A recurring theme in innovation-policy debates is whether the government should "pick winners and losers" or not. On one side of the debate, folks will argue that market forces are better at encouraging innovation. On the other side of the debate, folks will argue that the government has a pretty good track record of encouraging innovation.
Fortunately, in government contracting we can largely ignore that debate because, government contracting necessarily requires picking winners and losers!
But, hang on. What if government contracting itself leads to innovation? Has increased spending in government contracting led to greater innovation (as measured by patent counts and citations). That would be neat, right?
That's the topic of a really interesting paper entitled "The Missing Link(s): Government Contracting and Patenting". According to the author, increased spending in government contracting has, in fact, led to more innovation.
What's wild, though, is that author was able to test why government contracting leads to innovation. He draws the conclusion that "the competition between corporations interested in being awarded a federal contract are responsible for the increase in innovation following an expansion of federal spending".
Which, I guess makes intuitive sense? Increased competition, more innovation. Everyone can rejoice on having a broad consensus policy goal of increasing competition in government contracting. Yay.
If you step back, though, it's also sort of darkly amusing. You see, to reach that conclusion, the author needed to reject the hypothesis that government contracting leads to innovation after contract award. In other words, at a macro level, innovation is a distinctly pre-award phenomenon. Ouch.
The author also had to reject a hypothesis that, for me, caused an instinctive urge to reach for a bottle of whiskey before breakfast:
[T]here is a lot of evidence that innovation pertaining to a government contract is slow. As the federal agencies generally value an unchanging product, there is much less incentive for a contractor to innovate after having achieved a long-term procurement relationship with the government (slow innovation hypothesis). In conversations with employees at large contractors this became very apparent. According to one of them, even changing as little as the paint of an existing product resulted in a months-long approval process which ultimately ended in the denial of the request.
Too real? Oh well. I guess we should celebrate that that the "slow innovation" hypothesis was rejected? Sure, that's why I grabbed for the whiskey...
Anyway, it's an interesting paper and the bottom line resonates. Government contracting lead to more innovation. And more competition in government contracting leads to more innovation. But more must be said.
It also prompts some interesting, unresolved questions when you go down a level. The author, for example, raises concerns about the impact of government contracts on consolidation. Do contract awards lead to industry consolidation, leading to less competition, leading to less innovation? We don't know!
And what exactly is it about competition for government contracts that leads to more innovation, anyway? It is because the government is demanding more innovation with all of the new money? It is because the industrial base is rising to the occasion and offering better solutions? Are there spillover effects from procurement of commercial items? I have hunches, but that's what research is for!
The optimist in me wants to feel good about this paper! It really does. In government contracting, we can inspire innovation without worrying about the awkward politics of picking winners and losers... But on the other hand: maybe the conclusion is really just that we need to make sure there are enough losers out there?
See you at the bar?
 I'll leave it to the reader to evaluate the methods and hypotheses, but the author is really clever in using one of my favorite data sources (bid protests!) to deal with the fact that even though the government picks winners and losers, knowing who the losers are is not really public information.
 To be fair, I personally have many questions about the degree to which variables like data rights or contract type might influence these outcomes and this was not really addressed in the paper.