One govtech innovator argues that the fixes could create more problems than they resolve.
Note: This blog was written by David Eaves, Lecturer in Public Policy at the Harvard Kennedy School and Director, Digital@HKS.
During the launch of Healthcare.gov, media coverage centered around poor procurement as a contributing cause of the failure. It's a widely held sentiment. If you hang around the govtech sector -- people interested in starting tech companies that serve government, my friends over at Code for America and industry commentators -- procurement is often cited as the big bad bogeyman holding us back. If only we fixed procurement, there would be more choice, better vendors and less failure among government IT projects.
I don't buy it.
We can get better choice, better vendors and less failure without reform. I'm not saying procurement is perfect. But the benefits of reforming procurement are not needed and the dangers are significant.
Over the past semester, Steve Kelman, former White House Deputy CTO Ryan Panchadsaram, and I ran a course at the Harvard Kennedy School on why Healthcare.gov failed. The goal wasn't to find the line(s) of code that caused the website not to function, but rather to look at the lifecycle of the policy -- how the legislation was written, how the vendor was chosen, the project management, etc. -- to learn what conditions led to the failed launch.
In the end, the "rules" of procurement did not create the conditions of failure. Lots of choices did -- but none of them were required by procurement law.
So why do we think procurement is broken? People generally cite two problems: it's too complicated and it's too slow. Both of these problems allegedly drive away new players who can't navigate the system, or can't afford to wait around to win the work. This creates an oligarchy of providers who are slow to innovate and deliver value effectively.
On the one hand, I agree with this. My observations have been that learning to navigate procurement is one of the key elements that differentiates successful from unsuccessful companies in the government space. Sadly, it can sometimes become the core skill of success, more relevant than even the ability to deliver an effective product or service. That may be one reason you see so few recognizable names in the top 100 IT vendors to the federal government. Some of these players are niche players whose competitive edge is in navigating government purchasing processes. I suspect such skills are in part what prevents the list of players from changing.
But much of this is true of any enterprise-oriented company, government-focused or otherwise. Knowing how to sell into market is essential, and changing procurement won't dramatically alter that fact.
I often hear people say that if procurement were easier or faster, then their cousin’s/friend’s/son's/neighbor's cool startup would serve government. Maybe this is true, but I've also seen how many terrible IT vendors are out there. I'm not sure that making it easier to buy technology will ensure more good technology is bought. It may simply ensure more technology is bought -- and that could be a disaster.
What I've observed is that the number of people who can identify good technology and practices from bad ones is limited. This is actually a very difficult problem even for "experts" in the private enterprise sector; assuming it exists among public servants who rarely buy technology is asking a lot.
If I could wave a magic wand to improve govtech, I'd want better-educated and more knowledgeable buyers. No amount of procurement reform is going to get us that. (It's also a reason I've been supportive of 18F and the U.S. Digital Service -- they at least regularly pump fresh consumer-facing IT talent into the system that can identify good and bad technology choices.)
Moreover, the current procurement rules allow for a fair degree of flexibility. In our class we borrowed language that I'd learned from Jen Pahlka and Erie Meyer concerning red rules and blue rules. Red rules signify rules that you cannot break, as they are in the law (e.g. a procurement threshold amount). Blue rules signify the “lore” or wisdom accumulated over years (e.g. agile methodologies are not permitted).
What we observed is people often confuse red rules (which cannot be easily changed) with blue rules (which are, at best, guidelines). The reality is procurement law allows for a fair degree of flexibility. Procurement lore does not.
An as example of this flexibility, look no further than the TechFar handbook, created by the USDS. This guide to procuring IT distinguishes blue from red rules to give people more flexibility in procurement. The genius of the TechFar handbook is less that it is a handbook than that it is a culture-hacking tool. It's designed to deprogram perceived blue rules and replace them with newer, better ones.
And there have been wins in the procurement space from which lessons could be drawn and scaled. Companies like Nava and Ad Hoc show that small, innovative companies can serve government. I also recently wrote an Harvard Kennedy School case in which California altered an RFP to allow for an agile approach to procurement where multiple vendors would compete or different project components with more successful vendors earning more and more work. A lot of procurement lore was debunked. But no procurement laws were changed.
This culture change that USDS engaged in is not easy, but changing lore is probably easier and more effective than changing law.
And that brings me to the final caution around procurement reform. My friends in the tech sector assume that any effort to reform procurement will result in a net positive outcome.
I think this is fantasy.
Once you open a law up for reform, what happens next cannot be predicted. Is this the administration you want to lead a procurement reform effort? Maybe. Maybe not. But no matter who leads it, you won't be able to predict that outcome.
Who knows what influencers will emerge and what other ideas (good, but more realistically bad) will want to be included in any changes. However confident they feel, I have my doubts that the civic tech sector and agile advocates such as myself will have more clout and influence than, say, the established 100 top vendors who billed a collective $88 billion dollars in 2016. I'm sure they'll have lots of suggestions too.
Focusing on culture and delivering low-cost results within the status quo is maybe a less dramatic path to success, but it is a far better, and safer way to reform the system.
Opening up procurement reform? There be Dragons.
NEXT STORY: How State wants to innovate