The back end glitches in Obamacare

by on October 6, 2013 at 10:24 am in Current Affairs, Law, Medicine, Uncategorized | Permalink

Very few of the individuals trying to buy health insurance are getting through to all the steps of the federal ACA website and using it successfully.  But once they register, they still may not have actual coverage plans, with successes running at what is (possibly) a one in one hundred rate:

As few as 1 in 100 applications on the federal exchange contains enough information to enroll the applicant in a plan, several insurance industry sources told CNBC on Friday. Some of the problems involve how the exchange’s software collects and verifies an applicant’s data.

“It is extraordinary that these systems weren’t ready,” said Sumit Nijhawan, CEO of Infogix, which handles data integrity issues for major insurers including WellPoint and Cigna, as well as multiple Blue Cross Blue Shield affiliates.

Experts said that if‘s success rate doesn’t improve within the next month or so, federal officials could face a situation in January in which relatively large numbers of people believe they have coverage starting that month, but whose enrollment applications are have not been processed.

There is more here, via Megan McArdle, who for years has been predicting major problems with the web sites.  By the way, these are not fundamentally problems of high usage or high demand.

foosion October 6, 2013 at 10:32 am

It’s somewhat amazing seeing people who refused to support funding for implementation and who refused to help fix drafting glitches now criticize implementation and drafting glitches.

Yancey Ward October 6, 2013 at 10:35 am

This is bullshit, Foosion. The money for the implementation was abundant and already appropriated within the PPACA itself. The administration has completely bungled it, and had done it all by itself.

foosion October 6, 2013 at 10:50 am

The PPACA contained some funding for implementation. It was obviously inadequate.

Today’s screamers have been actively working against a successful implementation. They have no standing to complain.

Yancey Ward October 6, 2013 at 11:01 am

This is too rich. Nothing the “screamers” have demanded has actually been put into law- nothing, so they can’t be blamed for the shitty job HHS has done in setting the exchanges up. The “screamers” did warn you that this was being implemented in a terrible way and was likely to fail on day one despite all the promises of its supporters, and now you are here trying to blame them for being right.

Z October 6, 2013 at 1:02 pm

Whenever someone in a panic, like poor foosian, starts using the word “screamers” you know you are dealing win an hysterical moonbat. There’s no reasoning with someone who is having a type of psychotic break that causes them to think the raging voices in their head are actually the bogeyman coming to get them.

stubbs October 7, 2013 at 12:12 am

I’m still wondering what “somewhat amazing” means. Can anyone explain? Is it somewhat accurate?

Megan McArdle October 6, 2013 at 11:26 am

I don’t really think you can blame Republicans for the fact that the administration apparently didn’t even award the contract until December 2011, and delayed the deadline for states to declare whether they’d be running an exchange multiple times–the two pieces of information from which I predicted that they would have problems. The issue wasn’t money, but time; they didn’t leave themselves enough of it for the complexity of the task they’d set themselves.

AFAIK, the money the administration has been asking for (and denied) are for policy/outreach services, not IT; adding bodies to a programming project doesn’t always make it go faster, as any developer will tell you.

asdf October 6, 2013 at 11:41 am

“adding bodies to a programming project doesn’t always make it go faster, as any developer will tell you.”

Yes, but micromanaging everyone’s job to “keep cost under control” does make things go slower. I shouldn’t be getting reassigned from a project two weeks before go-live when its not done and still buggy because someone decided a particular charge code is out of money so now I need to be assigned to a whole new project (I’m not even qualified for BTW) because that is where the money is that week.

I went through four projects and four bosses in a month once entirely for charge code reasons. All of that comes back to cost, it certainly was never done for a business reason.

Claudia October 6, 2013 at 11:59 am

Well if it were so easy maybe they’d already have been insured? Expecting better doesn’t mean forgetting the counterfactual. Small steps, right?

Jan October 6, 2013 at 1:07 pm

Yes, these are really significant problems. Extending the deadline for states to decide whether they would run an exchange was a bad move in hindsight. Nobody had any idea so many governors would refuse to set up their own exchanges. I think HHS assumed much of this was just early posturing based on idealogical opposition to the ACA, not necessarily a flat out refusal to eventually build an exchange to help citizens purchase insurance in a way that best fits the particular state’s environment. It was a terrible assumption, even if it was intended give states more flexibility.

