First the right started talking up this post by AEI resident scholar Andrew Biggs showing that consumer spending on veterinary services is rising as fast as health care spending in the U.S. (see Cowen, McArdle, Kling, Mankiw, Manzi). Then the left responded (see Klein, Yglesias, Drum, and a sympathetic DeLong). There have been two main criticisms of the chart (reproduced below). First, there's funny business going on with the scales used (as Matt puts it, "Fun with the Y Axis"). Second, the figures aren't comparable because they should be per capita. Let's take a look at both of these criticisms.
First on scaling. You could write an entire book on how to lie with charts. But this one's fine -- both Y axes start at 0, and both end close to where the series hits its maximum. It's true the levels of spending are much different, but Biggs's point is about the change, not the levels. If you eyeball the numbers in the chart and compute % changes over the period, you'll see that they're similar.
Let's do that. Biggs shows national health expenditures rising from about $800 billion in 1984 to about $2.1 trillion in 2006 -- an increase of about 160 percent. The increase in veterinary service spending was from about $4.5 billion to about $11.1 billion, or 150 percent. Given these are rough guesses and that all of these numbers have considerable uncertainty, that's pretty similar. Point for the conservatives!
However....I tried to find Biggs's figures on the web but couldn't. I'm pretty sure they're wrong. [Update: Biggs emailed me to say that he had adjusted for inflation, which I should have guessed--so his figures are not "wrong". -srw] The official source of national health expenditure data is the Centers for Medicare and Medicaid Services. National health expenditures, according to CMS's figures rose from roughly $400 billion in 1984 to $2.1 trillion in 2006 -- an increase of 426 percent -- not 160 percent! Point for the liberals!
Except...it looks like Biggs's veterinary expenditure figures might be off too. Consumer Expenditure Survey figures on veterinary services expenditures are not, as far as I can tell, easily gettable. Instead, I went to the trusty Statistical Abstract, put out annually by the Census Bureau. And lo and behold, those crazy bastards have data from the American Veterinary Medical Association going back to 1983 on household expenditures on veterinary services for dogs and cats. When I add up expenditures on both, the increase is from $3.5 billion to $21 billion--an increase of about 500 percent! Point for the conservatives (if not for Biggs)!
A number of commenters noted that these figures really ought to be per person/animal. Otherwise, it's possible that the increase in the number of pets outpaced the increase in the number of people and that's what's really driving the figures for animals. From one perspective, if household pets were growing more numerous at a faster rate than their owners were, that would simply reflect that with growing affluence, people are choosing to spend more on the luxury good that is a furry four-legged companion occasionally needing expensive veterinary care, just like they are choosing to spend more on health care. But if the argument is about inefficiency in care, then per-capita expenditures are probably the most appropriate.
Anyway, when you look at the increase in spending per capita, health care spending per person rises by 350 percent, vet spending per dog rises by 335 percent, and vet spending per cat rises by 340 percent.. So on this one, I think the conservatives have the better argument, despite the flaws in the original evidence.
(Sources: For the vet and pet data, see http://www2.census.gov/prod2/statcomp/documents/1990-03.pdf, Table 400, and http://www.census.gov/compendia/statab/tables/09s1201.xls. For the health expenditure data, see http://www.cms.hhs.gov/NationalHealthExpendData/downloads/nhe2007.zip, and for the population figures, http://www.census.gov/popest/archives/1990s/popclockest.txt and http://www.census.gov/popest/states/NST-ann-est.html.)
RSS Feed