I decided to do a Health Alert – expanding on my comment – which is now posted at my blog.

As far as Uwe’s query, yes. Everyone’s IQ tends to fall when thinking about health care. The question is: How long does it take to come to your senses? Some people apparently take longer than others.

Cheers.

Age rating (varying premiums based on age), like medical underwriting (varying premiums based on health status), is one way for insurers to segment risk – separating the healthy from the sick, making insurance cheaper for some, but inaccessible and unaffordable for many others. In some markets today, premiums can vary as much as 11:1 based on a customer’s age. That means the oldest customer could pay as much as 11 times more than the youngest customer simply because of his or her age! In my view, this moves beyond actuarially fair into the immoral category.

Let us not forget one of the central goals of reform is to channel self-interest to serve the social interest. In the case of insurance markets, this means forcing insurers to compete based on price, value, and customer satisfaction, rather than avoiding the sick. Allowing insurers to charge older Americans vastly higher premiums simply because they are older is not part of this vision.

Ultimately, age rating should be eliminated. In the transition period, however, we could lessen its impact significantly by reducing – without eliminating entirely – the amount by which premiums can vary based on age. When combined with sliding scale subsidies based on income, this transition period will ensure health insurance is affordable for the young and healthy, while providing immediate relief to those Americans who find insurance unaffordable or inaccessible because of their age.

Collapse

How much more should insurers be allowed to charge older people than younger people? So far, health reform legislation in Congress has included language to allow insurers to charge older people twice as much as their younger counterparts. While that would lower premiums for people at, say, age 60, whose medical costs might be higher, it would raise premiums for the younger, generally healthier, generation. But legislators have given some consideration to allowing them to charge older people five times as much as younger people, which more closely resembles the differential today. Which makes more sense, and how important is this distinction? Is age rating even something the government should regulate?

Blue Cross and Blue Shield companies strongly support bipartisan healthcare reforms this year that improve quality, rein in costs, and expand coverage to everyone. One part of comprehensive reform must address insurance market reforms—that is, how to make the market work for everyone, regardless of health status. In addition to providing government subsidies to support a personal responsibility requirement for individuals to obtain and maintain health coverage, insurance market reforms also must allow for rating flexibility based on certain factors, such as age. This ensures that young people in particular will not flee the market, thus perpetuating the current problems of the uninsured.

While we are concerned that a 2:1 limitation is too narrow—especially given that so few states have that limitation today—we think that a 5:1 limitation would be good public policy. Today, people aged 18-30 are more than twice as likely as people aged 55-64 to be uninsured. Being able to offer appropriate age discounts will be an important part of getting young people to obtain coverage after reform. Young people typically use less health care and are less likely to appreciate the value and protection of health insurance than older people. If premiums are a lot higher than the medical expenses young people expect to have, they will forego coverage. This happened in Maine during the 1990s when 90% of individual market policyholders under age 30 dropped coverage after the state tightened its age rules. Even today, insurers in states with narrow age limits generally find that the average age of an insured individual is much higher than in states with broader limits.

Of course, the other side of the issue is what happens to older adults. On the individual level, we strongly support income-based subsidies so that assistance is available to everyone, including older adults, who need help paying for coverage. In the aggregate, older adults are going to benefit from younger adults buying coverage – even at discounted rates – because this will ensure a balanced, sustainable insurance pool in which more cross-subsidies are available to pay for their care than currently exists today.

Collapse

It’s worth looking at Utah’s recent reforms as an appropriate approach. Utah reforms limit age rating to a maximum difference by age of four to one.

Uwe Reinhardt argues that matching subsidies to age-rated premiums, "would quickly become administratively cumbersome, with a fair amount of cheating." I don’t think that is right. A 'percent of premium up to a maximum amount' approach would be a simple and compatible subsidy mechanism to accompany age-rated premiums. Furthermore, the amount of the subsidy cap could even be adjusted by the recipient's age (or the age of the head of household in the case of family coverage), setting a higher subsidy cap for older individuals. At Heritage we've drafted and modeled this type of tax credit and it can be done in a way that is easy to administer, with little scope for cheating.

I believe John Rother is wrong on two points.

First, employers can, and in some cases, do, vary premiums based on age, though that is admittedly the exception and not the rule. Indeed to the extent that most employers do not age adjust premiums (particularly when the employee is charged part of the premium for family coverage), younger workers and families are less likely to take up coverage.

Second, and more important, there is a large relationship between age and income. In fact, even most lower-paid occupations pay more for older, experienced workers. Maybe John could find a secretary at AARP with 20+ years experience and ask her if at retirement she would rather have her pension based on 80% of her average earnings for her last three working years, or 120% of her average earnings for her first three working years?

Collapse

Question to John Goodman:

Did your IQ also fall 15% when you wrote your post on health care in this round? Or did it stay up, because you were not thinking about health care when you wrote your post.

Just curious.

Best,

Uwe

Marilyn,

Why be so narrow? Let’s ask a much broader question: Suppose you could pay any premium you like for health insurance. What premium would you select?

a. Would you choose to pay the fair market price (reflecting the expected cost of adding you to an insurance pool), guaranteeing that insurers would vigorously compete for your business?

b. Or, would you choose a much lower premium, guaranteeing that no insurer would want you?

c. Or, would you choose a really, really low premium, guaranteeing that insurers really, really would not want you?

In thinking about this problem, it may be helpful to think about other products and services ― since IQ tends to fall about 15 points whenever people start thinking about health care.

If you could pay any price you like for dinner at a restaurant, what price would you select?

There is no objective answer to this question. Different people may have different views on it, and each warrants respect.

All health systems are structured on an underlying social ethic on how the resources of the health system are to be financed and shared. Most nations articulate that ethic quite explicitly and allow it to be a powerful constraint on their health policy.

Unfortunately, we have never done likewise, which can explain why our health system embraces unbridled compassion and amazing callousness side by side.

Among the ethical questions nations must answer in structuring their health system are:

1. To what extent, if any, and should wealthier citizens be made to subsidize the purchase of health insurance and health care by less wealthy citizens, other things (health status, age, etc.) being equal?

2. To what extent, if any, and should healthier citizens be made to subsidize the purchase of health insurance and health care by sicker citizens, other things (wealth, age, etc.) being equal?

Age enters this consideration only because it tends to be highly correlated to health status and the need for health care. Age is a proxy for them. It is well known that, relative to a 35-40 year old, people over age 50 and up experience, on average, much higher health spending – up to 5 times for people in their 70s.

To find answer to these ethical questions requires the services of a rabbi, priest or minister. Economists can merely raise them and suggest metrics on the extent to which sharing takes place.

One such metric, for example, might be the fraction of its disposable income (after essentials such as food, minimally acceptable shelter and clothing) that an individual or family must spend on its health are through insurance premiums or out of pocket.

In principle, at the level of theory, one could allow the health insurance sector to price its policies on what is known as an “actuarially fair” basis, which means that the premium charged the individual would reflect as accurately as it can the individual’s expected health status for the future period being insured.

If the body politic then desired, say, to equalize the fraction of disposable income individuals or families must devote to health care, that ethical goal would be achieved not through the health insurance mechanism, but through an explicit a tax and subsidy mechanism run by government.

Most nations have chosen not to go that route, for a number of reasons.

First, such a tax and subsidy mechanism would quickly become administratively cumbersome, with a fair amount of cheating, as in connection with the income tax.

Second, economists will argue that this mechanism will bring with it its own inefficiencies.

Third, and very importantly, it might be politically more difficult to enact such a massive, highly visible transfer of wealth. The cross subsidies paid and granted though the insurance mechanism are much less visible, and what people don’t know does not bother them.

Alas, building hidden cross subsidies into an insurance mechanism becomes ever more difficult, the easier it is for individuals to switch competing health insurers on short notice, especially if the purchase of health insurance is left as a voluntary choice on the part of the purchaser. As Rothschild and Stiglitz have shown in a seminal paper, in such a setting health insurance markets will segregate the insured population as much as possible by risk class and be highly unstable over time.

This is the reason why we now envisage imposing guaranteed issue and community rated health insurance premiums on the health insurance industry. But those strictures will work as intended only if the purchase of health insurance is made mandatory upon the individual. Furthermore, there must then be a risk-adjustment mechanism compensating insurers who have ended up with an above average risk mix.

Now, letting age be a variable on which the premiums charged the individual can be based reintroduces health status as a variable in this pricing. And once again, how far to limit letting age drive premiums is basically an ethical decision to be made by politicians. If there are no limitations, then community rating takes place only within a given age group. Perhaps this is what Americans want.

In the health insurance system of, say, Germany, younger people know that they pay higher than actuarially fair premiums, in both the Statutory and the private commercial market. But they also know that in older age, their premiums will be lower than actuarially fair. So far, this has been politically acceptable there.

Collapse

Insurance is the spreading of risk, not only between the healthy and the sick, but also among those in high risk occupations, those who live in high cost areas, and those who face higher risk because of their age. The broader risk is spread, the more affordable insurance can be for everyone.

Employers don’t vary premiums by age, and neither does Medicare, the federal employee benefit program, or insurance programs in other countries. Keeping premiums level for all is a key to keeping them affordable for everyone over the lifecourse. There is little relationship between income and age – a waitress or secretary in her 50’s makes about the same as one in her 20’s. So varying premiums by age would be regressive in that lower income older persons would end up facing much higher premiums than higher income younger workers.

Basing premiums on income, as the House and HELP Committee bills do, is inherently fairer. In a system with income based subsidies, however, permitting large premium variations by age just drives up the on- budget cost of subsidies, as many older persons would require higher subsidies to offset their higher costs in order to keep premiums affordable. The way to minimize costs to the federal budget is to minimize premium variations by age.

The ideal system is community rating with adjustments based on income. Permitting premiums to vary by no more than 2 to 1 based on age is a compromise that could soften the transition from very large age differentials today in the largely unregulated individual market. The bottom line for all ages, however, is affordable coverage based on a percent of income.

Collapse