I think there is enough blame to go around, however, as this was precisely the intended effect of the opt-outs. “The more states that opt out of the state-based exchanges, the harder it will be for the federal government to fully implement Obamacare and the more likely it will be that we can turn back the clock and reframe the health care debate,” Tea Party Patriots wrote to supporters ahead of the deadline for states to decide whether they would operate an exchange.

bmcburney October 6, 2013 at 3:52 pm


As a news consumer, I have been surprised by the lack of interest in the question of why we are having a Government shut-down given that Obamacare is obviously “not-ready-for-prime-time.” As I understand it, Obama refused to discuss a delay when doing so would have made the eventual roll-out of his signature achievement at least somewhat more effective and, at the same time, would have avoided a Govenment shut-down. Sounds like a win-win from Obama’s perspective. The only speculation that I have heard is that Obama wanted to use the the Government shut down disaster to distract attention from the Obamacare roll out disaster. Was it that or was Obama so out of touch that he was not aware that the roll out would be a disaster?

Shouldn’t somebody ask Obama or Carney or somebody about this?

Lonelylibertarian October 7, 2013 at 7:37 am

Am I missing something…

It would seem that ONLY the following are really needed…
Marital Status
Income [for subsidy calculation
Current Insurance – employer provided or other…

That would not seem to be hard…

Or are they trying to collect/add additional data that is “nice to know” or “would be nice to know” – but really doesn’t affect
the application – like pre-existing conditions…

Lonely Libertarian October 7, 2013 at 7:39 am

Oh yes they can blame Republicans – that is the way these folks avoid dealing with reality – in fact it is not just the GOPs fault it is Bush’s fault!

Careless October 6, 2013 at 2:54 pm

The PPACA contained some funding for implementation. It was obviously inadequate.

I’m just delighted by this comment/thread.

bmcburney October 6, 2013 at 3:05 pm


The “funding for implementation” was a part of the deal. Proponents got the CBO scores they needed to get the thing passed by lying about the costs of implementation (among many, many, other things). Opponents said the costs were grossly under-estimated from the beginning, how did they lose the right to complain when their original complaints are shown to be correct? It is proponets who now lack standing to complain about the “obviously inadequate” funding, they got exactly the amount of funding they claimed was necessary to make the thing work.

Jon17 October 7, 2013 at 10:42 am

Cheap solution….. roll out healthcare by alphabetical order of your last name. Those that starts with the letter A register on dates xxx-xxx. Problem solved. Cost = $0.

This is standard tactic for software developers. Everyone that has signed up for gmail via the early invite system knows this. I’m in IT and I really don’t understand why they can’t get this otherwise simple task done.

derek October 6, 2013 at 1:43 pm

It’s yours, you own it. Good heavens, there are 5 decades of experience with complex software implementations and example after example of failure. Poor initial design, complex and contradictory rules, scope far too large. Funding is not an issue, it is the task that is impossible.

But no, if we opposed this law we didn’t care about the children.

Go to hell.

Yancey Ward October 6, 2013 at 10:33 am

I tried 11 times between Tuesday and Friday to just create an account on the federal exchange, and failed 10 times to get past the security questions page (the questions were blank in the drop down menu). I finally made it past this page on Friday, but then got an error message that the account couldn’t be created at this time.

This is a total disaster in my opinion, and is far worse than anyone predicted.

dearieme October 6, 2013 at 10:46 am

Good God, y’all have come up with a system worse than the NHS. I’ll bet that took some doing.

Who’da thought you could find a President as bad as W? And yet you did.

And now I’ll return to trying to understand the significance of the government’s murder in Washington the other day of a poor madwoman who had had a minor traffic accident.

The Anti-Gnostic October 6, 2013 at 11:17 am

American democracy was built by and for self-sufficient capitalists who’d had enough of feudal, bickering Europe and the UK. That ethos no longer exists in the US. We are coasting on capital built up under a high-trust environment which enabled the rule of law. When that trust evaporates, as inevitably happens in a diverse polity, the US will be hell on earth. There will be nowhere to hide.

I think also we get the worst of all worlds because we’re trying to implement policies developed by and for Continental nation-states in a country that was never anything of the sort.

