Hacker News new | past | comments | ask | show | jobs | submit login
California’s $327 million web site in operation (law.harvard.edu)
94 points by tjr on Oct 2, 2013 | hide | past | favorite | 40 comments



I think I found their budget plan. Very, very big numbers.

http://www.healthexchange.ca.gov/BoardMeetings/Documents/CHB...


To be fair to them, California's population is 38 million. Divide the budgets by that number to get a per-capita cost and it doesn't seem crazy to me.

As an aside the way we citizens run our government, budget bloat is guaranteed. We complain a little about bloat, but we complain a lot when things don't go perfectly, and we also go looking for individuals to blame. The strong incentive is to be very risk-averse. That's especially true in organizations, like government, where stability is prized and careers are long.

I really appreciate a lot of the work that government people do, but I would never put up with half the bullshit they do.


688 Program Technician III... 88 Associate Gov. Program Analyst... 73 Office Assistant...

Numbers are FTE


Lots of snark here; very little thoughtful commentary.

If the document that steveplace posted [1] is accurate, the budget for this program includes things like the 800+ person call center to handle consumer requests, and the staff to figure out what kinds of health plans to offer in the first place. The line item for "technical" is $161M, which is only 23% of the budget.

I'm not going to suggest that's a low number for this kind of project, but then again, it's kind of hard to make it all about the horribleness of public projects when 94% of the technical budget is going to "contractual services". It looks like the software was mostly done by private contractors.

In other words: let's not allow facts and thoughtful discourse to get in the way of a good government bashing!

[1] http://www.healthexchange.ca.gov/BoardMeetings/Documents/CHB...


> In other words, it looks like the software was mostly done by private contractors.

Wait a second. Government over-spending is beyond reproach when the money is channeled to what are likely politically-connected private firms? To my mind, that is equally egregious, perhaps even moreso because of the taint of cronyism (or still worse than that, regulatory capture by entrenched large firms).

But even if I give the firm-selection process the benefit of the doubt and assume it was done without kick-backs or regulatory capture effects of any sort (seems implausible, but we'll go with that), private firms simply know that when you are bidding on government work, the budget is enormous. You're probably likely to be kicked out of the running for being un-serious if your bid is significantly less than the average bid. And the average bid is high because ... government.

I can imagine feedback of the flavor: "You think you can do this type of complex web application for a mere fifty million? Don't waste my time."


"And the average bid is high because ... government."

I'll admit that I was skeptical when people said that "everything is different this time because ... internet!", but I think your application of the theory is much more elegant.


How much did these contractors have to spend on kick backs?


It's dumbfounding this is the top comment. If we can't be critical of government waste in this case, when can we be? It seems you don't even understand your policy opponents' position; overpaying contractors is one of the chief complaints!


Well, $161m is more than 23% of $327m to start with.

The other thing is that although the government is using private contractors, that does not mean the latter are solely to blame. If the government ends up paying millions of dollars for a wordpress site sporting a $40 theme, the contractors charging for it may be immoral, but the government who awarded them the job is at best massively incompetent.


"Well, $161m is more than 23% of $327m to start with."

Read the linked document instead of just relying on Greenspun's numbers. The $327m number he cites appears to be incorrect.


That's all true and the $327M price tag is misleading, but they still spent on the order of several million on the actual buildout of the website and fucked it up. It does not inspire confidence that the rest of the money was well spent.


Yep. Because we all know that private industry never fucks anything up.


Ummm, what were you expecting? He went to the site, entered some numbers, and thereby discovered the following:

...for $327 million the government purchased a computer program unable to determine that $80,000 is more than $15,856.

Is there some sophisticated interpretation that finds this not to be a risible failure? If so, you could have provided that, as it actually would have been responsive to TFA. Don't allow what TFA actually says to get in the way of a good pro-government bashing!


Of course it was done by private contractors.

Oh yea - in India too.


It's amazing how even technical minded folks don't forgive launch day bugs and initial traffic overload.

That said, some of these bugs definitely should have been easily spotted during beta testing.

But I too would like to see an itemization of $327 million.


Initial reports were that insurance.gov got 1M uniques before 8AM on launch day. One can imagine that California's exchange site was also slammed.

I'm not going to argue that this project's budget wasn't bloated, but it's pretty naive to think that your average low-rent startup team could handle it any better. Doing tens of millions of uniques on launch day is a challenge for any team. Most commercial websites with that much traffic have technology budgets in the tens of millions of dollars, if not more.


