Important Points Made by the Speakers
• Relatively few people have purchased long-term care insurance, and many of the companies that have offered coverage are leaving the market. (O’Leary)
• The long-term care insurance industry will need to innovate in order to balance the risk to carriers with the needs of consumers. (O’Leary)
• Public subsidies for long-term care insurance currently benefiting the upper end of the income spectrum could be redirected to the middle of the spectrum to make long-term care insurance less costly. (Frank)
• Automatic enrollment in 401(k) plans by employers can substantially increase participation rates, and automatic escalation of contributions would substantially increase savings. (VanDerhei)
The private sector plays a role in the financing of long-term services and supports both through the provision of long-term care insurance and through participation in employee retirement plans, and three speakers at the workshop discussed this role. John O’Leary, president of O’Leary Marketing Associates, discussed private long-term care insurance and the difficulties it has faced in recent years. Richard Frank, the Margaret T. Morris Professor of Health Economics in the Department of
Health Care Policy at Harvard Medical School, described policy initiatives that could increase the use of long-term care insurance while easing the pressures on Medicaid. Jack VanDerhei, research director of the Employee Benefit Research Institute, explored how the private sector can help ensure that employees have enough money for retirement.
John O’Leary O’Leary Marketing Associates
Long-term care insurance is an $11 billion business serving more than 7 million consumers, but the penetration of the overall market remains low. Group long-term care insurance covers about 2.3 million people, and individual policies cover 4.8 million. Together this represents only 5.8 percent of the U.S. population above the age of 45.
Furthermore, many of the companies that have offered coverage are leaving both the group and the individual markets. According to O’Leary, only one insurance carrier is in the group market, and it has moved to longer-form underwriting rather than guarantee issues, which will reduce the appeal of its policies. Most group policies are now in what are called closed blocks, in which the product being administered is no longer being offered. Meanwhile, individual policies have undergone rate hikes as high as 90 percent.
The companies that have remained in the business are seeking to aggressively manage risks through premium hikes, more restrictive underwriting, limiting benefits, and reducing discounts and agent commissions. They are not trying to grow because, as O’Leary said, they have determined that making money in the long-term care insurance business is very difficult.
Several factors are behind this retrenchment, O’Leary said. First, historically low interest rates make it difficult for companies to profit from invested premiums, which was not the case when many of the policies were written. Also, the initial actuarial assumptions proved to be too optimistic. Fewer policies were dropped than expected, mortality rates were lower, and morbidity rates were higher. In addition, the initial plan designs were costly for companies. The result has been a “perfect storm” of problems for long-term care insurance, O’Leary said, with high risks and low rewards for companies offering plans.
As O’Leary pointed out, more than 23 million families are in the top quintile of income in the United States, with mean annual incomes above $114,000 per year. This constitutes a sizeable market for long-term care insurance, which historically has been purchased predominantly by higher-income consumers. People in the lowest quintile are closer to qualifying
for Medicaid. The people in the middle are the ones who fall through the cracks.
A Tipping Point
Long-term care insurance is at a tipping point, O’Leary said. Current actions reinforce a “niche” positioning for the industry. The most important question, then, is whether the industry will undertake the innovation that will be necessary to meet a broader range of consumer needs. The affordability of policies is the number one barrier to more sales, but it is not the only barrier.
What the industry needs, O’Leary said, is a way to balance the risk to carriers with the needs of consumers. It needs to broaden the appeal of long-term care insurance beyond “healthy/wealthy” consumers, which will require new thinking and better marketing and distribution. But the problems facing the industry have not yet been solved.
The health care system is also in the midst of dramatic changes, and health care is inextricably related to long-term care, O’Leary observed. How people take care of their health earlier in life helps determine whether they will need long-term care later in life, although, of course, the correspondence is not precise. In the future people are going to manage their own care more than they do today, O’Leary predicted. Higher deductibles and copays will be part of the reason, but innovative wellness programs also are becoming ubiquitous, as is the recognition that earlier knowledge and intervention can ameliorate later problems.
O’Leary outlined three approaches that he suggested the insurance industry could take to limit risk:
• First, companies can offer a life or annuity product with a long-term care rider.