Racist October 6, 2013 at 2:42 pm


Non Racist October 6, 2013 at 9:53 pm


TMC October 6, 2013 at 9:54 pm

= +2! Anti-Gnostic is a uniter!

anne October 6, 2013 at 10:46 am

I tried for three days straight and could never get past the first page. The message was something to the effect of “we will get to you as soon as we can.” I had never been put on hold by a web page before. Baffling.

AADL October 6, 2013 at 10:47 am

It was inevitable that Ocommiecare would be a disastrous failure. The question is when will it be repealed.

Yancey Ward October 6, 2013 at 11:03 am

See, it is people like you that are causing its failure!!!!

The Anti-Gnostic October 6, 2013 at 11:19 am

Saboteurs. Goldstein.

Chip October 6, 2013 at 11:12 am

Fortunately we have a president who can look past these minor glitches and has the strength to shut down national parks and other normally unstaffed facilities so that the US can finally enjoy its own version of the NHS.

And applause too for the American people, so thankful for the wealth and freedom they inherited that fully 40% can name the three branches of government.

MichaelG October 6, 2013 at 11:17 am

Despite the press reports, I’ve gotten through to “Covered California” every time I’ve tried. I’m not surprised there are load problems though. The pages are gigantic messes of Javascript and style sheets. For example, the plan comparison page is a 4 line form (number of people, income, zip code, ages). You could have written this as a simple HTML form with a few decorative images.

Instead, the page is 2.2 megabytes itself, plus loads 6 style sheets and 17 javascript files. That’s 24 hits (plus images) on the server, when at minimum, you could have done it with one. So they’ve created a lot of extra load by using way too much page design goo.

I also think the extra UI code isn’t helping. For example, instead of having a few blank lines and telling people to fill in the ages, it has a single age field plus an “add adult” button which generates the list. This is just confusing. There are also “helpful” popups for things like zip code that just get in the way.

Once you get through, you see the plans offered. If you input $20,000 as your income, you get offered Bronze plans which are only $1 a month after subsidy. But the maximum yearly out of pocket is $6,350 and the copays are very high. $300 for an ER visit, and 30% of test costs will be out of reach for anyone in that salary range.

For someone making minimum wage, it was a much better deal to just go to the Emergency room and not pay.

Steven October 6, 2013 at 11:19 am

I see an interesting paper coming out of this.

What’s the placebo effect of insurance? The Oregon Medicaid study revealed some evidence suggesting the just having insurance improved some self-reported health outcomes, even before any additional health care was received. If Obamacare results in a large number of people thinking they have insurance when in fact they do not, we could get more evidence on this question.

zbicyclist October 6, 2013 at 12:01 pm

But IIRC, the Oregon study revealed that it didn’t have much effect on actual health outcomes. If we just want people to feel better, legalizing marijuana would be a lot cheaper than a large expansion of Medicaid.

asdf October 6, 2013 at 11:35 am

Ironically, attempts to control spending in PPACA are often counter productive. When I first started working on the exchange project they just told us to get the work done and charge whatever we needed on whatever code. Then people complained we were over budget (probably people that don’t like the law and have been trying to defund it) so they implemented a retarded top down system of micromanaging everyone’s jobs based on charge codes. The net effect is that we all still got paid (costs didn’t go down), but that we weren’t able to work on whatever we actually needed to work on unless we happened to get assigned to the right charge code that week.

This isn’t some normal project where cost control is what’s really important. Getting the thing working was most important, and I doubt anyone bidding on the project had a real good idea what it would cost. If people go over budget just pay it. If they are skimming just pay it. Attempts to contain cost don’t contain any cost, if anything the inefficiencies of such a system increase cost as you have lots of people getting paid but not working on what they need to because some dude several management layers up that has no clue thinks you should be working on some different charge code.

If this was some non-critical simple project with a long time horizon then worry about cost. It isn’t. It’s a complex thing you need to make work quickly. The costs of it not working are orders of magnitude greater then the cost of implementation. Just pay whatever it takes.

zbicyclist October 6, 2013 at 11:57 am

A bit off the point, but I wonder if there’s an inverse relationship between the number of chargecodes (an indication of micromanagement) and the success of a project?

derek October 6, 2013 at 1:47 pm

>This isn’t some normal project where cost control is what’s really important.

Tee hee! This is betting beyond parody.

asdf October 6, 2013 at 11:17 pm

The goal is to reform the single largest sector in our economy. That is the revenue line. The cost line for the exchange is paltry in comparison.

