So the administration has given up on trying to implement the exchanges as they were envisioned in the law. I can’t say I am surprised. But I am disappointed.
I suggested that this was likely to happen in December of 2012, when I tweeted that “Like Michael Cannon, I think that there is roughly a 0% chance that Feds will have state exchanges running by deadline.” I linked a story by Dave Hogberg about the decision, by the Department of Health and Human Services, to extend the deadline for states to declare whether they’d be running an exchange, or letting the federal government do it for them. HHS extended the deadline several times, ultimately giving the states until March of 2013 to decide. Since the exchanges are supposed to go live in October, I didn’t see how the administration could possibly be ready in time.
I was challenged on this by supporters of Obamacare. The administration was saying that things would be ready on time; what reporting had I done to say differently?
The answer was no reporting; mostly, I was drawing on my experience doing IT implementations. Six months is an extraordinarily aggressive timetable to start doing an IT project, even one that’s what one of my colleagues used to call “shake and bake”: take it out of the box, add a few of your own ingredients, and serve it up.
Yes, the administration had been working on Obamacare since it passed in spring of 2010. But it seemed to me that just the build phase would take longer than they were allowing, because Obamacare had a lot of complicated parts; it’s basically a giant Rube Goldberg machine that breaks if any of its major pieces goes even slightly awry.
As the law (and all the wonks) had described them, the exchanges needed real-time data from the IRS, in order to calculate whether purchasers were eligible for subsidies, and how big those subsidies should be. In order to determine subsidy eligibility, they also needed to know whether your employer had offered you health coverage costing no more than about 10% of your family income, as they were required to by law (and subject to a fine if they didn’t). Only those who couldn’t get affordable employer coverage were eligible to purchase subsidized insurance on the exchange.
That’s two sets of big, secure databases that needed to be hooked into just for the subsidy part. One of them had to be created from scratch, and it wasn’t clear to me that the IRS was ready to deliver real-time payroll data, either.
Then there was supposed to be some sort of hook into the state Medicaid systems, for those below 133% of the poverty line. There was also, of course, the part where insurers sell you insurance. That seems like an enormous project all by itself.
As if that wasn’t hard enough, most of the exchanges were being set up under federal procurement rules, which make it slow and ponderous to commission software and buy equipment. When I was working for banks and other financial firms, if the developers needed a bank of new servers for an important project, I could get that signed off and shipped within a week, two at the outside, as long as the powers-that-be really wanted it to happen. That just doesn’t fly in the government.
On the other hand, it’s now been ten years since I worked on an IT installation, and longer since I worked on a large-scale multi-site project. For all I knew, the process had gotten a lot faster since the long-ago days when I was in the business. And of course, I did oppose the law, so I’m probably naturally predisposed to assume the worst. It wasn’t unfair for my critics to be skeptical of my judgement; I was a bit skeptical myself. Why would the administration lie about being prepared to move on time? Wasn’t it more likely that I was missing something?
It turns out we were all missing something. I was right that the administration could not possibly implement the exchanges that had been described when selling the law, described in the law itself, and described by all the wonks who supported the law after we passed it and found out what was in it. But I was wrong that the administration viewed this as an essential part of “getting the exchanges up and running on time”. Instead, they were going to start jettisoning large chunks of what they’d promised.
Last Tuesday, while my husband and I were driving around the Great State of Virginia, the administration announced that the employer mandate was being delayed by a year. We were on holiday, so we refrained from blogging . . . but not, I’m afraid, from chewing over our puzzlement. Why delay the employer mandate? As provisions of the law go, it’s relatively easy to implement, and it’s a significant component of both cost control, and coverage expansion. The administration’s explanation–that this was complicated and they wanted to be sensitive to business–made no sense. Nor did the follow-up explanations that were offered in the media, most notably that the administration didn’t want to see job losses in advance of the 2014 midterms. It’s been clear for a while that the law is causing lower-wage employers to rein in their full time employment, but delaying the mandate for a year doesn’t change that. No one wants to hire a full time worker you have to lay off in January 2015.
Besides, the employer mandate was a major part of subsidy verification; how could you rip it out of the system at this late date without compromising the whole verification process?
Question answered. On Friday, they announced that they were also not going to have the real-time IRS data, at least not for the state exchanges. Subsidy calculations in many of the exchanges will now be conducted on . . . the honor system. The most likely explanation for all of this is that the administration simply couldn’t get the system working in time, so they gave up.
So those of us who suspected that the exchanges were not going live as we (and the law) had conceived them, were correct. That’s why all the complicated functionality is getting ripped out of them.
(Though “ripped out” may not be the right word. The architecture needed to interface with all those disparate systems is pretty complicated; I’d be surprised to find out that they’re actually planning to rewrite all that code now, when they should be moving into the testing phase. Rather, I’d expect that they’ve known they were going to do this for months and months, and never put the functionality in there in the first place.)
And yet, my critics were not necessarily wrong either, because they didn’t see this as losing core functionality. Ezra Klein basically sums up what I take to be their point of view, saying that “Obamacare just got easier to implement“. Which is absolutely true–as long as you define large parts of the promised functionality as being inessential to Obamacare.
In one view, Obamacare really has only one feature: putting as many people as possible onto the insurance rolls. As long as you’re doing that, nothing else matters–not cost control, not fraud prevention, not the words in the Patient Protection and Affordable Care Act. Take them away and Obamacare still exists, as long as there are exchanges with subsidies. I understand why people see it that way, but as you might suspect, I see some problems with this view of last week’s news.
