Make your own free website on
A Theory of Price Control

Steven N. S. Cheung

PRICE or rent control is but one of many forms of legislative action which interfere with private contracting in the market place. To delimit the scope of my analysis, I shall use the term "price control" to refer only to any set of regulations which satisfies the following three conditions.

First, the control must fix the price (or income) terms of private contracts; this categorically excludes any laws which regulate the distribution of income among the contracting parties on a share or percentage basis.

Second, the control must involve no appropriation of proceeds to or from the government; taxation and subsidization are thus excluded. Finally, the fixing of price must not be associated with direct government sales, purchases, or manipulation of resources so as to maintain the regulated price; by this stipulation, price "support" is also excluded.

Even on such terms, legal regulations to control price are still many and varied.   How does price control lead to non exclusive income?

For illustration let us suppose that a tenement's constant monthly market rent of $100 is reduced by law to a controlled rent of $60. Assume for simplicity that this control will last to perpetuity. Who is granted the exclusive right to the $40 of rental income taken from the landlord?

One may be inclined to think that it is exclusively assigned to the tenant. Perhaps it is; but in general or part of that portion will have no exclusive claimant. The customary lack of exclusivity in this case may be illustrated as follows.

It would be relatively simple although at some cost to delineate and assign exclusive use rights to a portion of the tenement to the tenant. That is, instead of taking $40 from the landlord, part of the physical area of the tenement might be assigned exclusively to the tenant so that the landlord's remaining portion would yield a monthly market rent of $60.

It would also be relatively simple although again at some cost to assign exclusively to the tenant that portion of the rent taken from the landlord. This could be done, for example, by issuing stock against the market value of the tenement and giving 40 per cent of the shares to the tenant.

Abstracting from the higher arrangement costs that now would be incurred, either of the above situations would result in exclusive assignment of the entire rental income of the tenement. The landlord and the tenant would now be joint owners of the tenement, and constrained maximization implies that the tenement would be used as though no control existed. However, when the law governing the control avoids explicitly making the tenant a part owner of the tenement, the assignment of exclusive rights to the portion of rent diverted from the landlord is no simple matter. Of course, a sufficiently comprehensive set of legal provisions, abstracting from the costs of enforcement, could in principle endow the tenant with an exclusive claim to that portion of rent. When the actual control falls short of that comprehensive list, competition among contracting parties for the resultant nonexclusive income will tend to dissipate it.


Given the existence of nonexclusive income and its tendency to dissipate, each and every party involved will seek to minimize the dissipation subject to constraints, This will be done either through seeking alter natives in using or producing the good so that the decline in resource value is the lowest, or through forming alternative contractual arrangements to govern the use or production of the good with the least rise in transaction costs, or through the least costly combination of the two procedures.

This proposition is based upon the notion that, to the contracting parties involved, the dissipation of nonexclusive income constitutes a waste. The dissipation as discussed earlier will occur, but the postulate of maximization dictates that whatever dissipation must occur necessarily be a constrained minimum. Thus, instead of viewing the behavior as simply a reaction to the dissipation of nonexclusive income, we now see it as an attempted minimization. The effects are the same, but the change in view directly and restrictively leads us to investigate the constraints subject to which the minimization is achieved. Once these are ascertained and specified, equilibrium analysis further requires either that the specified gains and costs be equated at the margin or that certain corner solutions be reached. The particular forms of behavior through which the dissipation, or rather the minimization of dissipation, will occur become predictable: the effects of price control are thus explained.

For illustration, let us return to our example of tenement where the constant monthly market rent of $100 is reduced by law to $60, still assuming that the control will last to perpetuity. To exclusively assign the $40 of monthly rent to the tenant without explicitly making him a joint owner of the tenement will require the effective enforcement of a comprehensive program of legal provisions.

The simplest such program that I can think of will include the following provisions: the landlord cannot evict the tenant under any circumstances; the obligations of both parties in regard to maintenance and repair of the tenement are clearly defined; the tenant is assigned exclusive use rights as if he were the land lord, including the right to demolish and reconstruct the dwelling; and the tenant is granted the right to sublet, partly or wholly, to anyone he sees fit. Anything short of such a fully enforced program, or its equivalent, will likely produce nonexclusive income.

Suppose that a comprehensive program of this type exists. But now suppose that the landlord has vacant possession; that is, the tenement he owns is not occupied. With this added condition, no prospective tenant has an exclusive claim to the $40 per month. If the tenement is not to remain vacant, the landlord has the options of occupying it himself, selling it, or renting it to another party as he sees fit. In other words, he now has the right to exclude. Let us suppose that, in the absence of control, he would have chosen to lease the property annually at a monthly rent of $100, along with other stipulations. Under the control, this form of contract is no longer feasible, for once a tenant occupies the tenement he has the right to pay only $60 per month.