TallDave October 7, 2013 at 2:56 pm

Sheer nonsense. The goal of Obamacare is not to reduce health care revenue, it’s to consume more taxpayer dollars in order to insure the poor. The cost of it not working is less than zero.

Megan McArdle October 6, 2013 at 2:00 pm

I’ve long said that government micromanagement of its employees is a huge problem–makes the work worse, and more expensive in the long run. If you’re willing to talk about this (anonymously, of course) please drop me a line at

Jan October 6, 2013 at 3:17 pm

This is a catch 22. The micromanagement is required for any work that could be potentially controversial, to make sure some fatal mistake doesn’t get missed and used to fuel a politically charged oversight campaign by the administration’s opponents (e.g. IRS targeting scandal). If a business doesn’t micromanage its employees, the biggest risk it takes is potential failure. In government, such risks are unacceptable, even if 99% of the time it would be better to just let the civil servants and contractors do their jobs.

zbicyclist October 6, 2013 at 7:36 pm

I have little faith that micromanagement avoids political scandal. I’d put it a bit differently: micromangement / more bureaucracy / task forces / blue ribbon committees are things that are designed to look like there’s effective management, and in themselves are often barriers to effective management.

The IRS targeting scandal would seem to be an example of this, since I would imagine the IRS is extremely bureaucratic and procedure-driven. But I have not personally worked there.

Dan Weber October 7, 2013 at 9:09 am

Micromanagement doesn’t avoid scandal so much as it allows you to push the scandal of failure off onto someone else.

asdf October 6, 2013 at 9:29 pm

Failure = controversy. Having some CYA stuff after you fail isn’t going top protect you the way success does.

Richard Nixon October 7, 2013 at 11:14 am

Yes, muzzling politically charged oversight campaigns by the administration’s opponents leads to more efficient government.

J1 October 6, 2013 at 5:03 pm

“Then people complained we were over budget (probably people that don’t like the law and have been trying to defund it) ”

That need not be the case. My experience with government projects is that all of them have the characteristics you describe, except that normally they have the top-down micromanagement from the beginning.

TallDave October 7, 2013 at 1:06 pm

The costs of it not working are orders of magnitude greater then the cost of implementation

That might be true for say, Amazon or Google, but Obamacare portals don’t generate revenue, they consume tax dollars.

What people don’t appreciate is that there are literally thousands of failed attempts to imitate Amazon. It was therefore infinitely more probably the Obamacare portal would look more like the failures than the success.

The bottom line is this stuff is really hard to do, and government should leave it to the private sector where failure winnows the bad code.

If people go over budget just pay it. If they are skimming just pay it

That’s a fine philosophy when it’s your money. Using taxpayer money makes it problematic.

F. Lynx Pardinus October 6, 2013 at 11:58 am

There’s good comments from people who actually have worked on government software projects here:

Adrian Ratnapala October 6, 2013 at 12:06 pm

Private companies like to role these things out in stages. This failure-on-day was nearly inevitable. But it is just as inevitable that they will steadily pull their shit together.

But I am strucl by by adf’s comment about the inefficiency of cost-control beurocracy:

Getting the thing working was most important, and I doubt anyone bidding on the project had a real good idea what it would cost. If people go over budget just pay it. If they are skimming just pay it. Attempts to contain cost don’t contain any cost, …

Which is just about saying that if government demands something, it cannot be supplied efficiently.

dirk October 6, 2013 at 2:33 pm

The incentives (to individuals) on this kind of project mean the winning bidder will have been inevitably among those who offered the most optimistic and therefore least realistic timelines. Because that’s how you win a bid like this when dealing with a customer like this.

DKF October 6, 2013 at 1:20 pm

I tried only once, unsuccessfully. Since I only wanted a ballpark estimate of what my increased premiums would be under ACA, it seems that exchange functionality was beside the point–it appears to be geared mostly towards calculating subsidies and medicare eligibility, none of which will be applicable to me. Therefore, It was easier to simply go to my provider’s website (blue cross) and figure out what their bronze plan pricing is (compared to existing $10K/100% plan). Valid price-only comparisons (e.g. w/o subsidy) can still be made on providers websites or

I suspect that many people will migrate to alternative sources of information as they give up trying to use the main portal–but there’s still no excuse for this level of performance, and especially there’s no excuse for failures after an initial successful step…just poor planning all the way around. The fact that federal officials failed so miserably at this task, despite a clear understanding of its importance, seems to me a clear harbinger for their prospects at other, more difficult tasks, such as the collection of excise taxes, vetting of potential new entrants, or the enforcement of the individual mandate penalty.

