Harold Demsetz (1967) in his classic paper “Toward a Theory of Property Rights” makes the case that property rights arise endogenously when the cost of the commons problem begins to exceed the cost of exclusion, and illustrates with the case of Native American tribes and land rights. Once buffalo become sufficiently valuable (due to the new fur trade) as to make the commons problem acute, private plots of land become the less costly option, thus saving the buffalo from overhunting. Thus, the the tribes in the Northeast, where the fur trade was prominent, developed property rights in land, while those in the Southwest, where it was not as important, did not.
Demsetz concludes his paper by suggesting further lines of inquiry, for example intellectual property:
Consider the problems of copyright and patents. If a new idea is freely appropriable by all, if there exist communal rights to new ideas, incentives for developing such ideas will be lacking. The benefits derivable from these ideas will not be concentrated on their originators. If we extend some degree of private rights to the originators, these ideas will come forth at a more rapid pace. But the existence of the private rights does not mean that their effects on the property of others will be directly taken into account. A new idea makes an old one obsolete and another old one more valuable. These effects will not be directly taken into account, but they can be called to the attention of the originator of the new idea through market negotiations. All problems of externalities are closely analogous to those which arise in the land ownership example. The relevant variables are identical.
Demsetz’ analysis, like that of his Chicago contemporaries, is characterized by a distinct lack of process. Whether (as he claims for the West) property rights arose “largely as a result of gradual changes in social mores and in common law precedents” (p. 350), or whether it happens suddenly and discontinuously when an exogenous shock suddenly alters the relative institutional costs (as he claims for his example of the Plains Indians), change happens in the passive voice. The equilibrium states are important; the agents who respond to and drive forward the change are a black box.
Penetrating into this black box reveals that the relevant variables are, on the contrary, not the same at all. Not that Demsetz’ analysis is wrong so far as it goes, but that his analysis overlooks (or abstracts away from) a particularly relevant variable: the entrepreneur.
There is, of course, a reason for him to do this. The Kirznerian entrepreneur, for example – i.e. someone alert to arbitrage opportunities – is an extremely abstract concept. We can conceive of him existing in nearly any imaginable context, however much he may be impeded in practice. In the story of economic equilibration, we can reasonably take his existence for granted.
It is less legitimate, however, to take his existence for granted in other forms. This post will compare the forms of entrepreneurship necessary to bring about 1) economic equilibration, 2) institutional change, as in Demsetz’ example, and 3) political change, as would be necessary to establish an IP regime. Because of the differences in concreteness among them, social scientists have less warrant to assume an optimal result of the second than the first, and far less in the third than in the second.
The principle of methodological individualism constrains scientific explanations of social phenomena to those that are, at least in principle, reducible to the deliberate or emergent outcome of individual plan-makers. If hypostases and reifications are forbidden, abstractions must be judiciously employed: individuals must be regarded as similar to the extent required to explain the phenomenon, but no further. These abstractions are called, after the Weberian tradition, ‘ideal types’, and after Schütz, are said to vary in their ‘anonymity’ based on the specificity of their content. Thus ‘Human’ is more anonymous than ‘American’ (which contains specifiers like ‘speaks English’), which is more anonymous than ‘store clerk’.
The power of economics, the “queen of the social sciences”, lies in its restricting itself to only the most highly anonymous ideal types. These are types that apply strictly to all humans: prefers more to less, consciously evaluates options, etc. From these minimal assumptions, economists can tell a story of spontaneous impersonal coordination and the emergence of a single price for like goods across an entire economy. The story works because the driver of the process, the pure arbitrageur, is an ideal type sufficiently anonymous that economists no longer bother even to justify the assumption of his presence. Wherever humans prefer more to less, are alert to opportunities to profit, and face others similarly eager to profit, we can expect the law of one price to prevail.1 This is the central result of economics, which we can reasonably expect to obtain wherever there is a money economy.
Part of the reason economists can get such powerful results from such anonymous ideal types is that the story involves the emergence of a tight system constraint. As Koppl (2002, p. 14) explains, “The system constraint is the constraint imposed on individual action by the larger institutional system within which individual action takes place.” This lets us abstract away from the particular internal motives and mental processes of entrepreneurs and arbitrageurs because
under a tight system constraint, their actions will be driven into approximate conformity with the underlying situational logic. Those whose actions stray too far from this logic will suffer losses that, if uncorrected, will drive them from the market. . . . This is also the condition that lets observers rely exclusively on anonymous types.
The story of market equilibration is the story of the emergence of a tight system constraint through competition and arbitrage. Because money prices are a focal axis for competition, each competitor finds himself more or less constrained in his pricing decisions, regardless of his conscious reasons for doing so, or what he might otherwise like to do. Thus, economists can reasonably ignore these factors and generalize the story into “economic law”.
This result has been claimed with equal generality in other contexts besides the money economy. Demsetz’ story of the emergence of property rights in land draws an analogy from market equilibration to institutional equilibration with the claim that, like a firm, institutional outcomes change once the marginal costs and benefits of the various options change. The obvious question then is: what is the relevant system constraint for the plains Indians that compels them to adopt property rights in land?
