In the wake of the HHS announcement of Healthcare.gov enrollment figures Wednesday, many people in the media are relying on analysis of the Massachusetts health care law being proffered by supporters of the Affordable Care Act as proof the numbers are not as bad as they appear. Two specific claims are making the rounds. As I’ll demonstrate, these claims are either wrong or, perhaps intentionally, misleading.

Claim #1 – Only 123 people enrolled in MA the first month

On October 23rd Jonathan Cohn of the New Republic published a piece intended to calm the “media and political hysteria” about enrollment at Healthcare.gov. Using numbers given to him by ACA supporter Jonathan Gruber, Cohn created a chart saying only 123 people signed up during the first month of open enrollment in Massachusetts.

On November 1, the White House blog repeated this claim without citing a source:

In Massachusetts, 123 premium paying consumers – or 0.3 percent of the
eventual premium paying enrollees – signed up in the first month they
could enroll…

In an appearance on Jake Tapper’s The Lead show Wednesday Gruber repeated the claim once again:

When we opened our system in Massachusetts, the first month that people could pay premiums and enroll 123 people enrolled.

First let’s discuss what appears to be a bit of deception on Gruber’s part. The rollout of Commonwealth Care (which Gruber helped create) officially began on January 1, 2007. However, documents available online make clear that the state enrollment push did not begin until January 23, 2007.

This page offers a list of documents from a January 22 Connector Board Meeting. If you click on the Marketing Recommendation Presentation (a word document) you’ll see that Massachusetts hadn’t even selected a PR firm on January 1st. At the end of the document you see that they selected a firm called Weber Shandwick for a contract starting on January 22, 2007. And sure enough the first press release about sign ups for the new program was issued the next day, Jan. 23rd.

Why were the number of sign-ups low in January 2007? In part because PR did not begin until the final week of the month. This is not at all comparable to the situation with Healthcare.gov which had hundreds of millions of dollars worth of advance PR of every type, not to mention frequent discussion in the news media.

One might think, to be charitable, that Gruber wasn’t aware of this. That would be wrong. In fact, if you click on the Meeting Minutes found here you’ll see that Gruber was present at the meeting when Weber Shandwick was selected for the contract. In fact, the Minutes report that “Mr. Gruber expressed his desire that the board and outreach committee stay involved in marketing decisions.” I assume Cohn was not aware of this or he would have mentioned it.

Apart from the marketing, the number itself is wrong. Cohn admits as much in his post. The 123 figure is only possible if you exclude various categories of people in the Massachusetts system who would be counted under the federal enrollment figures for Healthcare.gov. Here’s how Cohn first explained it in his piece:

If you want to get a real sense of enrollment patterns among people choosing to shop and buy plans, it’s better to exclude the people
getting free care. (In the Massachusetts plan, that would mean people who ended up enrolling in what were called “Type I” and, with some
exceptions, “Type IIA” plans.)

Gruber and Cohn have a point about excluding type 1 category in Massachusetts. These people were being
auto-enrolled in a free program similar to (but separate from) MA’s Medicaid program. They had no premiums, no co-pays and no options. Since we aren’t looking at Medicaid enrollments in this comparison it’s probably best to leave type 1 folks out.

However, the people Massachusetts categorized as type 2a are different. These are people making between 100%-150% of the poverty line. And when their enrollment period began in January 2007 the plan they were offered had a small monthly premium (about $18). It was only a few months later (pdf, page 4) that the state revised its subsidies so that type 2a folks had no premiums.

But even without the premium, type 2a has a counterpart in the Obamacare system. People making between 100% and about 130% of the poverty line and who are living in a state which did not expand Medicaid (many of which are covered by Healthcare.gov) can be offered a premium which covers 100% of a bronze plan. You can see this for yourself using this Kaiser subsidy calculator.

So, just like the type 2a folks in Massachusetts, at least some of the 27,000 folks who enrolled on Healthcare.gov last month are getting a free ride. Therefore, for several reasons, an apples-to-apples comparison should include type 2a.

According to publicly available data, the first month of open enrollment in Massachusetts
actually signed up 226 people.
One month later the cumulative total was
4,186 people. As a percentage of the total who signed up in the first
year that would be 0.6% in month one and 11.5% by month two. As I’ll argue in a moment, the latter figure is the better one for making federal comparisons.

But there’s one more group Gruber and Cohn have left out. Again, Cohn
mentions this in passing but those who cite him, including the White House, usually do not. In
Massachusetts, anyone who made more than 300% of the federal poverty
line (and who did not have an employer plan) had the option to buy a
plan on Commonwealth Choice without a subsidy. The open enrollment
period for this program began in May 2007 for plans starting in July.

This Excel spreadsheet
says that as of July 2007 Commonwealth Choice had 2,799 enrollees. This is analogous to anyone who buys a plan on Healthcare.gov without a subsidy. These folks should be included to make a fair comparison.

Finally, there’s the problem of comparing two different enrollment periods. The Massachusetts open enrollment period lasted nearly a year. By
contrast, the current Obamacare enrollment period is just 6 months. Since the question being asked is who signed up when, it’s
misleading to directly compare the two without making some adjustments for the differing time scales.

One option to make the two comparable would be to look at the percentage of the total enrollment
period elapsed. Every month of Healthcare.gov open
enrollment is 1/6 or about 16% of the total time. That would be
(roughly) equal to two months of Commonwealth Care open enrollment. Put
another way, that first month’s figure in MA (226 people) would be
better compared to the first two weeks of enrollment in Obamacare, not the first month.

Claim #1 Conclusion

PR for the Commonwealth Care started in late January. As a result the first month’s enrollment figures were very low. But after two months there were 4,186 enrollments in Commonwealth Care, which this Massachusetts Excel spreadsheet labels “premium paying.” Those are the folks we should be looking at.

The enrollment period for Commonwealth Choice (the non-subsidized program) should also be included in an apples-to-apples comparison. So the best comparison figure for early enrollment in MA would be 4,186 people enrolled in Commonwealth Care plus 2,799 enrolled in Commonwealth Choice. That adds up to 6,985. And that’s vastly different from the 123 figure being pushed by Gruber and the White House.

Claim #2 – Over 20 percent enrolled in the last month

This claim is being made by the same people, including Jonathan Gruber and the White House

This claim is true but also somewhat misleading. For one thing, the spike in enrollment mostly happened among the type 2a people which Gruber and Cohn sought to exclude form consideration. Enrollments in the other plans were fairly steady over the entire period.

If a similar pattern were to exist with Healthcare.gov we would expect to about half of the anticipated sign ups by the end of December. Could we see 3.5 million by January 1st?

Claim #2 Conclusion

Gruber and the White House are highlighting the final month of enrollment to suggest that things might be sluggish before then. Cohn’s graph also gives the impression of a hockey stick by leaving out all the intervening months in which enrollment steadily climbed. (To his credit, Cohn added an update which makes clear that’s not what happened.)

The graph above shows what really happened. Enrollment climbed quickly early on and continued to climb steadily with an extra large jump in low income, heavily subsidized plans in the final month. So far Healthcare.gov does not seem to be following the Massachusetts pattern, not unless the final enrollment is going to be far lower than predicted.