Rusty Synapses October 6, 2013 at 2:09 pm

I’ve thought Obamacare was a mess from the beginning, but I am depressed at the tone of a lot of these comments – I’ve given up reading comments on the WSJ, it really is depressing the maniacal glee at any negative development related to the country (not just Obamacare glitches) you’ll find there. I hate to see it spread to other places. If the implementation of Obamacare weren’t such a mess that it didn’t need help, I’d wonder whether some on the right would actively work to make it fail (a la denial of service attacks). Good people should be hoping to be proved wrong, not happy to be proved right. I still hope that they get through these glitches, and then continue to modify the exchanges so that they are a more market oriented solution for healthcare (with eventually all employers dropping tax free health insurance, and just paying more cash – in my book, that’d be a feature, not a bug).

mike October 6, 2013 at 3:16 pm

Lots of people here think it is an extremely bad law which will become a disastrous but untouchable gimmedat program if it is allowed to take root. Big, spectacular early failures increase the probability that the law can be aborted or seriously altered before it metastasizes. This byzantine bill just barely limped through the legislative process at the last second in a 100% partisan power grab by a crumbling majority on the precise premise that once done it could not be undone. Now you’re feigning outrage that the people who were the victims of this usurpation haven’t just rolled over and died. Boo. Hoo.

Rich Berger October 6, 2013 at 3:44 pm

Ditto. This thing is fundamentally flawed, and anti-liberty, to boot. You would think that setting up the interface would be the easiest part of the project.

Willitts October 6, 2013 at 6:26 pm

The more general question is when government can cancel a boondoggle after implementation. Is there a minimum level of rot that must be obtained before you can get rid of it? It seems it either has to be an abject disaster or declare victory and quit.

Maybe Dodd Frank should have included a living will for government programs.

Jan October 6, 2013 at 5:19 pm

Every major government-administered program had serious opposition and problems rolling out. Of course it could be going much better, but in 3 years the law will be a ton more popular and repeal will not be a focus of the R presidential candidate (unless the Republican party really goes off the rails, which is the direction they are heading now, though I think sanity will prevail). In 5 years, the ACA will be nothing more than an afterthought for even quite conservative legislators. It will have helped tens of millions of people and moved us more toward a rate of health care coverage that other rich countries have had for decades. The biggest source of the boo hooing is same the group it started with. I wish them luck.

Kevin October 6, 2013 at 6:19 pm

Just out of curiosity, what makes you think the R presidential candidate in 2016 won’t run on repealing the ACA? I don’t think they’ll succeed (as in, I don’t think we’ll have a Republican president for a while), but to me it seems a no brainer that they’ll try.

Jan October 6, 2013 at 8:46 pm

After a couple years up and running, I venture most will actually like the law, and any major implementation issues will have blown over. Probably the Republican candidate will pay lip service to repeal — by then it may just be “reform” — but they will have fought and lost on it so many times that even the more conservative wing of the caucus will have mostly let it go. If they control House and Senate, they will be emboldened, but I don’t think that happens.

TallDave October 7, 2013 at 1:09 pm

At this point I suspect the Democrat will also run against Obamacare.

Cliff October 6, 2013 at 8:37 pm

So you have a crystal ball then?

TallDave October 7, 2013 at 1:08 pm

Every major government-administered program had serious opposition and problems rolling out.

That claim has already been debunked that several times. SS and Medicare were widely popular.

Ad Nauseum October 7, 2013 at 1:13 pm

Except many of these major government-administered programs have grown into large burdens on the public and political footballs for those in office (eg. Social Security, Medicare, Medicaid, etc…). The debate is not over, and another large program, which is much more poorly planned than the previous examples, may be more than the American public is willing to just deal with. The debate is far from over, I expect some form of mandate and pre existing condition clause to remain law, but the rest of the 2000+ page monstrosity that is the ACA to be stripped away and heavily reformed in the coming years.