• Second, by creating policies with flexible inflation designs, companies can share the risk with consumers while making insurance more affordable.
• A third approach would be to offer more modest protections while reducing premiums and streamlining the underwriting process.
Each approach holds promise but has potential barriers, such as typically not being available to groups. Consumer feedback, marketing analysis, and consumer research and segmentation may make it possible to hone future modifications.
According to a survey of actuaries conducted by a policy planning group to which O’Leary belongs, the approach most favored by actuaries includes both private insurance and a government-sponsored safety net or a social insurance program. Respondents to the survey disagreed on whether a mandate will be necessary or what the required coverage levels would be.
Other innovative ideas include a long-term care health services account, mutual long-term care, or shorter-term coverage. O’Leary was uncertain whether any of these would work, but people are exploring options and are open to change, he said. For example, he described a managed wellness program that would combine a wellness lifestyle program with financial protection so that, for example, premiums or deductibles would be tied to information on health status in a health record in the same way that automobile insurance rates are tied to driving records.
The key questions regarding long-term care insurance are who is going to offer it and what is going to be offered. O’Leary concluded by saying that funding is limited and that people are unlikely to get as much as they might want.
Richard Frank Harvard Medical School
A major problem with long-term care insurance, Frank said, is that people do not know much about it—what it costs, what it covers, or whether they will need it. They also have difficulty understanding the future implications of today’s choices, given their uncertainties about their risks and future. Furthermore, the products are complex and can be unpredictable, as when premiums go up unexpectedly. Frank offered four policy measures that might offer remedies for these problems.
The first, he said, would be to simplify and standardize products. Choices could be limited to 5 to 10 products with no limit on the number of sellers. Such an approach could reduce confusion and aid comparison shopping while also saving money on selling costs, Frank said. Electronic markets and decision aides could help people navigate the marketplace. State regulations could be altered to allow high-deductible plans to be part of the choice set, which could reduce premiums by 35 percent. This would appeal to the market, Frank said, because people could protect themselves against catastrophic outcomes at a premium discount.
The second approach, Frank suggested, would be to expand the employer role. Employers have many advantages in offering long-term care insurance, including lower selling costs, reduced concerns over adverse
selection, and the ability to filter products for their employees, he said. One possibility would be to offer simple low-cost products through employers. Another would be to mandate availability, so that the benefit would be offered to employees if they want it. Though politically difficult, this approach has been effective in some health care markets.
The third policy approach described by Frank would be to develop reinsurance pools. Many of the greatest risks and uncertainties for long-term care insurers are common to all participants and are hard to spread. As a result, insurers limit coverage or exit the market. Publicly organized and privately funded (or publicly and private funded) reinsurance pools have been used successfully in flood and earthquake insurance and are being proposed for some financial markets. By reducing the risk to insurers and stabilizing the industry, such a program could improve confidence by assuring consumers that the people who have qualified to buy into the program have met publicly agreed upon standards, Frank said.
The fourth option offered by Frank would be to provide targeted public subsidies for long-term care insurance.
Frank described four kinds of federal tax incentives that provide such subsidies: itemized deductions of medical expenses (which include long-term care insurance premiums), self-employed deductions, employer-sponsored long-term care insurance, and long-term care insurance purchased through health savings accounts. In addition, states have their own credits and deductions, besides carrying through the federal deductions.
Frank estimated very roughly that the medical deduction amounts to a subsidy of $1.4 billion, the self-employed deduction $1 billion, and state deductions $100 million. Furthermore, more than 40 percent of the medical deduction goes to the top quartile of the income distribution of older adults. Some of the other subsidies, such as the one covering self-employed individuals, go to younger people with higher incomes.
Frank asked whether it would be possible to target these more than $2.5 billion in subsidies to expand protection, encourage more private spending on long-term care insurance, and reduce the burden on Medicaid. He noted that the 40 percent of households in the United States in the middle of the income distribution had an average of less than $80,000 in total assets at age 65. Therefore, the potential for people to spend down their assets and quality for Medicaid is large, regardless of how many are actually doing so. And even if they do not spend down to qualify for Medicaid, their consumption levels and well-being are vulnerable to health and disability risks late in life. They have lived average American lives only to be left on the verge of destitution if anything goes wrong, he said. If the
subsidies currently directed toward the upper end of the income spectrum could be redirected to the middle of the spectrum to make long-term care insurance less costly, Frank argued that people’s lives could be improved, as could the prospects for Medicaid.
