I've stolen the title of this post from a very funny line delivered by Daniel Nocera (MIT, SunCatalytix) at the ARPA-E Summit in DC this morning.

Yes, I'm attending the first ARPA-E Summit, and I'm glad I did.  It's proving to be one of the best events of the year, in the cleantech sector.  A who's who list on stage and amongst attendees, a great mixing bowl for researchers and practitioners and investors, and some really impressively innovative ideas.  Kudos to the organizers at the DOE and CTSI, among others.

The idea of ARPA-E, of course, is to try to get the energy landscape shift Nocera was referring to, to happen more quickly.  To fund breakthrough, but practical innovation, that otherwise wouldn't get funded.  It hits directly at one of the capital gaps I've described before, at the very early stage, where VCs and angels and other private sector funders aren't fully able to get involved, for reasons I laid out a while back.  So it's an important effort.  In many ways, ARPA-E is just following in the footsteps of DARPA, another successful government program, and one that typically gets very strong support amongst legislators.

But while you can sense the optimism in the crowd here at the summit about this program and what it's doing, ARPA-E is going to face quite a few challenges going forward.

1. Demonstrating economic impact:  The work being sponsored by ARPA-E is necessarily forward-looking.  It's about investing in technology and commercialization efforts that are too far out or risky for the private sector to fund.  In theory.  But in this economic climate, the pressure will be on all programs to demonstrate economic impact, specifically jobs.  These technology development efforts don't lead to too many jobs being directly created.  Perhaps the funding helps hire another engineer or two at each company, but realistically, it's a bank shot to get to jobs growth -- we innovate and commercialize something that then eventually creates jobs once it starts being rolled out.  But here's the crux of the challenge -- many of the manufacturing jobs that get created via this innovation, when it happens, will be overseas.  This is just the way the world works, I'm not criticizing ARPA-E for this.  But my guess is others will, at some point -- and specifically when an ARPA-E recipient outsources manufacturing and gets visibly "outed" for it.


2.  Demonstrating additionality:  "Additionality" means a very specific thing in the carbon world.  It means proving that carbon emissions reductions wouldn't have already happened under the status quo.  ARPA-E is going to face a similar challenge.  Among the ARPA-E grant recipients are several startups that were already pretty well-funded by VCs.  It raises a critical question:  Is the role of ARPA-E to fund projects that wouldn't have been funded otherwise?  Or is it simply to accelerate development of ideas, regardless of how or if they've been funded to date?

If ARPA-E is only supposed to fill the capital gap, then there's a problem when it provides additional capital to a high-profile venture-backed startup.  The private sector had already demonstrated its willingness to put money into the development effort, so what's additive about ARPA-E's role, and why wouldn't they have better-deployed the money into something else with breakthrough impact potential but no venture funding to date?

On the other hand, if ARPA-E is only to fund efforts that haven't been able to get venture funding yet, doesn't that create a selection bias issue, where the agency is only funding ideas that the private sector has rejected, perhaps at times for good reason?  And why shouldn't the agency support really important ideas regardless of their funding status, because additional capital and visibility still helps?

Again, my point isn't to criticize ARPA-E, which I believe has done a great job to date.  But I do think the conundrum posed by these questions will be a challenge that the agency will continue to wrestle with going forward, and will likely face some scrutiny over at some point.


3.  Demonstrating good selection judgment:  This dovetails with the second point from above.  ARPA-E has been flooded with requests and applications for funding.  It will continue to be so, and many worthwhile efforts therefore won't get selected, it's impossible to back every deserving project.  And so as the agency staffers choose one recipient over another, it raises the likelihood of the agency getting criticized for their choices -- either as being wrong, or as showing favoritism.  The agency, as far as I've seen, has worked hard to help avoid this, bringing in a significant number of outside reviewers as part of the process, and now also reportedly providing a lot more transparency to the process.  This is very good.

But still, sooner or later one of these efforts will utterly and visibly fail.  VCs are used to this, but elected officials are not.  So sooner or later, competency of selection will be a debate that arises, whether justified or not.

And, sooner or later, someone is going to ask why a wealthy VCs' portfolio company just got more "free money" from the government.  This goes to that second point above, and has already been the subject of some external criticism.

Finally, sooner or later there will be an example of where the visibility inferred upon an ARPA-E grant recipient is perceived as having disadvantaged some other startups in the same subsector.  Government intervention affecting the competitive landscape, in other words.  There's a legitimate counter-argument to be made that any legitimacy for one player in a subsector generally helps all players in the subsector, but still, I know from talking with contacts at the DOE that they're already being bombarded with complaints from elected officials whose constituents were denied in their grant applications, etc.


I'm a big fan of the ARPA-E effort.  And, knowing some of the people involved, and knowing some of the major process overhauls they've been going through to try to get it right, I think we'll look back on the early days of ARPA-E as a really standout effort in terms of the level of execution of this versus other government programs in other policy areas.  I just also know that nothing is ever executed flawlessly, and thus ARPA-E will likely eventually face some of the criticisms I've described above. 

I hope that when that time comes, members of the cleantech research and investment communities will stand up and support it as the very valuable program it is.