Rusty Synapses October 6, 2013 at 7:47 pm

Um, I’m not feigning outrage at anything. I agree, it’s incredibly complicated. I agree, it was passed in a questionable process. I agree, Obama is on questionable grounds changing the deadline for the employer mandate and making other changes to the law. I’m also not a big fan of the 3.8% tax on unearned income – makes our tax code even more complex, and it’s amazing how little discussion it got. Also, I’m not surprised at all that that the government bungled a giant software development project – any rational person should be surprised if it didn’t. However, flaws aside, there are market elements to it, and don’t forget exchanges were originally market-based idea of the right (the “appropriate for states only” Romney dodge is silly). Also, I do think we have a real problem in healthcare, given that the market isn’t really set up to work in the existing status quo (i) tax-free nature of employer provided healthcare incredibly distorting, (ii) no way to buy insurance (outside of employment) that will protect you against multi-year risks (no “20-year fixed term life insurance” equivalent), and (iii) pricing/purchasing/market mechanism is really broken – no transparency, often no ability/time to shop – list prices aren’t true list prices. I just wish that (1) people weren’t so happy with the failure, and (2) the Republicans were actually trying to propose real solutions – then or now – either small tweaks or big replacements – I think there could be a pony in this pile of manure. (Sort of like I wish Obama was trying to propose real solutions on entitlements, or had taken the grand bargain – there’s a lot of cowardice to go around.)

lonely Libertarian October 7, 2013 at 7:51 am


You left out the “Elephant in the Room” that nobody wants to confront – and unless it is dealt with will doom this program…

The system we are about to “lock in place” is designed to pay Doctors, Hospitals, Drug Companies and Medical Device manufacturers a LOT OF MONEY.

The 18% of GDP that we spend on health care goes to REAL PEOPLE and REAL BUSINESSES – the money goes to them – and much of what was done to
get OC passed was to find ways to get the folks to go along quietly. Insurance companies were give bigger markets via the exchanges and guaranteed profits…
Medical Device companies saw great opportunity in the whole preventive care scam that was included [yes you will be paying a tax - but just think how many more scans
tests you will be doing!]. Even the Docs get more patients who can pay in the deal – and are not forced to post price schedules.

A friend of mine once suggested that we could understand what is wrong with our Health Care system by just comparing how nice on average our major hospitals are to how nice our schools are….

Follow the money…

Rusty Synapses October 7, 2013 at 9:05 am

But do you think it might be easier to follow the money with an explicit subsidy (the subsidy for insurance at least)? It seems like once you have variables, it’s easier to tweak them, or at least see them, if it’s not politically possible to reduce them. (Maybe I’m a dreamer, but I see this as closer to the “replace all safety nets with a market based solution, and just give people cash – explored in another item on this site a few days ago.)

joshua October 6, 2013 at 2:35 pm
NPW October 6, 2013 at 4:53 pm

Poor implementation does not convince me that a program is a bad idea. History is littered with great ideas that were executed poorly. In this particular case, I was never a fan, but still. The belief that because the website is bad, the idea is bad, is just poor logic.

The fact that it faces issues with being a federal government developed website is so far down the list of reasons to oppose ACA Nothing that people have been complaining about is anything new to people who work in the government sector.

The gov always tries to have the baby of the month by getting nine women together for one month, and it is dumbfounded when that doesn’t work. Charge codes are a constant source of annoyance.

J1 October 6, 2013 at 5:27 pm

“The belief that because the website is bad, the idea is bad, is just poor logic.”

That depends on whether the idea is the goal itself, or merely a step towards achieving it. If we agree on a goal, disagreement about how to achieve that goal isn’t necessarily poor logic.

john personna October 6, 2013 at 6:11 pm

The complaint is that data sets are incomplete. Tyler asserts that this is not related to site load. A great many commernts then say “Yes! I could not complete my process (and therefore data) due to site load.” Hmm. Possibly contradictory evidence.

Yancey Ward October 6, 2013 at 6:47 pm

I seriously doubt the load issue. I had to wait 30 minutes to get the create account stage at 2:30 a.m. on Wednesday morning, and still got hung up by the security question bug. Did all the people who failed on Tuesday morning/afternoon get up in the middle of the night to buy health insurance? I doubt it, and the inflated numbers of “visitors” is also putting lie to the claim of too many people trying to get in.

john personna October 6, 2013 at 7:10 pm

I have fixed large, busy, and broken websites. Not on this humongous scale, but national multivendor enterprise size. I am sure others here have as well. I hope they’d agree that you can’t say you have a fix until it is installed and tested. Before then the smoking guns and red herrings compete for attention.