Plus given the politics around it, we don't know what devious behavior might have been going on against it either.

Hopefully none, but you never know. Their stats may have been javascript based and not physical server hits.


Is there some balance sheet that explains how it cost $327M? Is this some massive contract awarded to some goverment guy's buddy?


It's actually the opposite. All of the regulations preventing it from going to some guy's buddy mean that only bloated, f'd up bureaucracies can clear the hurdles to land the gig. And in the end, it still helps to have some buddies, of course.


Yes. The risks of doing a government contract are enormous.

Once an old family friend with a lot of state government contracting experience brought me in to help out with some technical stuff. One of the clauses in the contract basically said that they could just decide not to pay me.

After I got done spluttering, it was explained to me that they needed the clause in case of certain weird emergencies that never happened. (I can't remember now, but I imagine government shutdown is one of those cases.) So I said, "Whatever," and signed, because I could afford to take the risk.

But there is no way I'd do that for a substantial project. The bigger the client, the bigger the crazy. To be able to take $5m gambles, I'd want to be doing $50-100m/year. Which means I'm running a bloated, f'd-up bureaucracy for sure. No thanks; if I need more crazy in my life, I'll just go read some comments on YouTube or newspaper sites.


Accenture seems to be the biggest contractor - http://blogs.law.harvard.edu/philg/2012/09/15/californias-st...


Having worked with Accenture on a government project in the past, they will milk CA dry and staff up with an army of worthless college grads at inflated bill rates and send their profits to the Cayman Islands to avoid taxes.


Storytime! Who wants to hear another story of going all-in on poor State IT management?

http://www.inthepublicinterest.org/case/virginia-statewide-i...


My small province wasted 1 million Euros in a bad looking, useless, magento site that will be close after next elections. This is how I imagine it works here in my local administration:

   -EU just sent 1 million euros for helping us out!
   -What's destined for? 
   -Technological development   
   -Ok, i'll call (my cousin) Jon, and ask him to make a website for 750.000, 
   you go and call (your brother) Mike and task him to make an iPhone app for 250.000.   
   -Good, what will be the site for?   
   -Mmmm, whatever Jon wants, does it matter?   
   -Guess not... Lets go have a drink a celebrate our good job?   
   -Deal.


Is this the correct site? https://v.calheers.ca.gov/

  Come Back Soon

  Due to the success of the Covered California Marketplace we are experiencing higher than expected users. 

  To make sure everyone has a great experience while on the site, we need you to come back in a little bit 

  We're sorry for the inconvenience and we look forward to your return in a few hours.


> we are experiencing higher than expected users.

Dammit, Californians, stop smoking all that pot so we can get to the obamacare website!


How much health insurance could have simply been bought with $327 million?


Probably not much actually. Google tells me that California's population is about 38 million people. So, on average, you get less than $10 per person as a one-time "credit" (or however you'd call this). Even if only 1% of people needed coverage, they could get $1,000 worth of coverage which doesn't take you very far.

I'm not saying that that budget is reasonable or unreasonable, but that puts the amount in perspective a little bit.


State IT project. Unbelievably overstaffed, massive budget. Mediocre results, if not outright failure. Par for the course. Nobody should be surprised.


At least CA's site is up and running. WA's site (https://www.wahealthplanfinder.org) has been down since the first. My project for down moments today is to try and determine how much it cost to develop. (Oops, I guess it's up now, just slow, and has an invalid cert.)


Apparently I need to find some government institutions for clients.


Join the rat race. Just make sure to either of the following: black, women, veteran, disabled if you want to go out on your own. If not, just sub contract from a minority owned business.


That was my conclusion too ;-) Why do we try so hard right?


Mine as well. Seems pretty lucrative.


Government business is great when you can get it. The problem is navigating the purchasing process.

In this case I like to think of it as stimulus for the economy.


Everyone's tax money goes to a select few who screwed the state over (or.. went through their loopholes successfully, however you want to look at it). None of that sounds like a good stimulus to me.


Is there anything being done here that looks like it should cost more than 500k ? ... Look over here, I've got a bridge to sell that I think you're going to love...


Sounds unreal, anybody have any details justifying it ?


They don't need to. Their customers are unable to switch.


I'd love to see a collection of how much these sites cost to build per state.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: