Editorials

This morning my esteemed Berkeley neighbor Brad Delong drew my attention to a Wall Street Journal article headlined: Health Website Woes Widen as Insurers Get Wrong Data: New Errors Indicate Technological Problems Extend Issues Already Identified. Because of my reluctance to swell Rupert Murdoch’s coffers, I haven’t bought my way through the WSJ pay wall, and therefore didn’t read any more of the article than the lead sentence Brad quoted on his blog, but the headline was enough to set me off.

The deafening chorus of OMGs being generated by the newsies over problems with the first generation of enrollment software for the Affordable Care Act (aka Obamacare) makes me wonder where they’ve been all these years. Sometimes I think business writers in particular should be required to take a real-world internship before they’re allowed to opine on things they just don’t understand. 

(Newsies is not a term of respect. It’s a somewhat contemptuous term of art which is a relic of my pre-journalism career as a political organizer. Not all newsies, however, are bad.) 

I myself benefitted from a lengthy sabbatical between my journalism bookends, when I served as general manager/marketer/house counsel/salesperson/dishwasher for our mom-and-pop software development company. In the course of this drastically understaffed venture (which was an overnight success after only 18 years of hard work) I learned a few useful maxims about technology development. 

First (swiped from IBM): “A demonstration is an opportunity to fail.” No matter how well you might think your stuff is working, something will go wrong when you try to show it off. Undoubtedly the software used to enroll customers in Obamacare worked well enough in the lab, but in the field lots can go wrong, and usually does. 

Second (my own observation)"Anyone who believes a programmer deserves what s/he gets.” Software developers (programmers, engineers, whatever) are working from ideal designs which exist largely in their heads. They sincerely believe that the code they’re writing will match their expectations, and usually, eventually, it will, but the timing can be off by a factor of 5. And sometimes it never works. That’s just reality—get over it. 

Third: “Capitalism is not your friend.” We were so cynical that we never did manage to connect with the venture capital suitors who courted our little company—the result was that when we finally sold it we retained a respectable profit since we had no greedy investors waiting for their cut. Friends in a similar mom-and-pop business who had big-league partners were not so lucky. 

Those of us who would have preferred a government-run single payer system to the Romneycare clone that became Obamacare are not surprised that the enrollment software which private companies pre-sold to the federal government didn’t work when delivered. 

As the French would say “ C’est normal.” That’s to be expected. 

A couple of case studies from experience: when we first started our company, long about 1980, we bought database software for resale directly from Larry Ellison himself. Didn’t work as delivered—never did work. So? 

Fast forward to 2013. A professor friend who works at San Jose State University reports that software from Ellison’s Oracle Peoplesoft company, which the university now requires her to use, takes 3 or 4 times as long to do simple tasks as the pre-Oracle manual system that she formerly used. And some jobs—room scheduling, for example—she can’t do at all on the computer anymore. She tears her hair every term. 

But Oracle has made Larry a ton of money, hasn’t it? And profit is king in the market-based universe. 

Too much testing cuts into profits. A not-for-profit non-competitive single payer system would work much better, no dispute there. But that’s not the American Way. 

A few of the newsies seem to get it. Or rather, a few publications have allowed some actual information to creep past the uniform chorus of booing. For example, in an op-ed in Bloomberg Businessweek, Paul Ford makes an intelligent case for open source software instead of the ultra-proprietary stuff which was used for the ACA rollout. But then, he’s a techie, not a newsie. A programmer, in fact—which is why, following Rule 2, I don’t believe everything he claims. 

Which brings me to Rule Number 4, one which has acquired a certain amount of currency in the wider world. I learned it in the high tech context as “The best is the enemy of the good.” 

What that meant in our business was that wearing my sales hat I sometimes had to wrest good if not-quite-perfected technology away from the programmers so that we’d have something—anything—to deliver in order to meet payroll. Happens everywhere, all the time. 

That’s why I have more than a little sympathy for President Obama’s current effort to deliver affordable health care to an ungrateful public. On a much grander scale, he too is a CEO/marketer/house counsel/salesperson. (Let’s hope someone else washes the dishes.) And yes, his product doesn’t work perfectly yet, but it’s a lot better than nothing. It’s time to cut him some slack. 

P.S. The reason that this piece was finished hours later than my (self-imposed) deadline is that my Comcast internet service took a five-hour unexplained vacation in the middle of the day. I recently shifted to Comcast because AT&T’s service, both phone and internet, had started failing frequently. 

Oh, and I have a lot of trouble buying theater tickets from some websites, and I think most people do. And they charge me two bucks for the privilege of wasting 10 minutes online with their dysfunctional front end. 

I’ve been using computers since 1969 and I passed the Bar exam on the first try, but I struggled for an hour trying to get a taxpayer I.D. number from an IRS website. When I finally gave up I was advised by my lawyer’s paralegal that the only way to do it successfully is to wait on hold to talk to a human. That took me an hour and a half, but I finally managed to get the number. 

Healthcare.gov may not work quite right yet, but that’s just “industry standard”, isn’t it? There’s a lot of non-functional technology out there, and we’re dependent on it. 

0