The honor system is a really expensive way to hand out subsidies. In theory, if you lie about your income (or underestimate it), the IRS can make you give the money back. But as I understand it, there’s an interesting wrinkle with Obamacare: the IRS cannot seize the money from your bank accounts, or take normal enforcement actions as they would with other sorts of overpayments. The only way they can get the money back is to take it out of your tax refund. If you don’t get a refund–if you owe a bit at the end of the year, or you don’t file a tax return because you’re too poor–then all they can do is keep a tab and hope you have a refund later.
In theory there is also a hefty fine for deliberately misrepresenting your eligibility. But the section of the law which deals with the penalties says this:
(3) LIMITATIONS ON LIENS AND LEVIES.—The Secretary (or,
if applicable, the Attorney General of the United States) shall
(A) file notice of lien with respect to any property
of a person by reason of any failure to pay the penalty
imposed by this subsection; or
(B) levy on any such property with respect to such
So it’s not clear to me that the penalties are actually any bigger threat than the remote possibility that the IRS will raid your bank account to recover subsidy overpayments. And in the case of the employer mandate, they’re not even collecting the data, so there will never be any way to know if you’re lying about the availability of insurance from your employer, or are just looking to see if you can get a bigger subsidy from the federal government.
Naturally, losing the ability to verify eligibility before you hand out subsidies is going to hike the cost of the program. Remember when Obamacare was passed, and it was trumpeted as a giant deficit-reduction program? Half of the deficit reduction has already gone away, as unworkable elements like the CLASS Act and tighter 1099 reporting requirements were trimmed away, and beacons of hope like Medicare pilot projects slowly flickered out.
Well, there’s a good chance that we just lost the rest of the deficit reduction. Doing away with all the verification is likely to push Obamacare deep into the red. At Time, Kate Picket lays out the cost:
Whatever political benefit the Administration accrues in the delay will come at taxpayer expense, as the move will increase the cost of Obamacare. The provision was expected to generate some $10 billion in revenue in the form of penalties that employers not offering coverage would have incurred in 2014. Adding to that cost is the fact that many workers whose employers decide not to offer coverage in 2014 will now be eligible to receive federal subsidies to buy individual insurance policies through state or federal exchanges.
When the law was passed, deficit reduction was supposed to be a hair under $120 billion, though that actually rose slightly after it passed. But then we started lopping bits off the law, and the projected deficit reduction shrank to about $7 billion a year, or $70 billion over a decade. A little arithmetic will tell you that the annual cost of losing the employer mandate will push this program into a net fiscal drain.
But it’s just a temporary delay, I hear you cry. And maybe so. But I’m not so sure. The administration seems to be floating trial balloons about getting rid of the employer mandate entirely–and no wonder, given that a lot of employers were clearly cutting back on their full-time staffs in order to avoid it.
You can make the argument that getting rid of the employer mandate is much more economically efficient, helping to finally break the link between health insurance and employment. But whether or not you agree with this argument, getting rid of the employer mandate is inarguably much more expensive.
Nor would I bank on the other income verification components going live next year. Setting up a secure system to gather and share confidential income data without, say, accidentally exposing half the nation’s private data to hackers, is an enormous enterprise. And the administration is not exactly inspiring confidence in either its competence or its candor.
After all, it’s really pretty stunning that the administration is suddenly announcing, at this very late date, that it plans to gut half the functionality from the exchanges, at least temporarily. These systems are supposed to go live in under three months and we are just finding out that they’ve given up on controlling costs in favor of indiscriminately handing out subsidies to anyone who asks? It’s hard to communicate how last minute this is, if you haven’t worked on a big project, but you can get some of the flavor if you imagine a newly minted MD arriving to take her medical boards and announcing that that she’s decided she’s going to be a Physician’s Assistant instead of a doctor, so please administer that exam, folks. Don’t get me wrong–we frequently rolled back the scope of projects. But we didn’t wait until a couple months before the drop-dead date when the whole thing absolutely, positively had to be live.
One of two things must be true: the administration knew this was necessary long ago, but concealed it from the public and the congress in order to limit the time they had to react; or the administration is so incredibly inept that it has only just now realized that it wasn’t going to be able to handle any of the complicated bits. Either way, why would we assume that anything else they say about the systems–like, “It’ll be ready next year”–is true? Indeed, why should we assume that this is the last such revelation?
In fact, we know it’s not; this morning we learn that the administration has somehow inadvertently made it impossible for insurers to charge older smokers more than non-smokers. This is a small thing in itself. But what else is going to break?
This is not just a rather unflattering reflection on the administration’s ability to implement its signature achievement–didn’t they, say, talk to any IT folks about this? It’s also a violation of the promises that they made when this thing passed. They swore up and down that cost control was central to Obamacare; the bill would have been even less popular had they suggested that it would be deficit increasing, rather than decreasing our Gross National Overdraft.
But as it turns out, they’re all too willing to throw cost control overboard if it becomes inconvenient. This is pretty much exactly what critics said they were planning to do–a charge which was excoriated at the time as baseless slander.
I understand why the administration, and Obamacare’s supporters, view coverage expansion as the whole point, and cost-control and anti-fraud measures as incidental side measures that should not be allowed to get in the way of the main task. I think they are sincere, and their goals are worthy. But that wasn’t the deal they promised, or the law they wrote. The deal they promised was that they would be cost-controllers, not the authors of yet another budget-busting entitlement–that they would fight fraud, not abet it. Ironically, as I was writing this post, I saw the new ads the administration is running–touting their efforts to fight Medicare fraud.
Obviously, I don’t expect the administration to pack up and end Obamacare. But if they can’t implement all the relevant pieces in time, then they should delay the whole thing by a year, not just little things like the fraud controls they promised. Yes, that would be embarrassing. But it wouldn’t be as shameful as promising one thing to get the law passed, and delivering something entirely different.