Redirecting these subsidies has both a technical component and a political component, Frank observed. Both raise hard issues, but the possibility should be explored.
During the discussion session, one workshop participant observed that simplifying and standardizing a product can quickly generate litigation directed at collusion. He also pointed out that requiring that long-term care insurance be offered is likely to increase adverse selection risk, which will make companies even less likely to offer policies, and that reinsurance pools face the problem of large deductibles. Frank countered that simplification would have to be sanctioned by government to avoid litigation, so attracting a larger pool of people is most likely to reduce, rather than increase, selection, and that many insurance companies are in favor of reinsurance, especially if the government sets standards so that companies can have more confidence. Frank also mentioned the possibility of indexing premiums to reduce the likelihood of unexpected increases and of using 401(k) savings or non-refundable tax credits to pay for premiums without penalties.
Jack VanDerhei Employee Benefit Research Institute
According to a research program conducted at the Employee Benefit Research Institute, 60 percent of the households in the lowest income quartile will run short of money in retirement. They will still have Social Security, but they will have depleted their defined contribution and individual retirement account (IRA) assets and any housing equity. Moreover, 41 percent of households in the lowest income quartile will deplete those assets within 10 years of retirement. However, the vast majority of this saving shortfall disappears if largely unpredictable and substantial health expenses not covered by Medicare, usually from nursing homes, are not included.
One thing that employers can do to help stave off this future is to change the 401(k) plans they provide, VanDerhei said. Since 2006 more and more employers have begun offering automatic enrollment plans, which automatically enroll employees on their first day of work. Employees can opt out if they want, but few do. As a result, participation rates among lower income employees have more than doubled since 2006.
The automatic escalation of contributions would make such plans even better, VanDerhei said. Very few people put into retirement accounts
the amounts they should be putting in—10 to 15 percent—and employers are very reluctant to start their employees at this level. With automatic escalation, a 3 percent contribution might go up to 4 percent after 1 year, 5 percent after 2 years, and so on, until the maximum was reached or the employee chose to opt out of further increases. According to VanDerhei, such a plan substantially increases the likelihood of replacing a significant portion of pre-retirement salary from Social Security and 401(k) accumulations upon retirement, depending on whether and when employees decide to opt out.
Education of employees can help convince them not to opt out of automatic escalation too soon. Also, when they change jobs they should be encouraged to keep their contributions at what they were in their previous job rather than starting anew at a low percentage. Finally, the escalation rate can be set at a 2 percent increase per year rather than a 1 percent increase, which would increase the chance that people will retire with an adequate replacement rate.
Employees also can purchase long-term care insurance. VanDerhei described a study showing that the people who have the most to gain from long-term care insurance are those in the second and third income quartiles (VanDerhei, 2005). The lowest income quartile are the closest to being eligible for Medicaid, and the highest income quartile generally have enough assets to self insure.
Asset allocations should be age appropriate, VanDerhei said. Young employees should not have zero equity exposure, employees close to retirement should not have equity allocations that are too high, and employees should not have an excess concentration in company stock (VanDerhei, 2009). In addition, working to age 70 instead of age 65 can increase the probability of success in retirement by 24 percent (VanDerhei, 2012), and annuitization at age 65 can reduce the replacement rate needed for 90 percent probability of success by as much as 24 percent (VanDerhei, 2006).
In the discussion session, one workshop participant asserted that the estimates of life expectancy generated by the Social Security Administration and the Census Bureau are too conservative. He noted that other estimates indicate that people will live as many as 5 years longer and said that this difference causes a dramatic difference in model outcomes. VanDerhei agreed that life expectancy has a “huge influence” on whether a household will run short of money in retirement but that most people run out of money in the first 10 years of retirement, which is reflected in the results of the models he described.