Are federal government expenditures on medical care transfers/consumption or government purchases?
About a week ago, Tyler Cowen linked (approvingly) to a blog post by by Garrett Jones which argues that federal government spending on health case (which is classified as a transfer payment from the federal government to households and then as personal consumption spending by households) should be counted, instead as a government purchase of goods and services. In 2001 (the year for which data are provided in the BEA document to which Jones linked), this amounted to $476 billion, about 5.5% of personal income. (The best data I can find on this issue are here, in Table 3.2.
I'm not sure how I feel about this, but there are two issues to be considered. The first has to do with how this would cause us to revise our understanding of personal income and consumption spending. If we decide to include Medicare and Medicaid spending as government purchases rather than as transfers and consumption spending, then we must also adjust personal income (those transfers are counted as a part of personal income) and consumption spending downward. Given the rapid growth in federal government health care expenditures (inflation-adjusted, Medicare spending has increased at an average annual rate of 16% since 1966, while other health care spending has increased at a 9% annnual rate), these adjustments will reduce both the level and the rate of growth of consumption spending and personal income.
For the population as a whole, these effects are not large, actually. In 2012, personal consumption expenditures (PCE) adjusted for reclassifying federal government health care spending, were about 93% of unadjusted PCE. Adjusted personal income (PI) would be about 94% of unadjusted PI, and adjusted disposable personal income (DPI) would be about 93% of unadjusted DPI. But this understates the actual effect of the reclassification. Medicare spending is for health care of those 65 and over, and Medicaid spending is for low income families (including low-income elderly). So the effect on how we measure the income and the consumption spending of the elderly and of low-income households would be significantly affected. In 2011, for example, households with a head of household age 65 or over had total income of about $1,277 billion (including Medicare benefits); those medicare benefits accounted for 37% of household income.
Let me be clear here. The situation of the elderly would not actually change. But our understanding of their income and consumption levels would look very different if we reclassified federal government health care spending.
The second issue is that this change would mean that we would be treating private health care insurance benefits differently from our treatment of government-provided health care benefits. Unless I am mis-reading things, private health cae insurance benefits are counted as private transfer payments (from the insurance company to the insured) and personal consumption spending by the insured; in the national income accounts, then, these benefits appear as additional to personal income as well as consumption spending. Jones argues for counting federal government health care spending, as noted above, as government purchases because of the degree of control that the federal government (allegedly) has over the providers of these services. But private insurance companies, it would seem, have similar influence/control over the providers of health care services paid for by private insurance. Consistency, then, would suggest treating them in a similar fashion.
But what I don't see is how treating these two types of health care spending differently from the way in which we treat them now would enhance our understanding either of how health care systems operate or of our national income accounts. But perhaps I am missing something...
So I am unpersuaded by Jones's argument that the reclassification "makes sense" as a way to understand what's happening in health care markets, or how it would improve our national income accounts.