Thursday, August 21, 2014

The best Obamacare data comes from a home office in Michigan

Florida shoppers wait in line to meet with Obamacare advisors on the last day of open enrollment. Joe Readle / Getty Images

For the past six months, the best Obamacare sign-up data hasn't come from a Washington think-tank or Health and Human Services number-crunchers.

The best, most granular accounting of Obamacare's sign-ups happens in a messy bedroom-turned-office in Michigan where Charles Gaba runs ACASignups.Net.

Gaba built the country's most trusted source of Affordable Care Act enrollment data

Gaba is a 43-year-old web developer who lives in Bloomfield Township, Michigan. He's an Obamacare supporter but one who finds the health insurance industry "ungodly boring." His eyes used to glaze over when his father, a doctor, would start talking HMOs and PPOs.

Charles_gaba01

Charles Gaba

But for some reason, he became interested in the Obamacare enrollment numbers and has spent six months tracking them obsessively. He usually does so at night, after his 8-year-old son has gone to sleep. And with no background in health policy whatsoever, Gaba built the country's most trusted source of Affordable Care Act enrollment data.

Gaba's enrollment counts are routinely more up-to-date than the federal government's, which is why they turn up on newspaper front pages and cable news shows. Paul Krugman has described him as "the invaluable Charles Gaba."

Gaba became well-known because he did something that, on the surface, seemed easy: He counted the Obamacare sign-ups. Lots of other people (myself included) tried to do this, and failed. Figuring out how many people have enrolled in Obamacare turns out to be a surprisingly difficult task.

"At a certain point, and I'm not sure when, I just kind of became the guy doing this," Gaba said in a recent interview. "I started to feel obligated to keep going."

'Sixteen different, moving sources of information.'

State_health_insurance_exchange_decisions
It was hard to anticipate how challenging it would be to keep track of Obamacare's sign-ups. Massachusetts' coverage expansion 2007 is usually the closest analog for predicting the health care law — but there you had one state tracking enrollment, not dozens.

Prior to open enrollment's start, Health and Human Services had never committed to a particular plan for releasing information. They eventually settled on monthly data releases, which left lots of space between each data dump.

But that was just for the 36 states where the federal government was running Healthcare.gov. There were another 15 marketplaces run by states and the District of Columbia who, as early as October, started releasing sporadic data.

Most people who tried to track enrollment — and weren't doing it for a living — gave up

"The challenge, on the exchange side, is you've got 14 states, the District of Columbia, and the feds all reporting different information at different times," says Caroline Pearson of Avalere Health, a research firm. "You've got 16 different moving sources of information. It's on different websites, states are tweeting it. It gets released in different formats that are all over the place."

Even the information that states spewed out was different. Some states, like Washington, included a breakdown of which shoppers had paid their first month's premium and who hadn't. Most states, and the federal government, didn't provide that granular data, instead releasing a lump sum of customers who had and hadn't paid for the plans they selected.

"We generally try to use the data from the carriers and states to come up with a reasonable assumption of who has paid and who hasn't," Pearson says.

Most people who tried to track enrollment — and weren't doing it for a living — gave up. It was too time-consuming; just when you thought you'd collected the most current data, another state would publish something even newer.

"You have 16 state exchanges that don't always keep to a schedule," says Aaron Strauss, who tried tracking enrollment in early October. He estimated he would spend about two hours a day — one before work, one after — trying to check in on more than a dozen state-run marketplaces.

At Avalere, Pearson has a team of 26 people working for her who try to keep tabs on state enrollment data and sort out issues like this. Gaba, on the other hand, is a team of one.

'For January, I ended up being off by about 500 people.'

There are arguably three things that set Gaba's site apart from others. The first is the speed: In a given day, Gaba will post enrollment reports from South Carolina, New York, Maryland, Nevada, and California. He'll work those into his spreadsheet and come up with new projections of, nationally, how many people have enrolled that day. He says he relies on a robust network of tipsters, who have sent in stories and state-level updates on a regular basis.

"The spreadsheet tells me what the average has been and I can extrapolate out nationally," he says. "So if I have data for 10 states for the first 10 days of March, I have a good idea of national sign ups."

The second is transparency: Every data point comes with a link, with every decision about how to count sign-ups — or retraction of data later found to have errors — announced prominently.

"I just kind of became the guy doing this. I started to feel obligated to keep going"

The last distinction is a sort of simplicity to his work. Gaba only tracks the number of people who pick an insurance plan on the exchange. He does, for example not attempt to estimate the percent of people who have paid for their insurance plans, a space where data is still sparse. This has drawn some criticism from Obamacare opponents, who argue paid accounts is a better way to measure who is actually gaining coverage through Healthcare.gov.

"It's a legitimate question and there's nothing wrong with asking," Gaba says. "But it's not the thing that I'm tracking."

Gaba's obsessive data collection has led to remarkably accurate data projection. The Obama administration won't release estimates of daily sign-up totals; Gaba will, and does quite well.

In January, Gaba projected that Obamacare would hit 3.3 million sign ups. The actual number, released by the Obama administration five days later, was 3.299 million. His February projection was 4.2 percent off from the final-sign tally.

"For January I ended up being off by about 500 people," Gaba says. "It was ridiculous."

This is how Gaba was able to predict on his blog that the White House had hit 6 million sign-ups in the last week of March before the White House had even announced the news. One administration official suggested on Twitter, after that he should "take a victory lap."

'I'm getting a little nervous.'

Gaba doesn't fully understand how he became so entangled in the enrollment numbers — except that it's become enough of a time commitment to negatively affect his actual web development business.

"My website clients are like 'that's great you're mentioned in the New York Times but where's my website,'" he says. "I'm a little behind on some of my work. I've had to give a few good-faith discounts as an apology."

We spoke the last weekend of March, just before open enrollment was slated to end. Gaba had cleared his schedule for March 31 and April 1, expecting a deluge of media requests and tracking data. At that point, he was forecasting 6.7 million sign-ups by the end of March (a number he revised upwards to 6.9-7.0 million mid-day March 31 , as reports of a sign-up surge grew).

"I'm getting a little nervous," Gaba told me. "There are so many variables. I know there's going to be a surge but don't know how much of a surge."

Gaba was initially planning to quit blogging enrollment numbers at the end of March — until the Obama administration announced it would allow some sign-ups to continue into April. So now he's committed to running his website through the end of this open enrollment period. He's not sure about next year.

"It seems to be filling a need right now, so I'll stick with it," he says. "By the end of the month, hopefully everyone will move on. I think, hopefully, the site will have served its function."

X
Log In Sign Up

forgot?
Log In Sign Up

Please choose a new Vox username and password

As part of the new Vox launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to Vox going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new Vox username and password

As part of the new Vox launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to Vox going forward.

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.
Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_10934_tracker