Innovation: Policy, Law and Economics

Fall 2007 Midterm October 17, 2007

1. If cost (respectively, value) of an innovation is observable, is there anything wrong with making a prize equal to the cost (respectively, value)? (Address both cost and value.)

Cost: First, we must make sure that we have the right notion of cost. For an economist, cost is the minimum payment required to get someone to perform the R&D. This would entail that the relevant “someone” is the most efficient researcher who is available, and there may be a discrepancy between cost in the economic sense and the accounting cost that is claimed ex post. If the accounting cost is reimbursed, that might give an incentive to be inefficient. A further nuance is how to deal with the probability of failure. If we say, as is sensible, that “cost is the minimum ex post payment, conditional on success, that is required to get someone to perform the R&D,” then the ex post payment must be greater than the outlay on materials and salaries to researchers. This is the most difficult aspect, as the ex ante probability of failure is hard to assess ex post, even if one can assess the outlay on materials and salaries.

Nevertheless, if these problems can be solved, and if cost is correctly defined, such a scheme would work. The problem is that cost in the correct sense cannot be observed, so it is not reasonable to try to implement such a scheme.

Value: It is easier (but not trivial) to observe value. The problem with giving a prize equal to the value is that it may attract may researchers in a race, which dissipates the value. If the prize is a billion dollars, reflecting the social value, and if participation by a single researcher costs $100m, then we would expect ten researchers to participate. The total social cost is a billion dollars, so what was the point?

2. Assuming that a particular type of innovation will be funded with intellectual property rights, is it optimal to give long enough protection so that no innovations are lost due to insufficient profit?

Not if there are many innovations that would be forthcoming even with shorter protection. For all those innovations, the lengthening of protection increases the social deadweight loss (exclusion of users) to no purpose. In addition, the longer protection will attract many participants in a race, possibly wasting resources.

3. Why might it be socially efficient to create a reward system in which an early innovator is given some of the value that is created by his successors?

If the early innovation has no market of its own (such as a research tool), then its only reward will come from other innovators who themselves create products for users. Without getting some profit from later innovators, the early innovation would not be made. Of course, this argument supposes that the early innovation has a social value, so that we do, in fact, want someone to invest in it. We are supposing here that the early innovation facilitates (or cheapens) the later innovations. If that is not so – if the early innovator is simply playing a hold-up game – then there is no reason to let him share the profit.

4. Describe three ways in which licensing is precompetitive, and give an example of how licensing can be anticompetitive.

Procompetitive (how could I misspell that?) (a) creates production efficiencies, (b) facilitates follow-on products, (c) reduces the pricing of complementary products. Anticompetitive: Licensing can reduce competition where Congress clearly intended competition, such as between firms that sell substitute products. We gave two examples of this. (a) If the substitute products have noninfringing patents, then they should have to compete, instead of joining their commercial interests through licensing. (b) In the case of cost-reducing innovations, the firms sell substitute products, but there is also an efficiency reason to allow licensing. If the royalty supports the monopoly price, the right holder is probably charging a license fee beyond the value of his innovation (the cost reduction). That would ordinarily be thought “sham licensing’’ and anticompetitive. Notice that users are worse off than before the innovation occurred. This should be a tip-off.

Here is where I discovered that I have real lawyers in my class. Some of you mentioned “nine no-no” practices. Most of those (such as tying) are now under “rule of reason,” recognizing that they have pro-competitive and anti-competitive uses.