A Brief Word on Innovation in Federal IT
Ok, I’ll put my cards on the table: I work for a federal IT department. I am a government parasite, feeding on your tax dollars. Hold your applause, please, I’m a professional. In all seriousness, we in federal IT do very important work putting in place systems that we try very hard to maintain and make available for the public good. In the process we deal with a lot of problems, and one of the big ones is the way that “innovation” is applied to government.
Today I’m going to make a radical statement: the problem with innovation in government isn’t that government doesn’t do innovative things in IT. The problem is that government shouldn’t do very many innovative things in IT, and we keep getting forced into it by politicians who want to “run government like a company.”
Here’s the thing about innovation — it’s very, very hard to succeed at. All it really means is that you’re building new things, trying new approaches, or doing old stuff in a new way. And innovation is really important, particularly in IT where the half-life of any given system is measured in just a few years or months. But innovation is time consuming, frustrating, expensive, and prone to failure: for every successful new system or process there are a double-handful of failures lurking in the background.
My agency has some experience with this. We were early adopters of a cloud-based business process management system (no, I’m not going to tell you which one). We used it to develop a new flagship application, in the spirit of being “innovative” and “forward looking.” Fast forward six years, and we made so many mistakes and wasted millions of dollars as part of this process, much of which was directly tied to the fact that we were reliant on contractor support to develop an unproven, evolving, immature technology stack.
And that was a successful system, mind you — we just went through all that headache because we were early adopters and we had to work through the technology’s growing pains. You don’t even want to know about the failures.
Part of this is down to the difference between government and private sector IT. The private sector has to innovate to survive. That’s because the private sector needs flashy new products, needs shiny new toys, needs hyper-sophisticated data algorithms, needs new AI routines, and desperately needs new ways to stand out. That’s the business model for anyone in private sector IT whose company isn’t named “Microsoft.” If you’re not new, you’re obsolete.
Government also has to stay relevant. It’s unavoidable; we can’t be asking federal people to use Windows 95 when the rest of the world is on Windows 10. We have to keep pace.
The problem is that innovation isn’t about keeping pace. Innovation is about taking risks to build the next big thing. Innovation is about leaping before you look, and hoping that you’ll land on a profit margin. And this isn’t the government business model.
See, here’s the thing about government — it is supposed to work, for as little taxpayer money as possible. We’re supposed to be constant, always-on, prepared to slog through whatever Congress wants us to do, and we’re supposed to do that on a fixed budget that is only marginally flexible on an annual basis.
We can’t get outside capitalization to finance big new things (because it’s illegal). We can’t go bankrupt and re-roll a company if we fail (because that’s illegal too). We can’t shrug, sell the agency to Google, and start over (because that’s really illegal). We have to work. If we don’t work, sometimes people die, which is a little worse than the normal penalty in the private sector. And even when that isn’t the case, IT failures in the federal government mean that taxpayer money is wasted, and people can’t get data or services they need.
That makes government a risk-averse environment by default, and it means that government agencies are not the appropriate testing grounds for innovative IT solutions unless, of course, you work for DARPA or a comparable research-focused agency. The rest of us don’t need the distraction.
Does that mean that government IT shouldn’t change? Of course not. We have to change, have to adapt to the times. That’s unavoidable. But we desperately, desperately need to stop getting handed “experts” from the private sector, who think they immediately need to “innovate” government IT, as our politically-appointed bosses. These poor idiots don’t know that they’re in a different game until half their tenure is done (because federal political appointees have a shelf life of about four years), and then they have a tendency to throw up their hands and just let things stagnate.
I’ll say this, after ten years of federal service I’ve seen four Chief Information Officers come and go, and not a single one of them did anything to actually make my agency better. Every. single. one. started up a new “innovative” solution or business model, watched it fail (because of the huge number of ways in which the federal government is not like private industry, but I digress), and then moved on to greener pastures.
What I dearly wish our dear leaders would do is wait. Wait for the private sector to develop mature process, mature systems, mature technologies, and then pick the best ones for federal use. Let the private sector take the risk; that’s what they’re for. And then move federal IT to a new solution once it’s proven to work.
If you think that’s a recipe for stagnation, my counterargument is as follows: two years. That’s all I’m asking for. After a couple of years, it’s not that hard to tell whether tech will work or not. It’s either already been broken in by the private sector, or it’s on the way out. Two years isn’t all that long to wait, and doesn’t put us that far behind the curve.
Doing otherwise, innovating for the sake of innovation, is a waste of your tax dollars, a waste of my time, and a waste of political capital within federal agencies. We could be doing so much more work that isn’t sexy, isn’t flashy, isn’t really that exciting, but which would actually benefit the nation. Instead we’re tasked to stand up a new IT solution because my new political boss is too up his own ass to realize that we don’t need to innovate a fix for something that isn’t broken.
This has been a brief rant on a waste of your tax dollars in the name of private-sector ideology, you’re welcome.
