Thursday, July 16, 2009

Conferees: Please Don't Harm the SBIR Program

If I had my druthers, the soon to be convened SBIR Reauthorization Conference Committee would be required to swear a Hippocratic style oath of "Primum non nocere" - First, do no harm.

There is potential that harm will be done. There's also potential for significant improvements. The problem is that it's not clear which provisions of the two SBIR Reauthorization bills on the table (H.R.2965 and S.1233) improve and which harm. It all depends on your perspective. Both sides of every issue think they're promoting improvement.

So what should we do? We're encouraged to write to our legislators and offer irrefutable logic to support a position. Does that work? Do they actually vote based upon our logical presentation? Evidently not very often.

We can argue details and numbers and statistics ad nauseum. Actually we have. I'm tired of that. Aren't you?

So let's take a different view. A qualitative view. My view of Primum non nocere. Without any numbers or statistics.


SBIR is a technology development stimulator. It's not a jobs program, although it does result in jobs being created. The competition for project selection is based upon whether the innovation furthers the mission of the sponsoring Agency, whether advancement of the state of the art is likely, and whether the company proposing the work is capable of executing the plan and eventually making money from it. No other criteria should be applied to project selection. Applying criteria that are unrelated to the innovation defeats the purpose of the program and will do harm.


SBIR provides the seed funding for jump-starting an innovation's development. The Government is the investor. In Phase I the innovation's feasibility is vetted. If the proof-of-concept after Phase II proves out, then it's up to the company to find the commercialization money to carry the innovative technology further. It may take additional investment. If the Government is eventually to become the customer for this, then it may be in the Government's interest to provide additional investment -- but not from the seed-funding pot. Keep SBIR money strictly for feasibility vetting (Phase I) and one proof-of-concept project (Phase II). Any other use of these funds harms the program by reducing the number of jump-start projects that can be funded.


Eligible companies for SBIR grants should be those who are unlikely to find seed funding anywhere else to demonstrate that the proposed technology innovation will actually work. Providing grant money to companies who have other sources for funding harms the program by reducing the funds available to those who need it.


The number of projects funded should be an appropriate number so that we're not leaving worthy projects behind, but also not providing funds to companies who really have no chance or intent to move to commercialization. This is a balancing act of making the allocation base (the set-aside percentage) and the funding caps (project maximums) appropriate to permit the funding of as many truly worthy projects as possible. Any provision that arbitrarily upsets this funding balance does harm to the program.


There needs to be a way of evaluating the return on the SBIR investment. There are three interested parties here -- each with their own perspective on ROI: (a) the Government agency making the investment, (b) the company receiving the grant, and (3) the taxpayer. If we can't measure ROI there's no way of evaluating whether anything we do improves or harms the program.

Just about every SBIR issue on the Conference Committee table can be related to one of these five categories.

So, Conferees, as you look at the details, please relate them to one of the above quality considerations and ask:

  • Do we preserve the purpose and intent of SBIR as Roland Tibbetts and others envisioned it?
  • Are we providing seed funding appropriately to worthy recipients, in the proper numbers, at the proper levels?
  • Can we calculate the return on this investment, so that when we go through this process again we have some real data to evaluate?
  • Does this provision do harm to the quality of SBIR?

OK, let's get to work. Please raise your right hand, and say after me: "Primum non nocere ut SBIR".

Hippocrates would be so proud!

1 comment:

Carl Nelson said...

Some good questions and some worthy ideals, Fred. But a few quibbles:

1. No jobs are created. Every SBIR dollar is taken from some other contract that would have been awarded for about the same number of jobs.

2. Sure, ROI should be measured. We already have 25 years of data. When do we start and how do we do it? No more lemonade reports on "best practices".

3. If it is such a great idea, why wouldn't the agencies do it without a mandate in a way that best fits the agency? Go ahead, put the grand theories of value to the test, make it optional!

4. Roland's idea was fine for NSF and NIH that didn't do companies. But the mission agencies knew all about companies, large and small, and their ideas. DOD had standing Broad Agency Announcements that invited innovation, although with no element of commercialization (which DOD pretty much ignores anyway).

5. Roland's fine sounding theory that government should have the same ratio of large to small company contracts as private business was nothing but a fine sounding theory. And after 25 years there is still no proof that he was right. If you can't prove your theory with 25 years of data, you should get yourself a new theory.

5. "truly worthy project" is in the eye of the beholder.

6. Extend SBIR for another year and pay Josh Lerner a pile to invent a worthwhile program.

BTW: Hippocrates was a Greek.