Any alternative contractual arrangements would, in general, involve higher transaction costs. In choosing among options, the landlord will weigh the rental value of the tenement for his own occupancy, which in general will be less than $100, against the rental value net of a higher transaction cost that he could obtain with an alternative contractual arrangement. In either case, some of the rental income will be dissipated.

One might be inclined to think that if the owner considered the rental value of his personal occupancy to be less than $60, he would simply lease the tenement on a first-come-first-served basis at the set monthly rate of $60 and that prospective tenants would then compete for occupancy, spending time and effort trying to get ahead of one another until most of the discounted present value of the nonexclusive $40 per month would be dissipated.

It is no doubt true that prospective tenants will compete; but since the landlord has the right to discriminate and exclude, the candidates will seek the least costly arrangement that is legally feasible, so that the rental income for the landlord will dissipate the least. As one solution, they may compete to buy the tenement outright instead of leasing. If this is effectively restricted by law or by a relatively high cost of contracting, they will choose the next least costly arrangements such as the offering of "key" money for the right to occupy the tenement, in which case competition will generate a lump sum offer to the land lord which approximates the discounted present value of $40 per month minus a higher cost of contracting.

If such an arrangement is again effectively prevented, the landlord may still reduce the dissipation by renting or selling furniture to the tenant at an exorbitant price, by arranging to employ him, or by some other such device.

One might also be inclined to think that the key money arrangement actually negates the control; on the contrary, it constitutes an effect which would not have occurred in the absence of control. Indeed, as a general class of observation there is no conceptual distinction between the key money arrangement and the familiar first-come-first-served arrangement.

Both are contractual arrangements associated with different costs of transactions. From the key money arrangement (which involves relatively small resources in the formation of the contract) to the first-come-first-served arrangement (which involves a relatively high cost of waiting, violence, and the like), there is a spectrum of alternative contractual arrangements. The associated costs are transaction costs, for the arrangements will not be present without the rental transaction.   The absence of exclusive claimants among prospective tenants to the $40 per month implies that the present value of this nonexclusive income tends to dissipate; the effort to establish alternative contractual arrangements is in effect, an attempt to define more clearly than is provided by law who has the right to some of the $40. While the higher cost of an alternative contract constitutes a dissipation of non exclusive income, constrained minimization of the dissipation implies that the least costly feasible arrangement will be chosen. In general, the more restrictive is the law on alternative arrangements the higher will be the transaction costs.

Let us consider a variation of the example. Suppose under the aforementioned comprehensive program of rent control, the tenement is now occupied. The resident tenant would have become an exclusive claimant of the rent taken away from the landlord (that is, the $40) except that, as a variation, the control is now imposed on domestic tenements only. The tenant may then be legally evicted if the landlord demonstrates bona fide intention to convert the structure to business use. Among a number of alter native uses that satisfy the legal definition of a "business premise", let us suppose that the highest valued option to the landlord is a warehouse, and that, net of all costs associated with the conversion, the warehouse will yield a monthly rent of $75. Thus, although the $100 market rent of the tenement under domestic use has been legally reduced to $60, the landlord actually has exclusive right to $75 per month from his tenement. Conversion into a warehouse will mean that only $25 per month is being dissipated.

As with the earlier situation, however, both the landlord and the tenant have incentives to minimize the dissipation. Facing possible eviction, the resident tenant will seek to dissuade the landlord from making the conversion. By law the landlord who evicts the resident tenant cannot lease again for domestic use; thus he will be willing to settle for any arrangement which will net him the equivalent of a monthly rent exceeding $75. However, the cost of forming such an arrangement will be higher than that of the key money arrangement discussed earlier, because the law has created a situation where prospective tenants are not allowed to compete. Neither contracting party is an exclusive claimant to the $25, and while there is room for negotiation so that the conversion to a warehouse may not materialize, each party will also seek a greater share of the amount. The absence of competing offers may raise the cost of settlement so high that conversion of the tenement into a warehouse becomes the only action consistent with the minimization of the dissipation.

Still another variation further illustrates the point while also demonstrating how non exclusive income may arise from many directions. Suppose, again that under a comprehensive control program the tenant has an exclusive right to the $40 diverted from the land lord and is granted all other exclusive rights to use the tenement, except to demolish and reconstruct it. These latter rest with the landlord. It will be recalled that the landlord cannot evict the tenant. Suppose, further, that owing to some identifiable changes in economic conditions a $5,000 net gain will accrue to all parties concerned if the tenement can be demolished and reconstructed immediately. Reconstruction requires eviction of the tenant.

