Wednesday, February 15, 2006
A Major Pathology Of Insurance "Donut Holes"
by Tom Bozzo
The issue is that unless everyone's expenses put them past the donut hole (making the donut hole, in some ways (*), a de facto first-dollar deductible), the donut hole deductible has to be much larger than a conventional deductible — about twice as large, in fact, depending on the distribution of claims and some other factors — to have the same cost for the insurer.
So, while a donut hole policy tends to appear attractive because of relatively low first-dollar costs, anyone unlucky to be sick enough to hit the 'catastrophic' territory beyond the donut hole will end up paying roughly twice as much out-of-pocket as they would have with non-donut hole insurance that would have cost the same to provide. And, of course, seniors with big prescription drug expenses are not exactly the best positioned to bear the greater cost individually.
Now, Jeremy had noted back when that there is a legitimate policy interest in getting people to sign up for the program, so an appearance of benefits to all is not irrelevant. (Less clear is whether the donut hole is a better way of doing this than fiddling with coinsurance rates.) However, having individuals bear risk instead of the government makes next to no sense as public policy. Government is the institution best positioned to bear, and to diversify most fully, health outcome risks.
(*) But not all. For instance, the cost signal for the first dollar of expense would be different.
Here's another observation for the old debate with Jeremy over the pathology of the Medicare Part D donut hole. Kate Steadman of Healthy Policy (h/t Ezra Klein) caught a Health Affairs article by Harvard's Meredith Rosenthal on the economics of donut hole policies — alas, no publicly accessible version of the paper is available from Rosenthal's homepage — showing that donut hole plans shift risk to consumers and cost to the relatively sick, relative to plans with "first-dollar" deductibles that are equally costly to the insurer.
The issue is that unless everyone's expenses put them past the donut hole (making the donut hole, in some ways (*), a de facto first-dollar deductible), the donut hole deductible has to be much larger than a conventional deductible — about twice as large, in fact, depending on the distribution of claims and some other factors — to have the same cost for the insurer.
So, while a donut hole policy tends to appear attractive because of relatively low first-dollar costs, anyone unlucky to be sick enough to hit the 'catastrophic' territory beyond the donut hole will end up paying roughly twice as much out-of-pocket as they would have with non-donut hole insurance that would have cost the same to provide. And, of course, seniors with big prescription drug expenses are not exactly the best positioned to bear the greater cost individually.
Now, Jeremy had noted back when that there is a legitimate policy interest in getting people to sign up for the program, so an appearance of benefits to all is not irrelevant. (Less clear is whether the donut hole is a better way of doing this than fiddling with coinsurance rates.) However, having individuals bear risk instead of the government makes next to no sense as public policy. Government is the institution best positioned to bear, and to diversify most fully, health outcome risks.
(*) But not all. For instance, the cost signal for the first dollar of expense would be different.
