Fixing Government IT Takes More Than Four Years
Published on:
Anyone who wants to make real, lasting impact to federal IT needs to understand something really important: it takes more than four years.
That's because the problems with federal IT are NOT technological. This is not a case of "the computers are old, let's replace them with new computers!" The roots of this problem are way further upstream than most people think. Those roots are:
HIRING. We don't pay well enough to attract and retain talented engineers. It doesn't matter if we outsource all of our IT. Without in-house talent to manage the vendors, the product of our outsourcing will always be poor. There's a number of tools to improve hiring, but the main one is legislation that allows agencies to pay people higher salaries so that the government can compete with private industry. That takes years. So if you assume a year or two for legislation, agencies will onboard people a year or two later (yep! it can take that long to hire people) at these higher salaries, and then several months before they are oriented enough with their enterprise to make a meaningful impact, we're looking at several years before the public notices a shift in the government's ability to deliver IT.
VENDOR CAPTURE. Let's say we just went through the above process and an agency now has the talent to build great IT, and/or manage vendors to do the same. Great. Unfortunately, many of the systems that drive federal IT services are captured by firms that do not want to let go of them. These firms were hired to run managed services, soup to nuts. Federal staff may not even have permissions to touch the code. So before an agency's great talent can get to work, they need a new contract that extracts the system from vendor capture and empowers the government to build a better new one. That takes a while! Years! (I'm radically underselling the extent of this problem and the pain of reversing it, but hey, it's social media.)
Both of these things take concerted, sustained effort across hundreds of agencies for several years. And that is simply too much effort and too long a timeframe for political organizations that a) want fast results, b) have lots of other priorities and distractions, and c) go through seismic shifts in strategic direction every few years.
And I haven't even gotten to the cultural, procedural, and regulatory issues that make government move slower than technology evolves.
Doing what it takes means driving forward for years without any noticeable impact, confident that you are laying the groundwork for something valuable...something that your successor might get all the credit for. So the question for an incoming administration becomes: Do you REALLY want to make real, lasting impact to federal IT? Or do you just want to create the impression that you have? The latter is easy. The former requires immense courage.
Originally posted on LinkedIn @govtech