If the law allows the landlord to find similar housing accommodation elsewhere for the tenant, then in effect the landlord has been granted an exclusive right to the $5,000 minus the cost of resettling the tenant.

In the absence of such or similar provisions, the control has created a situation where neither the landlord nor the resident tenant is an exclusive claimant to part or all of the gain from reconstruction. As with our earlier situations, both the landlord and the tenant will then seek contractual arrangements to realize the potential gain. Any delay in reconstruction owing to the transaction costs results in foregone interest on the $5,000 and constitutes part of the dissipation of nonexclusive income.

Dissipation occurs also in the cost of negotiating for some contractual arrangement for dividing the gains, and agreement is more costly to reach because, as in our earlier variation, no competing offers are available. The situation is complicated further if the tenement is occupied by more than one tenant. The absence of competing offers induces the problem of the "holdout" and the resolution of such a dilemma may require the services of a third party. The court, for example, may make a ruling which reduces transaction costs.

The difficulty of predicting whether some contractual arrangement will be reached in the above variation lies not in any inadequacy of our propositions for price control but, rather, in our present lack of understanding of the nature of transaction costs under certain peculiar situations. In the present state of the art, weaker but still refutable implications can be obtained if only we are able to identify one certain observable situation which entails higher or lower transaction costs, in total and at the margin, than another.

Furthermore, it is not necessary that transaction costs under different situations be actually measured. An assertion of the ranking of these costs under different observable situations will often suffice. With an asserted ranking, the test of any implication derived from it will also be a test of the ranking itself, and the validity of the ranking will be enhanced by the presence and confirmation of multiple test implications. Thus we have asserted in regard to contractual arrangements designed to minimize the dissipation of nonexclusive income that the following conclusions hold, other things being equal: the costs of reaching agreement will be higher, the more legal restrictions are imposed on these arrangements; they will be lower with competing offers than without; they will be higher with a larger number of contractual participants; and so on.

To illustrate other aspects of the mechanism of reaching equilibrium, let us consider a case where the market price of a theater ticket is reduced by law from $10 to $6. Available tickets might be allocated by a wide variety of methods including waiting in queue, "friendship" with the ticket seller, different forms of rationing, or even physical violence. Alternatively or concurrently, "black" markets of various shades might emerge, the quality of the show might be allowed to deteriorate in a number of ways, or as still another possibility the curtain might go down altogether. By adding or removing such assumptions, we will vary the example in a highly artificial manner to illustrate the desired aspects.

Let us assume that in a free market the show is produced monopolistically and that at the price of $10 per ticket, determined by competitive bidding among customers, no seat in the theatre is left empty. Assume, for convenience, that the monopoly rent per ticket is $4, an amount equal to the price reduction resulting from the control (that is, from $10 to $6). Assume further that the control effectively prohibits the producer from reallocating his resources either to adjust the quality of the show, to produce fewer or shorter shows, or to produce something else. In other words, by assumption we rule out the possibility of any adjustment in resource use to reduce the dissipation of non exclusive income; furthermore, the producer will remain in operation without subsidization.

To restrict other options lets us suppose that a customer can obtain a ticket at $6 only by standing in line on an orderly first-come-first-served basis. The producer of the show therefore cannot discriminate in selling the tickets. What was once a monopoly rent from each ticket (that is, $4) now has no exclusive claimant under the control. Note that the ticket itself is a contract which promises a show of a specific kind in a specific setting, and so on. The cost of waiting in line is therefore a cost of contractual arrangement through which the right to see the show is transacted. Suppose further that the purchased ticket is nontransferable and that each customer is restricted to purchasing only one at a time (that is, to buy more tickets, he will have to rejoin the line).

Finally, suppose that the waiting time required to acquire a ticket is known in advance, and that all customers have identical waiting costs.   Under the above specified constraints, not only will each customer wait an equal amount of time, but competition among them will lead to a waiting cost of $4 per ticket. Thus the amount of monopoly rent per ticket is dissipated or absorbed by the waiting (transacting) cost.

We may consider two variations which result in the same dissipation by supposing that the ticket is transferable and that there is no restriction on how many tickets a customer can buy. (1) If the costs of reselling tickets are zero, then a one man line will emerge, but under competition the total waiting cost will be the same with one man as with many. (2) Suppose the costs of reselling tickets are positive and the cost functions are identical for all "waiters," each with an identical U shaped average cost in reselling tickets. In this case, generally, more than one man will be expected to stay in line and each to purchase the same number of tickets for resale. As the costs of reselling tickets increase, the total waiting cost will decline but the total transaction costs (waiting plus reselling costs) will be the same as in (1).

