In this special episode of the Acquisition Talk podcast, we listen in to a great conversation on intellectual property in defense. The panel is moderated by my colleague Jim Hasik (who, by the way, has a great blog) and features Richard Gray, the DoD’s IP chief — Shay Assad, former DoD pricing chief — Kelly Kyes, Boeing IP expert — and Bill Elkington, former IP director at Collins Aerospace. Find the transcripts here.
There were tons of great insights throughout, including:
- What causes companies to “buy-in” on production
- Whether there is a level playing field for commercial and defense contractors
- The difference between OMIT and DMPD data
- How the DoD is moving to negotiate specially tailored deals early
- Whether government should get IP rights for a modified commercial product
The conversation is teed off by Jim’s third IP white paper where he looked at the effects of IP rights on the prices of military trucks using a natural experiment. The Army bought the IP rights to the FMTV cargo truck while the Marines did not buy the IP rights for their equivalent MTVR. When they both came up on follow-on procurement contracts, both services saw price increases roughly 20 percent, despite the fact the Army bought the IP rights. Jim has a good bit to say about this and more nuance in the paper.
A fascinating debate that caught my ear was the difference between IP rights related to defense and commercial contractors. Shay Assad argued that defense contractors have their independent R&D reimbursed by government, which puts them on an uneven playing field with commercial companies who funded R&D our of their gross profits. That dichotomy in risk should have consequences on pricing and on IP rights. Kelly Kyes argued that defense IR&D doesn’t have commercial applications, so needs government reimbursement. And in the case of Other Transactions contracts, the cost-sharing requirement is waived for nontraditional contractors which evens out the competition.
Level playing field
There were just so many interesting ideas that the discussion is hard to excerpt. However, in my usual style, I’ll focus on three areas of disagreement that I have. First, with Kelly Kyes. Here’s the relevant quote:
If there’s no commercial application for the technology, then if I develop that out of profit, which is what Shay was describing, then if I can’t make a commercial sale or a direct commercial sale, then there’s no way for me to recoup that investment because the cost accounting standards don’t provide for that, recoupment vehicle other than through IR&D.
… If you’re a non traditional contractor or you’re a small business, you actually don’t have to provide a cost match, for prototype OTAs. but if you are a traditional defense contractor, you do have to provide a cost match. And I think that was one of the ways that Congress leveled that playing field by saying traditional defense contractors.
Consider the program office’s perspective. They put out an OTA solicitation for a capability. A nontraditional does not need cost-sharing, so it offers a proposal for let’s say $50 million. The defense contractor can offer the same proposal at, let’s say, $25 million with a $25 million cost-sharing from it’s IR&D. That other $25 million doesn’t come from the program office’s budget. It is paid for by the DoD at-large, reimbursed through G&A rates on hundreds or thousands of contracts. So the program office prefers the traditional prime to the new entrant — it’s cheaper to the PMO but not the government. [Is there a flaw in this reasoning?]
There are a lot of aspects, like the fact a defense firm can waive cost-sharing requirements by having a nontraditional on its team. But consider another way it might work out. The nontraditional gets the OTA, and then the government competes out the requirement in a FAR-based contract to get into a program of record. Then the defense contractor enters with its government-paid IR&D, gets some proprietary information in the RFI/RFP from the OTA prototype, and because its big and has a portfolio of programs, can “buy-in” on the development to win production and sustainment profits.
Cost vs. Value
Second, I disagree with Shay Assad. He indicated at several points that the criteria for thinking about a fair price is what is the cost and a reasonable profit. Here’s Shay:
What you’re measuring is measuring a price index. The real question would be what was the profitability based on the actual cost that it took.
I actually think what matters is the price index — or the price relative to the relevant alternatives available. That shows you what matters. Who cares about “what it costs”? If it cost Toyota $100,000 to produce a Camry, that doesn’t mean it’s a fair price. So Shay’s framing only makes sense if the buyer knows that the seller is producing the right item in the most efficient way possible.
Moreover, there’s a huge problem of even knowing what the cost of something is. In modern multi-functional organizations, most of the expenses go to indirect costs like capital equipment, bid and proposal, facilities, benefits, etc. Peanut butter spreading those costs is not indicative of actual absorption rates.
One aspect of this is that for tech firms that are based in software, data, enterprise tools, product design (i.e., intangible assets), the profit margins can be huge — and that’s fair! That’s because outputs are not correlated to inputs in knowledge work. For industrial work, where known items are produced in known ways (i.e., assembly lines), then inputs correlate to outputs and profits are razor thin. For more on this economic reality, see this paper and this post.
Here’s Shay recognizing the matter:
But if I’m in a competitive environment and I’m talking about the industrial sector, I’m not talking about. Apple and Facebook and all that stuff. I’m talking about people who build things, or manufacture things. It’s a very different profit margin scenario.
As software and data “eat the world,” the world of tangible objects and manufacturing will start to look like software programs, and firms like Apple and Facebook will dominate. The DoD needs to lean into that, and not expect old industrial firms to look the same way for the next century.
The need for lifecycle IP planning
The last area of disagreement I’ll bring up is with OSD’s chief for IP, Richard Gray. His focus area seems to be aligning IP strategies between the government and industry early in the program. That production, sustainment, and other long-term considerations have to be hammered out at the start.
And we’re heading into, I think, an unprecedented area where we’re going to have, routine, more robust, more detailed discussions and engagement with industry to be able to discuss these issues early and have these hard discussions about how is your intellectual property strategy, allowing us to have appropriate competition downstream.
For a while, I’ve seen the Army and now OSD’s approach to IP as requiring more upfront planning and definition of the program lifecycle in terms of technology and costs. That aligns with traditional waterfall methods, but not agile methods that some have been advocating for, and seemed to be the intention behind the Middle-Tier and Software Acquisition pathways of the AAF. Further, I think it biases the competition toward the traditional defense primes, who have embedded relationships and can be in those early conversations. If all this needs to be hammered out before a development competition, then certain insiders can shape the RFP.
I think Mr. Gray’s approach to solving this problem is relying on modular open systems architecture. This allows the government to take a more incremental approach and use competition throughout the program, requiring them to meet the open standard without requiring buying the IP rights and data packages from the incumbent. Here’s Mr. Gray:
I’m just going to go ahead and say MOSA early on and get it out of the way. A modular open systems approaches is going to be one of the ways we are approaching this… by taking a modular approach, try to engage with our industry partners so that the government’s intellectual property strategy and the private sectors intellectual property strategy for any given program are synchronized.
I think this is only a partial solution. (1) Open standards are very difficult to generate a global consensus on, and then become rigid to change; (2) The various open systems working groups, which have industry and military co-chairs, seem to be run largely by defense primes. So they have the inside track to those open standards; (3) If the standards aren’t really interoperable with commercial standards, then you get the same ecosystem of players; (4) I’m really hoping STITCHES is able to create ad hoc interoperability rather than requiring global standards — standards emerge over time and are not laid down by fiat, and need flexibility to do particular new things.
Thanks, Jim Hasik and panel!
I’d like to thank Jim Hasik for writing a great white paper and getting together this panel for a really stimulating discussion. I’d also like to thank Jerry McGinn, Charlie Dolgas, and Alison Langford as Mason GovCon for putting the whole thing on. You can find the webinar video here. Find the transcripts here.
Leave a Reply