For that reason I know that the folks trying to back trace errors under this crush of usage have a very have job.

john perdonna October 6, 2013 at 7:13 pm

s/have/hard/ (stupid phone)

john personna October 6, 2013 at 7:26 pm

(I skipped past your questions because they require server logs to answer, something none of us have.)

JonFraz October 6, 2013 at 6:35 pm

Tyler, have you ever bought insurance online? I have (my current auto policy with Geico). You don’t have a policy when you done with the application. Instead a representative calls you to verify your information and your intention and obtain any additional details that are needed (in my case Geico needed info about my health insurance so I could avoid being charged for personal injury protection, which is mandatory here for those without health coverage). I would be surprised if the exchanges do not envisage a similar call back process.

Careless October 6, 2013 at 7:32 pm

I’ve gotten insurance with Geico and never talked to them on the phone.

joshua October 6, 2013 at 8:22 pm

Ah, so that’s why Obama was saying it would be as easy as buying airline tickets.

Dan Weber October 7, 2013 at 9:38 am

In fact, I’ve bought health insurance online from Blue Cross without talking with anyone on the phone. Just last month, too.

Bill October 6, 2013 at 7:38 pm

I am a bit skeptical. The persons claiming difficulties in the newspapers have been insurance agents, who, interestingly, will lose money if persons buy on the exchange because they will not be getting a commission.

Dan Weber October 7, 2013 at 9:38 am

Journalists have been trying, too. It’s not like it requires a whole lot of investigative background to use a website and say that it had problems.

8 October 6, 2013 at 11:02 pm

Eventually, bad stuff will be blamed on Republicans having bad juju.

Steve J October 6, 2013 at 11:35 pm

Thought I would go through the process to see what people are complaining about. The second step in the process is picking your username. Here are the username requirements:

The username is case sensitive. Choose a username that is 6-74 characters long and must contain a lowercase or capital letter, a number, or one of these symbols _.@/-

The way I read this requirement any email address that is at least 6 characters should be valid – yet mine was not. Is it possible the “or” should be replaced with “and”? I see a chat button. I’m sure a friendly Obamacare representative will answer my question:

[10:49:45 pm]: Thanks for contacting Health Insurance Marketplace Live Chat. Please wait while we connect you to someone who can help.
[10:49:48 pm]: Please be patient while we’re helping other people.
[11:15:09 pm]: No such user exists

Add capital letter and number to email address and successfully get to next page. 3 security questions. Answer all with “Obamacare”. Sorry your answers must be unique. Get frustrated close tab. Remember the people using this site do not have employer provided healthcare. Back to Site is overloaded we will save your place in line. Close tab. Sorry poor people go get a decent job.

Brian Donohue October 7, 2013 at 11:19 am

Heh. I always thought a Kafkaesque world could be largely automated.

Gordon October 7, 2013 at 12:53 am

When news came out that the site was having performance problem, I had a suspicion that one of the causes was a mistake that I’ve seen many web application teams make. There’s a feature that’s enabled on web servers by default that improves performance for the users IF there are only a small number of users visiting the server at any one time. I can see from the http response from that this feature was left enabled and is probably playing a big part of the performance issue. The fact that they don’t know this does not inspire confidence that they know what they’re doing with other aspects of the site architecture.

Gordon October 7, 2013 at 12:59 am

I should have elaborated a bit further on the feature that I mentioned. Even though this feature is meant to enhance performance, it will cause users to have terrible performance if a large number of users are attempting to access the web server. When I’ve encountered coworkers that had trouble with their web application performance due to large numbers of users, turning off this feature solved the problem.

cyd October 7, 2013 at 7:29 am

> “By the way, these are not fundamentally problems of high usage or high demand.”

Please to elaborate. It’s pretty strange to put such a strong statement at the very end of the post, completely unsupported and unsourced.

Ape Man October 7, 2013 at 9:47 am

I imagined that he was referring to something like this.

Though I have to say, calling some 8 million users in a nation of 300 million people “high traffic” seems a bit rich to me. Just how many people did they expect would use a “national exchange”?

