Having built a model to tackle the old-age care market, Atlanta Fed experts turn to health insurance puzzles

June 22, 2017

a-hero with banner

A tangle of mathematical formulas crowds a whiteboard in Toni Braun's office at the Federal Reserve Bank of Atlanta. The thicket of numerals and symbols is the raw material of economic research. Here, Braun and colleague Karen Kopecky are dissecting the enormous complexities involved in U.S. markets for health insurance.

It can get arcane. The math on the board is part of an extension of the model Kopecky and Braun devised to help decode the market for long-term care (LTC) insurance, which mainly insures nursing home stays. They aim to use the extended model to analyze the individual health insurance market, which is for people who don't get insurance through their employer.

Drawn to the puzzles

Their work on the U.S. LTC insurance marketplace was motivated by some perplexing features. "There are just a million different puzzles in the market," Kopecky says. "That makes it interesting." Adds Braun, "You look at the real world, and you see something that looks puzzling. Then you look around and see if anyone has offered an answer to this."

About half of Americans over 50 will stay in a nursing home at some point, and a substantial fraction will experience a long-term stay. Long-term nursing home stays are about three years on average and are not covered by Medicare. Moreover, a nursing home stay is expensive, averaging more than $80,000 a year for a semi-private room. Given the size of this risk, one might expect that most Americans would have private insurance. Yet only about 10 percent of those over 65 have LTC insurance coverage.

A second puzzling feature of this market: even though the number of retirees is growing, as the first wave of U.S. baby boomers turned 70 last year, the size of the private market for LTC insurance is stagnant. In fact, many insurers have left the business altogether. The remaining insurers use strict medical underwriting standards that result in many applicants being rejected. For those who are offered insurance, premiums are expensive, typically about $250 to $420 a month.

Producing clear theories that can make sense of insurance market puzzles involves crunching mountains of numbers and overturning some classic theoretical results from earlier research. Ultimately, Braun and Kopecky hope to devise standard contracts, or insurance policies, that would prove sustainable for insurers and insure virtually everyone, with some government involvement. The federal government, of course, has long been a major player in supplying health insurance to low-income and older people.

After spending the better part of two years constructing a model that can account for the puzzles in the LTC market, at a quantitative and qualitative level, Braun and Kopecky are now looking into the individual health insurance marketplace.

They face many challenges. Health insurance markets are fraught not only with economic and financial challenges but also with political and social complications. Yet economists like Braun and Kopecky are drawn to the intellectual challenge the way mountaineers are drawn to dangerous peaks.

Starting to understand dysfunctional long-term care insurance market

Braun and Kopecky recently published their latest working paper from their ongoing research on LTC insurance. It's important work. The deep dysfunction in that market is troubling, as millions of Americans will eventually enter nursing homes or require other forms of expensive care late in life, Kopecky says.

"It's a difficult risk to insure well," Kopecky says. "As a result, people prefer to self-insure. However, from a policy perspective, that does not seem to be optimal."

A full half of nursing home care is paid for out of pocket. (Estimates of total annual spending on nursing home care range from about $200 billion to $320 billion, according to the Centers for Disease Control and Prevention's National Center for Health Statistics.) Most people—more than 60 percent—will require some type of care in old age, whether it's in a nursing home or at home. So from a policy standpoint, it is daunting to have such a large, expensive risk—the possibility of funding nursing home and other late-in-life care for millions—insured in something of a patchwork fashion.

Kopecky and Braun set out to untangle the reasons why it has proven so difficult to insure against late-in-life care expenses. After all, insuring against many other risks—car wrecks, house fires, faulty appliances, dental care, and so on—is relatively straightforward.

LTC insurance might seem an attractive market. Consider that the number of Americans 65 and older will more than double between 2010 and 2040, according to the U.S. Census Bureau. Yet the private-sector market is small and shrinking for two main reasons, Kopecky and Braun explain.

First, among those with lower earnings and wealth, Medicaid crowds out demand for private insurance. Unlike Medicare, the public insurance program for the elderly, Medicaid covers nursing home expenses for those with few if any assets. Since Medicaid is essentially free, lower-income seniors have little reason to purchase private LTC insurance.

For the middle class, Medicaid can kick in when they have exhausted their savings and other assets. In this sense, middle-class individuals are also partially covered by Medicaid. Since Medicaid is a secondary payer, any private insurance individuals buy will just reduce their Medicaid benefits. In other words, individuals cannot buy private insurance to top-up Medicaid benefits. As a result, even wealthier individuals who only are partially covered by Medicaid may find private LTC insurance unattractive.

The second major problem is that LTC insurers reject lots of applicants. "Supply-side frictions" lead the companies to turn away the frailer prospective policyholders, for example. Surprisingly, those tend not to be the most likely to ultimately enter a nursing home and thus make insurance claims, Kopecky says. That's because the sickest applicants tend to die before needing nursing home care.

Rejections in general grow out of a friction that economists call "private information": applicants know information about their health status that the insurer doesn't know. For example, an individual knows how likely their family is to care for them, but the insurer probably doesn't. Braun and Kopecky find a more severe private information problem among highly frail individuals. Some very frail individuals are very likely to soon enter a nursing home. But insurers cannot easily distinguish these individuals from other highly frail people, who won't enter a nursing home. Faced with the private information dilemma, insurance companies in many cases decide rejecting applicants is more prudent than taking a risk based on incomplete knowledge.

Predicting nursing home entry is hard

LTC insurers tend to have market power. That is, consumers have few choices for buying coverage and therefore little bargaining power. Braun and Kopecky find that rejections among middle-class and more affluent individuals can be attributed to private information once one accounts for market power and administrative costs.

Despite their market power, LTC insurers apparently have had trouble making profits. One reason is that pricing LTC insurance is a tricky actuarial job. On average, people buy LTC insurance in their early 60s and, if they use it at all, enter a nursing home about 20 years later. Thus, the insurer must set a price based on a forecast of what they may have to pay out two decades later, Kopecky explains.

Sometimes these forecasts go wrong. For instance, the unanticipated decline in interest rates since the 1990s has hurt insurers. LTC insurance companies make money mainly by investing premiums they collect. Because of regulations limiting risk, much of the LTC insurers' investments are in interest-rate-sensitive instruments. Insurers wrote policies before the financial crisis that were not priced to produce profits in such a low-rate environment, Kopecky says. What's more, she adds, even most LTC insurance products sold in the 1990s have proven to be underpriced, as insurers also underestimated growth in nursing home costs. At the same time, insurers appear to have overestimated the number of customers who would simply quit paying premiums and let policies lapse.

In fact, securities formed from bundles of long-term insurance policies sold in the 1990s are considered toxic, Kopecky says. Amid this jumble of issues, several companies have exited the LTC insurance market in recent years, while others have gone before state insurance regulators to request significant premium increases.

"I would say right now, if anything, the market is looking worse, not better," says Kopecky.

Still other forces are muddying the LTC insurance market. One is the trend of fewer people entering nursing homes. Nursing home care is far more costly than home-based care, either informal family care or paid service. With this in mind, insurers have started to offer coverage for LTC services provided at home. But the administrative costs of home care insurance are steep. For one, the companies must check on patients frequently to verify that they really need the services they are receiving.

The puzzles in LTC and other health insurance markets are abundant. Ultimately, Kopecky and Braun aim to better understand insurance markets broadly. In the current phase of their research, they seek to understand how the market for individual polices will affect insurance plans for the more than two-thirds of Americans who get health insurance through an employer. A major step is to model the Affordable Care Act, then presumably what, if anything, replaces it.

"The world's a very complicated place," Braun says, "even more complicated than our models."

photo of Charles Davidson
Charles Davidson

Staff writer for Economy Matters