Set-asides for innovation in government IT

A lot has been said and written lately about the need to inject more innovation into US federal government IT, and government IT in general. We talk about the urgent need to innovate in Gov2.0 and Open Government. As a taxpayer and someone who has worked in this field for quite a few years now, I couldn’t agree more. But some of what I read is a bit cynical, painting the government as something so different from the private sector that it just cannot get out of its own way to innovate. There is a sense in some circles that innovation is simply anathema to the entrenched bureaucracy or that major restructuring is needed.

I think, in reality, the barriers to innovation in large private sector companies are really not all that different from what we see in the government. If you worked for a large hotdog bun maker and your big idea is to give away the bun recipe and rent the hotdogs instead, you are going to face an uphill battle at MegaBuns, Inc. And that battle will be just as hard as you would face at MegaGovContractor, Inc. who has a strong business relationship with MegaLegacySystems, LLC.

The difference, though, is that if you are sufficiently fanatical about this great idea, you can leave MegaBuns and start your own company. That works because your disruptive business model can find investors who are willing to take that bet. But in the government, if you have a brilliant idea to automate the Cans for Beans program, you can’t just leave and start your own Department of Agriculture. Unfortunately, it also turns out, you usually can’t even get a parallel contract for the program either.

One way to address this issue is to evangelize vendor independence, open standards, SaaS, open source and other great ideas from the top down in government. Those efforts are to be lauded, but I wonder if there aren’t also some more structural ideas that can accelerate innovation in general.

The issue is that right now we, as taxpayers, if we need a program to track Cans for Beans, will invest $X million in a single program to build that system. We have decades of great work in the FAR to make sure the program is awarded competively and fairly, but at the end of the day we invest in one team, one idea, one architecture, and one outcome, and then hope that works out well for us. The Department of Defense has a better way of approaching huge weapons programs, where they will at least fund a substantial “bake-off” and pick between two prototypes for the next fighter jet or armored vehicle. We could adapt that idea to government IT, but that could be a big challenge for program directors and acquisitions staff that would then have to manage twice as many contractors. This may or may not scale well to smaller programs.

Innovation Set-Aside

But what if the government took a smaller step first? What if for every new IT program of size greater than $X, (X doesn’t have to be very big), we tell the chosen contractor that they must set aside some amount $Y to fund a prototype “Plan B” architecture for the program using different technologies or models than “Plan A”. So the program and acquisition staff will still choose one contractor and one technical approach like they do today, but then the government invests in a built-in hedge on a consistent basis.

This doesn’t need to be as complicated as putting two hyper-competitive contractor firms in the same pen and expecting them to play nicely together. And it doesn’t have to put the government in the position of having to manage two programs. (Imagine a VC investing in two directly competitive companies, knowing one will kill the other, and then trying to sit on the two boards and help where he can.)

Instead, the winning contractor will invest the $Y internally (this can involve subcontractors, or not) and if the “Plan A” fails or the “Plan B” is just a rocking success as a prototype, then the government can switch to “Plan B” with no harm done to the prime contractor (and therefore no incentive to suppress it). “Plan B” becomes the new “Plan A” and gets built out to scale. No contract cancellations, no scathing GAO report, not even a re-compete required.

We have a lot of experience in using financial set-asides (incentives for women-owned business, minority-owned businesses, firms started by service-disabled veterans, etc) to effect social good and encourage more open competition between companies. Can we apply a more limited version of this to encourage competition between ideas as well?

UPDATE: Bill Brantley also wrote about some other barriers to innovation in government IT

UPDATE: Some thoughts on why innovation isn’t rational for a lot of government programs today.