Years ago, when I first started writing about health care, I came across a press release that said three new cardiac centers had opened in a Midwestern city and that, as a result, the costs of heart care in that town were expected to rise. This seemed contrary to all I had ever learned about supply and demand. But it was a powerful lesson. Health care economics, it turns out, is an oxymoron. The normal rules don’t apply.
That explains in part why it is so hard to model the costs of the massive health bills Congress is debating. And why those estimates from CBO and JCT that will drive the legislation in the coming months will almost surely miss the mark. This is no criticism of the congressional estimators—or of those in the private sector who are trying to do similar work. They are doing as well as anyone can. But accurately measuring the ten year effects of huge changes in the medical system is simply not possible.
It should be no surprise. Health care is a $2.5 trillion industry—nearly 18 percent of the economy. It includes many providers whose relationships to one another are not well understood. At the same time, consumers of health care are often not the payers. This changes their behavior in a big way. On top of all that, patients are often driven by non-economic incentives. How does one model fear, for instance?
Medicare Part D drug insurance may be the best recent example of how health care budget estimates can go wrong. That debate, like this one, eventually came down to dollars. But as complex as it was, creating Part D was simple compared to broad health reform. Instead of remaking the entire insurance system, Part D changed the way seniors paid for just one product—pharmaceuticals. Thus, forecasters needed to project only three big things: How many people would buy the insurance, what premiums they’d be charged, and how much they’d spend on drugs.
The debate itself included a highly-charged argument over cost estimates. But even after the law passed in late 2003, government experts struggled to project its ten-year price tag. The official estimates came from the Medicare actuaries—a highly respected and professional operation that knows the program as well as anyone. They hedged their forecasts as much as possible, even showing low-, intermediate-, and high-cost projections. They warned that estimates of a brand-new program were highly uncertain. Despite all the caveats, their forecasts were far off course.
The March, 2004 Medicare trustees report predicted 2008 premiums would range from $12.7 billion to $19.7 billion. Actual 2008 premiums: $5 billion. The trustees figured 2008 spending would range from $77 billion to $131.4 billion, with a mid-range of $101.9 billion. In reality, the program paid just $49 billion in benefits last year—half the intermediate projection and far below even the low-cost estimate. For context, the $53 billion difference between projected and actual spending is equal to half of the total estimated annual cost of all of health reform.
Keep this story in mind when, a few months from now, President Obama and congressional leaders find themselves haggling over some last-minute idea that is intended to keep the cost of the bill below what has become the target of $1 trillion over 10 years. In truth, nobody will actually have a clue about what these changes will cost—or save. If we are lucky, they’ll get the direction right.
