Adam Smith on Why Patents Are Preferable to Prizes

Not to mention that, historically, governments have reneged on prizes when the “wrong sort” solved the problem. See the maritime clock, Bessermer’s steel process, etc. In many cases, years later, they did settle up, but that is hardly incentive.

Would we have the steam engine given that those who invented it and key improvements weren’t from the right institutions or the accepted religions?

And what of Robert Hooke, one of the greatest minds and inventors, but unable to gain membership in the Royal Academy because he had to work for a living. This was solved by finding him a job at the Academy, but still.

@JK Brown,
Not to mention that, historically, governments have reneged on prizes when the “wrong sort” solved the problem. See the maritime clock, Bessermer’s steel process, etc. In many cases, years later, they did settle up, but that is hardly incentive.
Good point. Read the linked post by Taylor for more on this.
@Matt Moore,
This doesn’t apply to private prizes, which are a good means of matching technical problems with a wide source of potential different solutions.
You’re right.

BTW, Gordon Tullock remarked on giving prizes:

Direct rewards, in the form of prizes, for scientific work would also seem desirable. It is disgraceful that the Nobel prizes and the new Balzan prizes are practically the only substantial ones that a scientist can win. There should be many more prizes, and they should be much larger. There is no reason why a man who has made a really significant contribution to scientific knowledge should not be rewarded by very large sums of money. A system of prizes should be aimed at two objectives: specific discoveries and unspecified developments. The difference between the two may be neatly summarized by two discoveries of considerable importance to astronomy: the chronometer and [168] the discovery that the apparent direction of stars shifts slightly according to the direction in which the earth is moving. The first was seen as a need by the British Admiralty, and a prize for the first successful chronometer was offered. Eventually an expert clockmaker succeeded in producing an instrument of the required degree of accuracy and, after some trouble, collected his reward. The second discovery, by Bradley, could not have been specified in advance because no one suspected that it was true. Further, if anyone had suspected it, he could very easily have checked his suspicions. The discovery, although of great importance, could not have been the subject of a specified prize offered in advance. It could, however, have been rewarded by a prize for an “advance in astronomy.”

The specified reward is an excellent way of directing research toward some specific end, whether that end is large or small. As an extreme example, surely offering a reward of $1 billion for the first successful ICBM would have resulted in both a large saving of money for the government and much faster production of this weapon. At the other extreme, there are large numbers of minor practical discoveries which would be desirable but which, for one reason or another, are not patentable. Offering rewards for such discoveries would appear to be an effective method of encouraging this type of highly useful science. Much of present-day agricultural research could, for example, be done in this way, and thus the vast bureaucracy which now both carries out and impedes research in this field could be eliminated. Other areas where little or no research is now done could also be fertilized by this method: criminological procedures, for example.

Nor would specified prizes necessarily be confined to applied science. In the pure field, too, innumerable problems are suitable for such awards. But the non-specified award seems on the whole more suitable for the pure field. The patent, of course, is a non-specified reward for research in the applied field, and a most successful method of encouraging research. Non-specified rewards in the present-day world, such as the Nobel prizes, are rare and more valuable for the publicity and prestige than for the money. While I would not decry the value of publicity and prestige, most scientists would, I think, rather have more and larger prizes. But while such prizes are desirable, it would be highly undesirable to have them distributed by the same bodies. The maximum possible dispersion of decision-making powers on the question of who gets a prize is desirable. We discussed previously the advantage that the scientific publication system derives from the fact that there are many journals with diverse editors, so that work rejected by one stands a chance of being [169] accepted by another. The prize-awarding process should be similar. There should be a large number of individuals or boards that could award a scientist a prize, but no one of these should be able to say that he would not get one.

There is little else to say about the organization of these prizes except that diversity is desirable in every way. Not only should there be a number of prizes available in, say, physics, all offered by different bodies, but there should also be some prizes restricted to certain fields of physics, and other prizes for which physicists’ discoveries must compete with, say, biologists’. The award-giving process should be so organized that no scientist could greatly benefit his chances for an award by any “politicking.” Publication of his work should be all that is needed to make him eligible for an award. The people who decide who gets the award should not require applications, but should simply read the literature and reward the best items they see. Further, since it is sometimes difficult to see the real importance of a discovery immediately, a good many of the awards should be given only some years after a work was originally published. This is particularly important for the largest rewards. Frequently, some line of research which appears important when it is completed is seen four or five years later to be a dead end. The type of premonition which seems to lead some scientists into work which will lead to further great discoveries in the future is a rare and valuable gift, and it can be rewarded only by delaying the granting of rewards.

Some will object to all of this on the grounds that scientists are motivated by other things than a desire for money.8 In my experience, scientists themselves are particularly likely to make such statements. Their tendency to talk about their lack of interest in monetary rewards is equaled only by their tendency to bemoan their “low” pay. I have divided motives for scientific investigation into three categories: desire to make practical applications, curiosity, and induced curiosity. Clearly, the first is already largely motivated by material [170] considerations. The second would not be greatly affected, but a man who was genuinely interested in some scientific problem might just as genuinely be interested in money. If presented with a choice between investigating the problem that has engaged his interest at a low salary and some other problem at a high salary, he might well choose the latter. If the compensation for his scientific work was the same as for both alternatives, he would probably choose to work on the first. Thus, even for people motivated by curiosity, the amount of money likely to result from scientific or non-scientific activity is relevant.

How is selecting the proper duration for patents any less of a central planning problem than picking the right prize?
This is also assuming some kind of homogeneity between inventions (how fast they can be profited from, how fast they could be emulated, …).

@Ann Sherman,
The market IS a prize system.
Sure, but I think you understand the context, right Ann?
@Silas Barta,
Slightly off-topic, but why is Wealth of Nations so hard to parse?
From the reading I’ve done of books written between the late 1600s and the early 1800s (which is probably more than 5 and less than 15), that seems to be the way they wrote. I think they assumed, correctly, that readers had more patience for so-so prose because they had fewer alternatives. Think of the cost of books back then, which was huge.
@Julian Couvreur,
How is selecting the proper duration for patents any less of a central planning problem than picking the right prize?
This is also assuming some kind of homogeneity between inventions (how fast they can be profited from, how fast they could be emulated, …).
Both are central planning problems but, since you asked how it is less of one, the patent has zero value if not used in the first 17 (now 20) years. So the market is telling you the value over the next 20 years. As you say, that’s central planning too, but it’s less central planning than somehow trying to figure out what to give a prize to.

Since (targeted) prizes cannot be set for unanticipated inventions or unknown problems, the policy choice between patents and prizes exists only when a particular need or problem is known. In any such case, a fixed patent length is likely to be either too long or too short whereas the research prize can be tailored. While designing optimal research prizes is obviously very difficult, unlike patents, these need not be set by the government.

It seems to me that when there is a real choice between prizes and patents more things speak in the favor of the former. For example, unlike patents, prizes do not create monopoly deadweight losses. Unfortunately, this choice between the two does not often exist due to the limited scope of targeted research prizes.