• What does a GDP+1% growth cap mean?

    It’s cliché to say the devil (or God) is in the details. Nevertheless, it’s true for so many things and is particularly apt for Medicare premium support proposals. In recent months, I’ve been corresponding with Henry Aaron about various premium support specifics that are not pinned down by proposals, including that by Rep. Paul Ryan and Sen. Ron Wyden. Below is just one example, pertaining to what a GDP+1% growth cap might mean. If you think you know what it means, I bet you’re wrong.

    The Ryan/Wyden plan says that “program growth” will be capped by nominal GDP plus one percentage point (GDP+1). (See their proposal (pdf) or my summary.) This sounds unambiguous, but, in fact, by itself is meaningless. Some reasonable interpretations are problematic. The following illustrates these points. It’s a bit raw and mathy. If needed, I will return to the points made and make them again in a less terse form. If you’re a true health policy wonk, you’ll want to read this now anyway.

    Let’s assume Medicare enrollment is growing 3% per year and total GDP is growing 2% per year. Let’s also assume that the per person cost of traditional Medicare is rising 3% per year and that GDP per capita is growing at 1% per year. Then, Medicare spending is rising 6% a year (3% for enrollment growth plus 3% in per person cost, ignoring interactions, which are small).

    Let’s now consider what could be meant by setting the target for Medicare spending growth to GDP plus one percentage point (or GDP+1, as it is normally written). What does that mean?

    Possibility 1: The GDP+1 target applies to total Medicare spending. In this case, under the assumptions above, traditional Medicare spending growth (6%) exceeds the GDP+1 (2%+1%=3%) target by 3 percentage points. To bring Medicare spending down to the target without affecting those grandfathered into traditional Medicare, the growth in the  voucher offered for a premium support program would have to be quite low. It would have to grow slow enough and apply to enough people to offset 3 of the 6 percentage points of growth in traditional Medicare. Under some conditions that would be impossible. For example, if traditional Medicare was not an option for new enrollees (i.e., they all had to enter the voucher program), the assumptions above make meeting the GDP+1 target impossible. Enrollment is growing 3% and we’d need to offer vouchers to all those new enrollees that grow in such a way to offset traditional Medicare growth by 3%. This can’t even get off the ground.

    So, this interpretation is absurd. The GDP+1 target cannot apply to total Medicare spending.

    Possibility 2: The GDP+1 target is interpreted in a per capita sense and applies to Medicare spending per person, which, as assumed, is rising 3% per year. GDP per capita is going up 1%, so GDP+1 is a 2% growth cap. So, overall Medicare spending per person has to be held to a growth rate of 2%. But the cost for all of those grandfathered into traditional Medicare is going up at a 3% rate. How does one get the overall program average of per person growth down to 2%? Either one has to hold the amount of the voucher below the cost of traditional Medicare or one has to scale back the value of traditional Medicare. The second approach destroys the meaning of ‘grandfathering.’ So, this hardly seems a viable interpretation. Consequently, it means that the voucher must erode relative to what is offered those in traditional Medicare. Is everyone on board with this idea? It seems at least a little politically problematic to me.

    Possibility 3: The cap applies to those in the voucher program only, either in total or per person. This approach severs the voucher program from traditional Medicare as far as cost growth bookkeeping is concerned. Depending on the numbers, it could lead to growing or shrinking vouchers. Nevertheless, this seems like the most feasible interpretation of a GDP+1 cap.

    The third possibility may be most feasible, but it still leaves the question of how much money it would save. I’ll address that another day. For now, note that none of the above hinges tightly to the specific numbers selected, so long as they’re in the realm of reasonable (as those above are). Make it GDP+1/2 or Medicare growth of 2% and the qualitative conclusions are the same, for instance.

    Given the ambiguity inherent in the Ryan/Wyden plan, among others, It strikes me as premature to say there is a growing bipartisan consensus on premium support.


    Comments closed
    • What is the point of the RYan/Wyden plan? Do they believe it will somehow put pressure on providers to reduce costs?

    • Your analysis is spot on. Although neither plan is crystal clear, my understanding is that both the Ryan-Wyden proposal and 2013 Ryan budget would apply their respective caps to the voucher amount–your Possibility 3, on a per person basis. Wyden and Ryan could readily resolve this and other ambiguities in their proposal by providing legislative language, but they have said that they do not intend to do so.

    • What happens to Medicare expenditures in the case of a recession? Suppose GDP goes down 3%. Medicare expenditures would then need to decrease by 2%? In a recession/depression that would be a disaster.

    • I am afraid you missed two important points in your email. First is that the Office of Actuary at CMS assumes GDP+1 for per capita spending growth, before you consider the age-sex factor or enrollment growth. Second is that Trustees have assumed GDP growth much higher than the 1% you used; they assumed 1.7% for the next 75 years in TR 2012.