I’d like to call attention to Paul Krugman’s misleading use of a statistic in a recent Op-Ed.
In reaching for some sign of ACA (Obamacare) success Krugman states “so far, so good: in October, 22.5 percent of California enrollees were between the ages of 18 and 34, slightly above that group’s share of the population.”
Presumably this is a positive sign since we want young, healthy people to sign up for health insurance.
The trouble is that Dr. Krugman’s reason is blinded by his partisanship combined with desperation for something quantitative to plug into his narrative. Dr. Krugman is very intelligent and I’m sure he’d see through the meaninglessness of that statistic (statoid?) if it were put forth in support of something he opposed.
Let me elaborate. There are three problems.
First, it is a very preliminary number. It’s from the first month. A slight preponderance during a brief period contains nothing of statistical significance.
Secondly, he sidesteps adverse selection by implying that “young” = “young and healthy”. I’ve seen this conflation many times during these discussions. It’s as if we have a new word, “younghealthy”. It is reasonable to expect that within each age cohort there will be some adverse selection. In other words, among the previously uninsured a young person who is sick is more likely to sign up for healthcare under the ACA than a young, healthy person. This is both “common sense” and economic sense. So a claimed slight preponderance of the “young” does not mean we’re getting lots of healthy people into the system.
Third, and perhaps most importantly, there is a false comparison between the cohort’s ACA enrollment proportion and the cohort’s proportion in the entire population. The appropriate baseline is the cohort’s proportion of the ACA-eligible population. This excludes people over the age of 64, because they have Medicare!(*)
Aside: Because of this, the proportion of every ACA cohort could be higher than its total population share. Indeed the data source he cites, Covered California, states “Oldest age brackets, 45-64, are overrepresented compared to the population”. So add another dose of deception to Dr. Krugman’s narrative – omitting a key fact that contradicts his narrative.
From the Census data I find the following for the 18-34 age group in California (2010)(**):
% Total Population: 24.8%
% <65 Population: 28.0%
Thus, the 22.5% rate is substantially below the relevant <65 proportion of 28.0%. This meaningful comparison is quite different from the misleading claim that 22.5% is “slightly above that group’s share of the population”.
Also, 22.5% is actually below even the total population proportion of 24.8%! Dr. Krugman receives a pass on this one though. The Covered California data source shows 7% + 14% = 21% for the 18-34 group. This source cites the “2010 US Census data”, but for the life of me I can’t get to 21%. (***)
Aside: The California exchange site is still clunky and slow though improving. More importantly, a consumer cannot find out which providers are in-network. This is a critical piece of information needed to make informed choices between the competing plans. The fact that Krugman and his ilk ignore this or swat it aside reveals their true colors. They really don’t seem concerned about people making informed choices. The plight of the currently insured doesn’t phase them. The lack of concern, the inability to even fake some empathy for the plight of millions, is a real “tell”.
None of this means the ACA is not “working” or that it will not “work”. It simply means that Dr. Krugman’s factoid-driven approach is unsound and misleading.
* Actually, there is a minor weakness in my protest. Not everyone aged 65 or older is eligible for Medicare. Those people may use the exchanges. (You must have worked 10 years paying into the system or have a spouse who had.) Still, only 0.5% of sign ups during this period were 65 or older – less than 1/20th of this groups population proportion. Still, I expect California has a higher than average proportion of seniors who paid little or no Medicare taxes.
** The 2012 California estimates were 25.0% and 28.4%, respectively, so that actually moves in the opposite direction of Dr. Krugman’s claim. The 2012 national estimates were 23.5% and 27.2%, respectively. So it’s not clear where Dr. Krugman found an estimate of the 18-34 age cohort being less than 22.5% of the population.
*** I downloaded DP-1-Geography-California: Profile of General Population and Housing Characteristics: 2010 from census.gov for the 2010 data. I had to do some calculations because there is a “15 to 19 years” cohort. There is also information for “18 years and older” as well as “20 to 24 years”, “25 to 29 years”, “30 to 34 years”. But the share of just those latter three (20-34) is 21.7%. It is possible that the Covered California folks miscalculated something by excluding instead of including an endpoint in age ranges or visa versa. Perhaps they’re using old data that were preliminary estimates? (They also show “AGE <18” as 28% while the official census data show 75.0% for “18 years and over”. Strange. (Incompetent?)