How might agency problems limit innovation?

Holmstrom’s work has provided me with a great deal of understanding of why innovation management looks the way it does. For instance, why would a risk neutral firm not work enough on high-variance moonshot-type R&D projects, a question Holmstrom asks in his 1989 JEBO Agency Costs and Innovation? Four reasons. First, in Holmstrom and Milgrom’s 1987 linear contracts paper, optimal risk sharing leads to more distortion by agents the riskier the project being incentivized, so firms may choose lower expected value projects even if they themselves are risk neutral. Second, firms build reputation in capital markets just as workers do with career concerns, and high variance output projects are more costly in terms of the future value of that reputation when the interest rate on capital is lower (e.g., when firms are large and old). Third, when R&D workers can potentially pursue many different projects, multitasking suggests that workers should be given small and very specific tasks so as to lessen the potential for bonus payments to shift worker effort across projects. Smaller firms with fewer resources may naturally have limits on the types of research a worker could pursue, which surprisingly makes it easier to provide strong incentives for research effort on the remaining possible projects. Fourth, multitasking suggests agent’s tasks should be limited, and that high variance tasks should be assigned to the same agent, which provides a role for decentralizing research into large firms providing incremental, safe research, and small firms performing high-variance research. That many aspects of firm organization depend on the swirl of conflicting incentives the firm and the market provide is a topic Holmstrom has also discussed at length, especially in his beautiful paper “The Firm as an Incentive System”; I shall reserve discussion of that paper for a subsequent post on Oliver Hart.

That is A Fine Theorem on the work of Bengt Holmstrom, do read the whole thing.


And these firms obviously need protection from competition through regulation and the odd tossing in of a few trillion dollars to bail them out from their stupidity.

In other words, left to their own devices they would die a natural death.

CNTRL + F + "patents" and not a single f-ing mention of patents. Typical of eggheads in their ivory tower.

The reason no firm does 'moonshot' R&D is that it does not pay to be first: you'll be ripped off due to weak IP (patent) protection. This is well known in the invention industry. GE ripped of Damadian on MRI (and actually perfected it better) with the attitude "just sue us, we'll fight about the damages later in court, and we'll both make money" (which is exactly what happened). Being a pioneer does not pay. The term of protection is too short for pioneer inventions (20 years, used to be 18). The only exception was Patlex, Gordon Gould's firm to capitalize on Gould's laser patent. The only reason Patlex made money is that the US Patent Office mistakenly delayed his patent for about 25 years, and by that time the market for lasers became mature. Further evidence that the term of protection for true pioneer inventions is too small (folks, we're NOT talking about troll patents here, please).

Do I win a Nobel Prize now? It's so obvious to anybody who's worked with inventors like I have...but I guess not to Nobelists.

Back in the 70's there was a popular MBA case study. Harshaw Chemical developed a way to grow crystals (for some reason growing doped silicon crystals looked to be a promising growth area). One of their employees went into business for himself. He made boat-loads while the patent case went through the legal process, and was allowed, after being found guilty of infringement, to sell through the warehouse full of crystal he had produced. (Back then the global market was very much less important). As an ex-chemical R&D worker, I mostly agree that it is very difficult to protect yourself when you have a really innovative patent - the space of "similar, but not quite the same" rip-offs is huge. But it isn't only patent law that doesn't provide sufficient protection & enforcement (patent law has changed, some for the better, since then). There are two other facts that must be brought in: innovation is enormously disruptive to an organization's status quo. Most organizations do not restructure their operations (and thinking) to facilitate the implementation of innovation. [Let me tell you about a conversation I had with two IBM salesmen (was it blue pin-stripe they wore back then?) about how PCs were going to revolutionize the computer (main-frame) market, and that IBM should get on-board (which they finally did 5 or 10 years later; too little, too late)]. IBM's problem wasn't stupidity or reputation, I think I'd boil it down to their inability to hold two mutually contradictory ideas in their heads at the same time (a capability every single one of us personally has). Refrigerator, automobile, PC, cell-phone - the guys who SHOULD have owned the market failed miserably to innovate in a self-disruptive way. The other problem is that the chance of actually going to market with a "break through" product is small. Some claim that for every 100 good ideas, only one makes it into development, and for every 100 development projects only 1 would return a positive ROI, return on investment. I suspect we've actually gotten better at this, although I've not seen the statistics, but never-the-less, the reason why small companies do this better than large is simply because they've got nothing better to do, and they've little to lose.

Maybe Harshaw was actually growing ruby and other crystals, not sure about doped silicon. They made windows for NASA and for USAF's laser programs (were classified when I was there). Harshaw was also a subcontractor for the Manhattan Project, (Uranium hexafluoride) and made a bunch of experimental rocket fuels in the late 50's early 60's.

""""For instance, why would a risk neutral firm not work enough on high-variance moonshot-type R&D projects,"""

Because while the narrow goals of the Apollo program were accomplished in the end it proved to be a dead end.

The narrow focus of the project meant that huge amount of money was spent on getting 3 people to Moon orbit and 2 on the moon surface for a few days and back to earth. Once that was accomplished a few times then most of the things built for it were scrapped.

They then built the Shuttle program which now too is scrapped

They are trying now to build Apollo 2 with the SLS and Orion programs which are Apollo on steroids.

Luckily there are now a few private or semi-private efforts to build launch systems which are more conservative in their funding requirements and more radical in their design and operation

Hard to have innovation when you have Obama taxing the bejesus out of people in order to fund Solyndra.

This is why I worked at small biotec companies instead of major pharmaceuticals. Small biotec companies had to focus on one or a few innovative projects based on specific knowledge developed by the company. Big companies had a very political process for selection and progression of projects that favored incremental improvements over major leaps. I liked swinging for the fences.

Comments for this post are closed