Boonton October 7, 2013 at 5:05 pm

I believe the estimate is about 7 million or so. If you’re covered through your employer (say 50%), then the exchange will probably not offer you a better deal. Medicare/Medicaid likewise wouldn’t be a factor (another 40% say?). Even after that there is still private insurance which people can and do buy which is neither employer based nor purchased through the exchange as well as people who simply will not buy coverage no matter what incentives there are.

Boonton October 7, 2013 at 10:28 am

The website is not Obamacare, the website is an implementation of it. If you told me Social Security’s website was bad, I’d shrug and say they should fix it but that in itself says very little about social security (which as enacted decades before before anyone had a computer let alone a web site).

The exchange is a very simple idea. A person seeking insurance goes there to buy. Companies seeking to sell insurance go there to sell. In theory you could do this with an 800 number or on paper application until the web site’s kinks are worked out. Those hoping Obamacare dies because of website glitches are, IMO, going to be very disappointed.

TallDave October 7, 2013 at 11:23 am

By the way, these are not fundamentally problems of high usage or high demand.

IIRC Amazon handles roughly 100x the load, with an error rate around .01% (as opposed to 99% for Obamacare). This includes a vast network of back-end connections with affiliates dwarfing what Obamacare has to deal with.

What people don’t appreciate is that there are literally thousands of failed attempts to imitate Amazon. It was therefore infinitely more probably the Obamacare portal would look more like the failures than the success.

The bottom line is this stuff is really hard to do, and government should leave it to the private sector where failure winnows the bad code.

Boonton October 7, 2013 at 1:07 pm

Except something very much like this has already been done. Medicare D has a gov’t run ‘portal’ where you go, put in the medications you commonly use and it spits out a list of different insurance plans and their associated premiums letting you pick which one you want. is running a database of millions of products, customer reviews, purchase histories etc. That is not necessary for the healthcare portal. Once you end up choosing an insurance company, it’s up to that company to give you a website portal to track your account and spending. In essence the healthcare portal is more of a referral service, a bit like lending tree. You put your info in and maybe it will pass some or all of it along to the company you ultimately choose but the hard work is done after the ‘handoff’.

TallDave October 7, 2013 at 1:11 pm

So your argument is that Obama is much less competent than Bush, and vastly less competent than Bezos? No argument here!

Boonton October 7, 2013 at 2:46 pm

See previous comment, Obamacare is not a website nor does a buggy website have anything useful to tell us about Obamacare. If you want to discuss javascripts and HTML coding I suggest consulting an IT orientated blog.

TallDave October 7, 2013 at 2:52 pm

Not making much sense. The issue isn’t whether “Obamacare is a website” (clearly it HAS one) nor were we discussing “javascripts.”

The Obamacare portal has been ineptly implemented by the executive branch of the government, which is headed by Obama. That’s just a fact.

Boonton October 7, 2013 at 5:02 pm

If your point is a website doesn’t work as well as it should in the first weeks of its launch, then yes I give you that point. Somehow I’m pretty sure the Medicare D website also had downtime and problems, but to be honest I don’t remember. In the long run the exchange is not at all that complicated and there’s no reason the web portal should be a major problem. If this was the ‘diaster’ Obamcare critics were hoping for to sink the reform then you all will be very disappointed. That’s all.

But other than that yes yes a new web site launched by the gov’t should work perfectly from the moment of its launch and the IT people and contractors should be red faced and get yelled at until the thing is fixed.

TallDave October 8, 2013 at 10:37 am

The Medicare Part D portal never had a 99% failure rate.

So, your argument is that Obamacare can’t even get the easy parts right. Again…

TallDave October 8, 2013 at 1:46 pm

And it’s not just the technical issues. In SC, people have posted screenshots of Obamacare policies with $10K deductibles. CA isn’t much better.

Ad Nauseum October 7, 2013 at 1:31 pm

In the end, no matter what happens, it will probably be Bush’s fault.

Floccina October 7, 2013 at 2:28 pm

It is early yet. If the things are still not working in a few months then they screwed up bad.

Lee A. Arnold October 7, 2013 at 3:42 pm

Wow, the software has glitches and there are no such things as hackers for hire, to fix stuff in a few weeks… It’s all jus’ too complicated! It’s the end of the liberal ascendancy!

Comments on this entry are closed.

Previous post:

Next post: