn’t do much. Brown and Finkelstein estimate that if every state were to adopt the eligibility rules used by the strictest state, demand for long-term care insurance would increase by only 2.7 percentage points. Shifting Medicaid from secondary to primary payer of benefits would make more of a difference, but it would substantially increase the program’s total costs—a political non-starter, given that soaring Medicaid expenses are already the biggest threat to the sustainability of state budgets. If we want to confront effectively the looming long-term care crisis, we’ll have to do more than reform Medicaid.
Toward a New Model
As we work to move beyond Medicaid, there are some European approaches worth considering. For example, Germany instituted a mandatory, universal long-term care system in 1995, and Howard Gleckman of the Urban Institute has analyzed its principal features. The funding mechanism is a payroll tax, initially set at 1.7 percent—divided equally between workers and employers—and increased to 1.95 percent in 2008. Benefits are indexed to inflation and are reviewed for adequacy every three years. The public system covers about 70 million people, while another 9 million (mainly higher-income individuals) choose to purchase private insurance. According to law, however, all workers are required either to participate in the public social-insurance program or to obtain private coverage; no one is allowed to opt out. About two million Germans (2.4 percent of the population) who have passed a needs assessment receive benefits; two-thirds of them opt for home-based rather than institutional care.
According to Gleckman, the German program “has succeeded in substantially reducing the number of long-term care patients on means-tested public assistance.” This suggests that a similar program in the United States (a fully funded, actuarially sound Part E of Medicare, for example) could reduce the ranks of Medicaid beneficiaries. The German program has remained financially viable, albeit with a close balance between revenues and expenditures. And the program provides a considerable measure of choice that enables families to tailor benefits to their individual circumstances. They may receive in-kind benefits, where agencies under contract to the insurance program provide care directly; cash, which, though significantly lower than the monetary value of the in-kind benefits, they can use for a wide range of purposes; or a combination of the two. Public satisfaction with the program seems pretty high.
Still, the benefits provided under the German program are spare by Medicaid standards, and the cost of even this modest initiative is projected to rise in coming decades—to 3.2 percent of payroll in 2040, according to one estimate. The inexorable rise in the frail elderly as a share of total population will make it impossible for Germany to hold the line on long-term care spending as a share of GDP. There is no reason to believe that the future will be any different in the United States.
It’s also worth examining what we’re already doing here at home, modest as it is. In the first place, according to Gleckman, the federal government and at least 34 states provide tax credits or deductions for the purchase of long-term care insurance. But evaluations find only a modest impact on the purchase of policies, and models suggest that tax credits as large as 25 percent of total premiums would increase demand by only 11 percent.
Another U.S. initiative is the Partnership for Long-Term Care program. Sponsored by the Robert Wood Johnson Foundation in the early 1990s, the program allowed an individual who purchased a long-term care policy to be able to qualify for Medicaid while retaining assets equal to the policy’s value. In this system, private insurance reduced Medicaid outlays, while the Medicaid backstop cut the cost of private premiums. Only four states—California, Connecticut, Indiana, and New York—implemented Partnership programs before Congress, fearing that high-income individuals would use it to abuse the Medicaid system, acted in 1993 to prevent new states from participating. Twelve years later, the Deficit Reduction Act (DRA) of 2005 allowed other states to create Partnership programs. Thirty-nine states now offer policies under this aegis.
Here again, the results have been modest, verging on disappointing. Policies sold in the four original Partnership states totaled only 259,000. While the DRA brought another 35 states into the system, their policies have collectively added only 230,000. Evidently, the incentives the Partnership offers have been unable to catalyze a broad response.
Another step was taken in 2000, when Congress passed the Long-Term Care Security Act. It required the federal government to offer long-term care insurance to its employees and their families. The Office of Personnel Management (OPM) conducted a competitive bidding process and contracted with large carriers selling long-term care insurance in the private market to offer an array of insurance products. The contract ensured not only a wide range of choices but also the reliability of information about each option and enough standardization to permit ready comparison among options. As of 2011, the program had more than 268,000 enrollees, making it the largest private long-term care program in the country.
The cost of policies has contributed to the program’s success. A 2006 Government Accountability Office study found that annual premiums averaged 46 percent lower for single people and 19 percent lower for married couples who were both the same age than for comparable policies in the private market. In 2009, an unanticipated increase of between 5 and 25 percent affected about 150,000 participants, sparking widespread outrage and charges that the change reflected misrepresentations during the initial offering. Still, costs remain reasonable, at least for those who purchase plans well in advance of probable need. As of mid-2012, a 40-year-old federal worker would pay only $107 per month for a policy with expansive benefits. By contrast, a worker who waited until age 60 would pay more than twice as much—$231 each month, while a worker who put it off until the threshold of retirement would pay nearly $300.
Unlike most of the health insurance the federal government offers its employees, long-term care insurance receives no subsidies. The government does negotiate, however, with carriers who want to offer policies and makes information about costs and benefits available in an understandable, user-friendly format. And while most long-term care insurance is purchased on the individual market, the federal program enjoys the cost savings of a group plan.
That said, the program’s results should be kept in perspective. More than four million civilian and military personnel are eligible to participate, along with spouses and retirees. Even with the program’s convenience and cost advantages, the participation rate is barely 5 percent. If the objective of long-term care reform is to move away from traditional public programs toward a new insurance model, the OPM program represents a very modest step toward that goal.
The most recent effort to improve long-term care coverage was an instructive fiasco. The CLASS (Community Living Assistance Services and Supports) Act, appended to the 2010 Affordable Care Act (ACA), was intended to provide a benefit averaging $50 per day, mainly to support home-based long-term care. The legislation’s language was unusually sparse and general, delegating most key design decisions to the secretary of Health and Human Services. Republican Senator Judd Gregg succeeded in adding an amendment requiring the Administration to certify that the plan it developed would be actuarially sound and self-sustaining over 75 years. But the underlying legislation made that impossible, by allowing everyone—even those with serious health problems—to sign up, by permitting policyholders to collect benefits for life after only five years of contributions, and by limiting premiums the poor would pay to $5 per month. In the end, the Obama Administration had to admit that it could not square the circle: The Act’s mandatory features made it impossible to design a program that would achieve long-term sustainability. So the Administration threw in the towel and suspended implementation of the program.
The CLASS Act’s failure illuminates some basic truths about long-term care insurance. A voluntary program that must accept all applicants is bound to trigger adverse selection that will require higher premiums, public subsidies, or both to keep the program afloat. If neither is permitted, the program quickly enters a death spiral. Something has to give: Either the program must abandon guaranteed issue and select participants based on their health status, or it must get everyone to participate, generating a predictable pool over which it can spread the cost of benefits (much like the ACA). And if one of the purposes of the program is to limit or eliminate public subsidies out of general revenues—as should be the case, in order to avoid the costs associated with the ACA—then it must be allowed to adjust premiums so as to achieve long-term sustainability.
What’s Next for the United States?
If we think that the aims of the CLASS Act were worthy but the means fatally flawed, what should we do instead? My proposal, in brief, is to be bigger, bolder, and more fiscally realistic. There are two basic models for fundamental reform—one that mimics the structure of the ACA, the other that proceeds along the lines of the German program.
The first option would work as follows: At age 40, every adult would be required to purchase a long-term care insurance policy or to pay a penalty equal to 2 percent of wage and salary income for each year without coverage. The policy would need to have certain features: a term of five years, a benefit of at least $150 per day, an automatic annual inflation adjustment of 5 percent, a 90-day deductible, and benefits that could be received in cash or in kind and used for both home-based and institutional care. The government would provide subsidies for purchasing the plans. Individuals with household income between 150 and 300 percent of poverty would receive income-related premium subsidies; those below 150 percent would be enrolled for free.
The federal government would create a competitive bidding process along the lines of, but broader than, the current system for federal employees (and the exchanges under the ACA), with the aim of creating a large menu of carefully vetted, readily comparable choices. After the five-year benefit period expires, Medicaid would assume full financial responsibility for any remaining costs. Individuals in this category would not be required to spend down their assets to be eligible.
