October 14th, 2013

A Few Remarks About Shutdown/Debt Limit Coverage

While T. and I were at McDonald's yesterday, I heard a familiar voice and looked up to the TV (I know -- the Maynard McD's has free wifi and a big TV. Wait for it . . .) to notice that Rogoff was on CNN, arguing that reducing the debt too quickly -- austerity -- was not a great idea. Obvs, this is a tough idea for the anchor on CNN and/or whoever is generating her talking points, so she went on for a while about, how can this be true? We've never even _tried_ austerity. Also, belt tightening.

On the one hand, if you're an anchor on TV, you are familiar with going hungry, so I guess she may know a lot about belt tightening. But starving yourself so you meet ridiculous appearance standards for a TV gig is very different from going to school hungry because your family doesn't consistently have enough money for (or isn't there -- because they are already at work -- to make sure you eat) breakfast. So: annoying anchor, yes. And Rogoff is right, but _wow_, this is the guy who put together a dodgy spreadsheet to be published in the non-peer reviewed issue of a journal arguing that OMG WE MUST KEEP DEBT UNDER ARBITRARY X% OR OH NOES, thus leading to our recent experiments with austerity, preventing the recovery from making better progress and leading directly to where we are right now.

Nice that he finally got religion, but really, a day late and a dollar short.

Today's entry in, _this_ is how you're going to fix the trouble you got us into:


An "anti-debt" foundation has put out something about how expensive the shutdown mess is, and how that would be a _really bad idea_ if engaged in repetitively. Because, financially irresponsible. _Duh_. But again, it was all that anti-debt rhetoric which made it so hard to do decent policy post-bust. A day late and a dollar short to say, wow, maybe took it a little too far.

On the one hand, I applaud these efforts to unravel the knot they have created.

On the other hand, I would prefer to just tell them all to go f*ck themselves.

Define "failure"

I got a degree in Computer Science. It was an undergrad degree, not an advanced degree, but it was from a Pretty Good School (University of Washington). I had decent grades. I worked in the field for a few years and then retired, so my actual, personal experience is a little limited, but I read. A lot. I've been reading about how various large software projects went since the mid 1980s, when I was reading my sister's CompSci textbooks (she got the same undergrad degree I did, altho 7 years earlier, and then she got a Master's at MIT, overlapping there with my now-husband, while he was an undergrad. As near as we can tell, they never met). So I have a Pretty Good Idea about what Failure in a big software project looks like. I've participated in some.

The rollout of the exchanges for the Affordable Care Act is _NOT_ a failure. It happened on time. People have been able to get the software to eventually produce what it is supposed to produce (let them create an account, research, and even sign up for a policy -- not clear on whether payment is working yet or not). Calling this a "Failure" is Being a Jackass.

All right. I'm done with being an Old Fogey for now. Kids these days. Don't know what a Failure really looks like.

Here's Ezra Klein being an absolute idiot on the subject. He's too young to have any clue, plus no real experience in the industry:


He quotes this guy:


Me, I'm impressed anything is getting through. I know perfectly well what happens when transactions are large and unwieldy and difficult to encapsulate (which is definitely the case here). I could get into ridiculous detail about stuff I've seen go wrong -- just as a customer -- with Disney's ticketing and reservations systems (both at the parks and at lodging). If they cannot fix these problems, there is trouble. But having this kind of trouble at the start?


ETA: Wow. I thought the above coverage was bad. This is _pathetic_.


"One highly unusual decision, reached early in the project, proved critical: the Medicare and Medicaid agency assumed the role of project quarterback, responsible for making sure each separately designed database and piece of software worked with the others, instead of assigning that task to a lead contractor."

For the win! I am not being sarcastic. _This is why it actually rolled out_, like, ever, rather than dying a horrible death years overdue and a billion dollars over budget. You _cannot_ outsource co-ordination. And everyone who tries to do so discovers why that doesn't work the hard way. So writing an NYT piece about why this was the source of some relatively unimportant problems is ... well, I guess that's just what we have to expect from people who have avoided technology as much as humanly possible for as long as humanly possible.

I'm increasingly suspicious that the carriers who are predominantly interacting with the federal system are carriers who themselves have created minimal customer-facing technology. They have no experience with it, so their first experience strikes them as horrifyingly bad. Once you've been a few rounds on this stuff, you sort of get used to it, and expect the Badness.

There's some sort of tech virgin joke here. I'm going to not actually make it, because whatever you come up with will be funnier than anything I'm coming up with.