The national debate on health reform has me thinking about a particular angle of the question: paying for health care in retirement. Let’s put aside for the moment long-term care costs (i.e., nursing homes) and focus on regular medical care—doctors’ bills, hospital fees, drugs, and so forth.

Health care costs in retirement, like health care costs generally, continue to explode. The numbers are daunting. In a June 2009 report, The Employee Benefit Research Institute in Washington estimated that a 65-year-old couple retiring in 2009 would need $268,000 to $414,000 in savings to pay for out-of-pocket health expenses during retirement. This is in addition to the costs that Medicare would pay on behalf of the couple.

Wow! These are amazing numbers. They are larger than most people’s nest eggs.

But here’s the question I’ve been pondering. Is this actually the right way to think about health care costs—as one huge, unmanageable lump sum? I don’t think so, for two reasons.

First, all health care projections are plagued by a problem of compounding—or, as Herbert Stein, the noted economist, once said, “If something can’t go on forever, it won’t.” All long-term estimates of retiree health expenses assume rapid growth in health care costs for the foreseeable future, well in excess of growth in incomes and the economy. When health costs compound at a high rate, they explode and swamp everything. It’s not just retirees who are impoverished—so are governments and private employers and active workers. Entire countries even.

John Maynard Keynes famously observed: “In the long run we’re all dead.” Today’s variant would be: “In the long run, we’re all bankrupt—from health care.”

The question is, can health care costs compound at high rates forever? The short answer is, no. We will not have an economy devoted exclusively to health care and nothing else. But because we cannot see a world in which health care costs grow at slower rates—we cannot today imagine the mechanism by which costs will eventually slow—we take past history, namely rapidly growing costs, and simply project it forward.

Here’s the second reason. When it comes to health costs in retirement, there seems to be an assumption that health is a unique category of expenditure, somehow separate from or independent of our income in retirement. Actually, it’s another category of expense that our retirement incomes can and should be used for. As we age, we spend more on health care. Retirement isn’t just about spending our money on housing, utilities, transportation, or leisure; it’s also about spending money on medicine and doctors and health. And we derive positive psychological (as well as medical, hopefully) benefit from that spending—just as we derive positive benefits from our home or from the dinner we eat or from taking a trip.

Think of it this way. Consider a 65-year-old couple, earning the median household income of $30,000 for retirees, who live for 25 years. That’s $750,000 in total income. Suppose they spend 20% of their income on housing and transportation, and 20% of their income on health. That’s $150,000 each on these two items.

Why is one a problem (health) and the other not (housing and transportation)?

Ah, you say, the problem isn’t spending $150,000 on health—it’s the fact that the $150,000 is growing too quickly, and will bankrupt retirees. I refer you to reason one above. We come full circle.

In the end, I can’t tell you what mechanism will slow costs. Nor can I tell you whether it’s right to spend 10%, 25%, or 50% of your lifetime retirement income on health.

But it’s for these two reasons that when I read of these large estimates, I pause, think about it, and tend not to panic.

Note: The link to will open a new browser window. Vanguard accepts no responsibility for content on third-party websites.