The five-year requirement is intended to track the distinction between normal expectations—the maximum amount of time that 80 to 90 percent of Americans will spend in nursing home care—and unusually long stays that represent the equivalent of financial catastrophe. The average nursing home stay is 2.4 years—higher for women, lower for men. As we’ve seen, relatively few stays last longer than five years. The choice of age 40 to mandate the purchase of insurance rests on a judgment as to when it is reasonable to expect adults to begin providing for events that may occur in later life.
Insurers in the current federal program offer 40-year-olds the policy on which my proposal is based for an annual premium of $1,285. With mandatory participation and more competition among insurers, premiums under this system should be substantially lower. Regulations would permit premium increases in only a narrow range of circumstances that companies would be required to document and subject to strict review.
Under this plan, Medicare expenditures would be reduced somewhat since private insurance would finance a portion of the 100-day rehabilitation period that Medicare now covers. The impact on Medicaid would be far larger in fiscal terms, and much more significant structurally. Medicaid’s revised role in long-term care would be to: 1) create, through a competitive process, the options among which individuals could choose and provide information to assist individuals in making that choice; 2) use regulations to ensure appropriate standards and safeguards; 3) provide premium subsidies for low- and moderate-income individuals; and 4) serve as insurer of last resort after the expiration of the five-year private benefit period.
The design of this proposal deliberately mimics that of the ACA and therefore should raise no constitutional issues under the congressional taxing power. Like the ACA, it involves the private sector, which would increase choice and competition but also administrative complexity.
That’s one idea. The alternative to the ACA model is an approach that adapts Germany’s program to American circumstances. Under this approach, workers over age 50 would remain in the current system. All workers 50 and under would be required either to contribute 2 percent of their wages into a long-term care fund or to purchase private long-term care insurance with at least the coverage features specified below. For all workers making less than twice the median wage, the federal government would make a sliding-scale contribution to ensure a minimum annual contribution of $1,000 per worker.
The government would establish eligibility criteria for companies wishing to participate in the new program’s market exchanges. All companies meeting the criteria would qualify for inclusion, and the terms and conditions of their policies would be arrayed in standard formats that make it easy to understand and compare them. The government would ask every worker to choose a program and a policy offering no less than five years of coverage at $150 per day with a 5 percent automatic annual inflation adjustment, a 90-day deductible, and fully flexible benefits as in the first option. Drawing from the long-term care fund, the government would subsidize that choice up to the value of the lowest-cost policy with the basic features specified above. Individuals who select more expensive policies would pay the difference into the fund, while individuals who fail to make a selection would be defaulted into the lowest-cost basic package. Workers who directly purchase at least a basic package for themselves on the individual market would be exempt from the mandatory payroll contribution.
Every five years after the new long-term care program went into effect, an independent long-term care actuarial commission would evaluate the program’s financial soundness and recommend whatever changes it deems necessary to ensure its long-term sustainability. Any proposal involving changes in benefits or the payroll tax rate would be submitted to Congress for an up-or-down vote within 90 days. If Congress failed to act, a package evenly balanced between revenue increases and benefit cuts would be triggered and would remain in effect until such time as Congress accepted the commission’s recommendations or enacted its own package making financially equivalent changes to the program.
A Moral and Responsible Goal
Our current circumstances of fiscal constraint, intense controversy over federal health programs, and pervasive skepticism about government’s effectiveness may seem like an odd time to suggest extending our patchwork system of social insurance. But reforms along the lines I’ve suggested represent the only way of achieving a morally essential goal—providing for the long-term care of the physically dependent elderly—while also meeting the equally essential practical goals of preventing unsustainable claims on general public revenues and ensuring that the states retain the fiscal capacity they need to invest adequately in the future. Americans will have to accept a broadened principle of personal responsibility as part of any viable twenty-first century system of social insurance. If the need for long-term care is an insurable event—and I’ve argued that it is—then we must all do our individual part, as our means permit, to help insure against it.
If we do what we should, my son’s generation will not have to face the burden so many members of my generation do—struggling to help parents avoid an old age of destitution and dependency. Few children want to consign their parents to institutional care if there is a reasonable alternative. Few want their parents to be forced to divest themselves of what they have managed to accumulate through decades of hard work, a process that strips the elderly of dignity and pride. But that is what our current system does. We can do better.