In 2018, expenditures by Medicare and Medicaid combined grew 4.9%. To put these rates in perspective, U.S. gross domestic product (GDP) grew only 2.9% that same year. Medicare and Medicaid expenditures have grown faster than the U.S economy for a long time.
Needless to say, this situation is not sustainable indefinitely. Eventually the growth of Medicare and Medicaid will have to be constrained in some way. Most likely by imposing limits on coverage as single-payer countries do.
Proposals to indiscriminately expand Medicare without imposing limits on coverage, i.e. Medicare For All (MFA), test the bounds of credulity. The Urban Institute and Commonwealth Fund estimate the cost of MFA would be $3.2 trillion per year from 2020 to 2029. That is an eye-popping amount, but one that is almost certainly too low. According to that same study, even without MFA, government expenditures on healthcare in 2020 are expected to be $1.6 trillion. We’re talking about doubling something that is already growing unsustainably.
Proponents of MFA counter with the defense that higher taxes will be compensated by the elimination of premiums and out-of-pocket costs. In the aggregate this argument might seem valid, but it relies on an “apples-to-oranges” belief that a billion dollars of private expenditures on healthcare equal a billion dollars of tax revenues. They don’t.
This argument also ignores an even more onerous consequence of MFA than its total dollar cost, i.e., a large redistribution of healthcare. When robbing Peter to pay Paul, Peter might object.
My idea for an alternative to MFA, Medicare Prime (MP), would achieve many of the same goals without MFA’s ruinous costs. To demonstrate this, I created a model using data from the 2017 Medical Expenditure Panel Survey (MEPS). Make no mistake, this model would not pass muster with the Congressional Budget Office (CBO). It relies on several assumptions about healthcare demand elasticity and consumer preferences that are too numerous to describe in detail in a blog. Also, the MEPS is known to undermeasure total healthcare costs. But the model does give a good back-of-the-envelope estimate of MP’s likely cost compared to our current system.
A Model of Medicare Prime
In a previous post, I offered a simple formula for determining a household’s MP deductible. It relied on three parameters: a, b, and c. Parameter a determines the deductible’s size as a percentage of middle-income above the federal poverty level (FPL) while parameters b and c determine the lower and upper bounds of middle-incomes. A fourth parameter, d, determines the fraction of the Medicare rate that would be paid to out-of-network providers.
When b = 1.38 and c = 4, the middle-income range is the same as that for households eligible for ACA premium subsidies. So, for comparison sake, the only two parameter values we need to explore is for a and d.
From a budgetary perspective, the interaction of these two parameters determine how generous we can afford to be with middle-income households versus out-of-network providers. The smaller the deductible, i.e. parameter a, the less we can afford to pay out-of-network providers, i.e. parameter d.
From an incentive perspective, the more we encourage households to use less expensive in-network providers, i.e. set a low value for parameter a, the more we need to encourage providers to choose to be in-network, i.e. set a low value for parameter d. Therefore, the values of feasible values for a and d are directly related or positively correlated.
According to the MEPS in 2017, Medicaid expenditures for people under age 65 were $195 billion. This is likely an underestimate, but since this is a comparative study, that shouldn’t matter much. ACA premium subsidies were approximately $45 billion that year. Total healthcare expenditures by the government for the non-Medicare population, therefore, were $240 billion in 2017.
Figure 20.1 illustrates the government healthcare expenditures under MP for the range of values from zero to 1 for both parameters a and d. For a given value of a, the amount of government expenditures would increase as we increase the value of parameter d. For a given value of parameter d, the amount of government expenditures would increase as we decrease the value of parameter a.
Figure 20.1 Annual Government Expenditures and Values for Parameters a and d
The horizontal dotted line indicates the current level of government expenditures for healthcare for the under age 65 population. The points where it intersects with the different values of parameter a indicate the budget neutral combinations of a and d. These are illustrated in Figure 20.2.
Figure 20.2 Budget Neutral Combinations of Parameters a and d
A policy that greatly incentivized low and middle-income households to use only in-network providers would be represented by the bottom point of the budget-neutral curve, where a = .25 and d = .03. At the other extreme, a policy that maximized low and middle-income households’ choices of providers would be represented by the top point of the curve, a = 1.0 and d = .85.
The objective of this exercise was to demonstrate that MP would not necessarily require a large increase in government expenditures and taxation the way MFA does. The optimal point along this curve would depend on factors such as how much low and middle-income households value choice of provider over out-of-pocket costs and premiums and the number of providers willing to be in-network. I made some reasonable guesses, but that is all they are, guesses. Additional research is needed to discover better informed policy options.