Fake Chart?

Many people have been commenting on a chart that shows annual per capital health care expenditures in the U.S. by age group. The chart seems to say that this figure is about $3500 until people reach their mid-50′s and then rises exponentially to about $30,000 in their 70′s and $45,000 in their 80′s.

Tyler Cowen is among those pointing to the striking chart, which is creating a frenzy in the health-care-wonkosphere.

At least two folks, Austin Frakt and Kevin Drum, are skeptical about these numbers. I am beyond skeptical. I call baloney sandwich.

Finding the most reliable data for 2010 takes two seconds. Just go to the trusty Medical Expenditure Panel Survey. The mean expenditure per person for people with expenses is $3866 for people under 65 and $10,274 for people 65 and over.

It takes another five minutes to generate your own table using MEPS. I wanted to break down the over 65 group into finer categories. So here are the means for each age group:

45-54: $4816
55-64: $6823
65-74: $9265
75-84: $10,175
85+: $11,233

Those are the facts, as best I can determine.

[Update: a number of bloggers have now backed away from the chart (see Tyler's comment, posted below), on the grounds that it does not include private health care spending. But that makes it sound as though the problem is that the chart understates spending on the young, when in fact the problem is that it overstates spending on the old. The most charitable interpretation of how the chart emerged is that somewhere along the way somebody started with TOTAL spending on health care by people in a middle-age bracket (say, 45-54), multiplied this by the ratio of GOVERNMENT spending on people in a higher age bracket (say, 65-74) to GOVERNMENT spending on the middle-age bracket, and arrived at an alleged TOTAL spending figure for people in the higher age bracket. But really, trying to figure out how bogus numbers made it into a chart is a mug's game. Regardless of how they got there, they are bogus.]

9 thoughts on “Fake Chart?

  1. Tyler: I do not see any “repudiation” on your blog. It looks like the post was just deleted. That’s reasonable although doesn’t help the people who already read the post.

    Arnold: Thanks for fact checking that, especially since the obvious conclusion (that medicare spends way too much and healthcare spending ages 0-65 is very reasonable) sits easily with “our side”.

  2. I went to the website you highlighted and I came up with higher means than you came up with. I am not sure what I did differently.

    The median is relevant to this discussion since the gap between the median and the mean for all age groups is significant. An examination of the median for all age groups demonstrates how low health care expense is for the majority of the US population.

    In the 18-44 group, half of the population spends less than $875 per year. In the 45-64 age group, half of the population spends less than $2124 per year. For policy discussions, this is proof positive that most of the population is better off with very high deductible plans.

  3. Also in the 18-44 age group, 25% of people spend zero. This shouldn’t be unexpected. Healthy adults usually don’t need medical attention. And in response to Jardinero’s comment, I would certainly have a high-deductible plan, if the family HMO that I do have wasn’t so heavily subsidized by my employer (but we all know that story, too).

  4. Well, it’s not a bogus chart. It’s an incomplete (several countries dropped) and inaccurate (public spending only) if you’re looking at overall medical care spending, but that wasn’t its original context.

    It’s “fake” in the sense that my hammer is a “fake” screwdriver. Using it to open the bottom of my laptop to add memory may be a Really Bad Idea, but it would work. (To borrow a Zevon line, “I can saw a woman in two/but you wouldn’t want to look in the box when I’m through.”)