The analysis in this case must resort to more concrete ideal types, such as learning man – not quite so anonymous as acting man, but still mostly taken for granted in price theory. Property rights in land would be a new convention in practice, but hardly a leap of faith, as they could observe European settlers organized along such lines, apparently with much success. The ideal type of the person who learns from those around him is also highly anonymous, and with determinate results so far as we can assume that both parties prefer one outcome over the other.2 It is no stretch to assume that members of the tribe preferred the preservation of buffalo to their extinction, and that they recognized the consequences of the looming commons problem.
More importantly, however, the property rights which arose there were would have been the self-enforcing kind, meaning we have a plausible system constraint. Land is typically considered ‘excludable’ – that is, the costs of negotiation and enforcement of its property rights are both relatively low. In addition, the success condition is fairly easily to observe (more buffalo), leading to rapid memetic spread once a first mover exists. We do not know whether a chief instituted property rights by decree, but we do know he did not have to in order for them to prevail. In short, though we cannot take it as quite so foregone a conclusion as the partial equilibrium story in price theory, the system constraint is sufficiently tight that we can still be reasonably confident that property rights in land would have prevailed, without regard to the specific characteristics and culture of the tribe.
Other economists have stretched the analogy further. Priest (1977) demonstrates that, given a relatively weak system constraint (the increased incidence of cases under an inefficient rule), the common law process results (in the very long run) in efficient rules. He explicitly argues that this constraint allows him to avoid using the concrete ideal types usually involved in judicial analysis: “the tendency of legal rules to become efficient over time is independent of judicial bias or the method of judicial decisionmaking.” Whatever we may think of the practical strength of his stipulated system constraint, the argument is methodologically sound.
This would carry over quite nicely to intellectual property, as Demsetz suggests, if a similar process could be expected to lead to such a regime. In fact, however, IP is highly non-excludable – a limiting case, even – meaning the costs of negotiation and enforcement are prohibitively high for private parties. Because an effective IP regime requires uniformity ex ante, there could be no “first mover”. Especially in the internet age, where distribution costs are practically zero, the costs of defecting are low, and the benefits potentially very high.
This precludes a self-enforcing memetic spread of IP in the way traditional property rights have been observed to spread. The requisite ex ante uniformity can only be imposed by the decree of a central government which is itself willing to bear the enforcement costs, costs that become progressively higher as distribution becomes cheaper.
Negotiation costs, of course, are lowered by having the decision imposed centrally. But that centralized locus of the decision also prevents the use of anonymous ideal types in our analysis of the political process. Becker (1958), for example, sets out the argument for why the democratic process might in theory be considered a system constraint on political actors, but then argues against its practical relevance. Caplan (2007) makes more explicit the many failures of the democratic process to constrain political actors in a way that might make it tractable for the use of more anonymous ideal types, as in economics. To the extent that anonymous ideal types are insufficient to understand the dynamics of the situation, the analysis loses generality. We have to “get inside the heads” of the relevant decisionmakers in a way we did not have to for economic actors.
The fact is, because discretion is such an irreducible feature of political decisionmaking, there is no guarantee that its results will tend toward anything resembling an optimum. Freedom of entry and exit may serve as a weak constraint, weaker perhaps than even the constraint of relitigation for judicial decisionmaking, depending on the size of the territory. But in our world of neutered federalism, migration restrictions, and international cooperation, policymakers have wide discretion within broad constraints.
For this reason, attempts to integrate political considerations into economics at the same level of abstraction are abortive. Unless the monetary rule is strict and binding, for example, central bank reaction functions are extremely contingent. “It will make a great difference,” Machlup (1936) noted, “whether Mr. Keynes or Professor von Hayek is governor or expert adviser of the central bank.” Where discretion is the rule, the concrete personalities and institutions cannot be abstracted away from.
We may note as well that it is precisely where discretion and imagination are the dominating factors that generality fails in the private economy as well. Price-quantity equilibration over a given set of potential products can be reasonably expected to result in a determinate equilibrium, and thus modeled. But when the range of products is not given – that is, when the economy is characterized by technological progress – the fate of modeling can only be regarded as an ignominious failure. Lachmann (1977) notes, regarding the fate of neoclassical capital theory in the wake of the Cambridge Capital Controversy,
It is not the subjective nature of expectations, any more than that of individual preferences, which makes them such unsuitable elements of dynamic theories, it is the fact that time cannot pass without modifying knowledge which appears to destroy the possibility of treating expectations as a data of a dynamic equilibrium system.
In other words, the problems of technological growth and the accumulation of knowledge necessarily involve us in more concrete ideal types that impair the generality of economic theory. The Kirznerian arbitrageur is predictable and tractable; the Schumpeterian entrepreneur is not.
Similarly, the constraints on political entrepreneurship will have little in common from one institution to another. For this reason there can be no general and emergent theory of politics in the same way that economics constitutes a general and emergent theory of markets. The form of government, in this sense, is not “neutral” with respect to policy, as we may imagine it would be in the long run under free entry and exit.3 A general theory of a particular form of government could, in principle, be articulated. But the constraints and procedures that constitute a form of government are, by design, loose, and permissive of a great deal of discretion. There is no avoiding the particulars of time and place in political analysis.
The relevant variables in the two cases of property rights in land and IP are, contra Demsetz, entirely dissimilar. The relevant analogy is not the changing
For this reason, Demsetz’ story – while valid so far as it goes – does not have nearly the generality he claims for it in his conclusion.