In the above cases, as in a number of others, the monopoly rent will be fully dissipated by different behavior associated with different constraints in each case.   Let us return to our earlier set of specifications, except that we now suppose that the waiting costs of the customers are not the same. In this case, the monopoly rent per ticket will be the same for all customers, but the intra-marginal customers, or those with lower waiting costs, will be capturing some of the monopoly rent. This is the general case, and it brings out a point which we did not elaborate in the preceding section.

In our discussion of the key money arrangement, it will be recalled, competition among prospective tenants generates an exclusive payment for the landlord. Yet with other tenements under similar situations, some "rent" will be captured by intra-marginal tenants whose costs of transacting the key money contracts are lower than the marginal contract. Of course, the transaction costs for the marginal contract will be lower with the key money than without.   We may again vary our ticket example to illustrate still another point.

Suppose not only that the costs of waiting differ among individuals, but that customers are allowed to hire "waiters" (who have no intention of seeing the show) to obtain the tickets for them, while again each "waiter" is limited to one ticket at a time. In this case, the transaction costs will include the waiting cost plus the cost of the hiring arrangement. Transaction costs for the marginal transaction will equal the monopoly rent of $4. The intra-marginal transactions, however, will not only capture some rents but, to the extent that some "waiters" are hired, the total rent captured will in general be larger than in the earlier situation where the hiring practice was not allowed. The reason is that the increased flexibility in contractual arrangements permits the utilization of individuals whose waiting costs are lower for some intra-marginal transactions. This is consistent with the discussion in the preceding section: other things being equal, the less restrictive is the law governing contractual arrangements under price control, the less will be the dissipation of nonexclusive income.

The above variations should have amply demonstrated the simplicity of reaching equilibrium solutions under price control once the relevant constraints are known. The mechanism is equally straightforward if (in our ticket example) the show is produced competitively, or whatever other alterations are made in constraints. Equilibrium requires that nonexclusive income be dissipated at the margin (or at comer solution, as when the theater might be converted into a concert hall as a result of the control).

Associated with the same margin of dissipation may be a variety of observations (or test implications) each corresponding to a change in constraints.   Assuming such arbitrary constraints as we do with the ticket example makes the equilibrium solution easy. What is difficult and important is to identify and simplify the constraints as they actually exist and are relevant to the control under investigation. Only by this latter discretionary choice of actual constraints, together with the equilibrium solution they imply, can we hope to predict the specific forms of adjustments in resource use and in contractual arrangements to be expected under price control.


Economic theory has seldom stepped beyond the constraints of private property rights. The importance of the costs of transactions in affecting behavior has yet to be popularly recognized. Contractual arrangements as a class of economic phenomena have only recently been subject to intensive tilling. In a paradigm where private property is taken for granted, where transaction costs are assumed away, and where exchange is discussed without reference to contracts, any approach which analyzes behavior through an examination of the constraints of property rights and transaction costs is irrelevant. But my choice here of such an approach is exactly because price control, by interfering with the terms of private contracts, imposes constraints on decision making that differ from those of private property. Thus, with modifications, the theory offered here may be extended to analyze any other regulations affecting the rights associated with private property.

Throughout this paper, legislative actions on price or rent control are taken as given. But treating the behavior of legislators as exogenous, as we did, is only an attempt to avoid a separate problem. The lack of theory to explain legislative actions must be my chief defense for using the term "dissipation," which implies economic waste. Yet in a world where each and every individual is asserted to behave consistently with the postulate of constrained maximization, economic inefficiency presents a contradiction in terms. Even outright mistakes are traceable to constraints of some type. The world is efficient, if the model describing it sufficiently specifies the gains and costs to make it so.

Consider, as discussed earlier, the waiting cost incurred in queuing for the theater ticket under price control. The "rent" thus dissipated constitutes a waste in the sense that valuable resources are allocated to the waiting, which produces nothing of specifiable value. Yet without any price control, we notice that queues also form during rush hours in supermarkets. We do not, however, consider the latter waiting cost a waste if, in our hypothesis intended to explain the waiting behavior, we are able to specify the gains involved: the customers as a whole are not willing to pay higher prices for the products, or to make separate payments, to cover the costs of hiring more cashiers to reduce the waiting time."

Thus the world is inefficient only when the system chosen to analyze it fails to fully specify tha gains and costs of every action described. But the specification of constraints sufficient to yield refutable implications may only be a subset of the specification sufficient to yield an efficient solution. Inasmuch as we have ignored the constraints binding legislative decision making, the implied solution in our analysis of price control falls short of satisfying the Pareto condition .

Back to the Index