Olfier Price-Sfardms markets

out. Organizations of colluding firms are termed cartels. While many different types of cartels can exist, a joint profit maximizing cartel seeks to extract maximum collective profits for the cartelisis. Tliis means it essentially acts zs a collecdve monopoly. Such a cartel joindy produces that output which equates the indus-fS: try marginal revenue curve with die cartels collective marginal cost curve. The situation for two producers is illustrated in Figure 111. Each seller has his own average and marginal cost curves, MCj and MC2. Their joint marginal cost curve is obtained by horizontally summing MC] and MC2, and is labeled MCj. Clearly, joint profit maximization requires the minimizadon of the total cost of producing a given level of output. Consequendy, the firms output must be allocated so that in equilibrium the marginal cost of one more luiit from firm 1 (MC] ) just equals the marginal cost of another unit from firm 2 (MC2). Otherwise it would be possible to decrease total cost - or what is the same thing increase total profit - by reallocadng producdon from the high marginal cost to the low marginal cost firm. (See Resource Conservation in Secdon V if you dont remember why). Joint profit maximizadon further requires that the industry output, Q, be such that industry marginal revenue just equals the firmsjoint marginal cost, MCj. This determines output quotas for the firms, q] and q2, which sum to Q. The common price is E To avoid clutter, weve not depicted the profits of each cartelist. As always, profit is given by the difference between price and average total cost dmes quandty. Of course, in Figure Ilf, the equilibritmi quandty and average total cost are different for the two cartelists.

In practice, it can be extremely difficult for the participating firms to maintain such a collusive agreement. First of all, when the cost stricture of the -telists ), those with lower costs and higher outputs will enjoy larger profits. Thus, in Figure 111, firm 2 is the more profitable. More important, while die cartelists have a collective interest in maintaining the output restrictions, each has an indiridual interest in cutting the price, secredy expanding output, and enjoying bigger profits at the expense of rirals maintaining the agreement. Of course, the smaller the chance of detection, the greater

the incentive to break the agreement.

In any case, collusion by cartelists has a close analogue in your section when grading is done on the cur\e. If everybody agreed to cut study time by, say, 50%, grades would be the same, yet the study cost of achieving them woidd be cut in half. Yet, ethics aside, could a class of fifty effectively collude to reduce study time? Could e\en five? Even if you entered such a collusive agreement, woiddnt you or others secredy break it? The answers, of course, may vary. Nonetheless, the threat of student cheating cartels ofthis type clearly seem more imagined than real. Business cartels are no different.

One final and crucial point even if a cartel is successfiilly organized, its high prices and profits may well attract entry as dme goes on. Therefore, imless the cartel has some means of keeping profit hungry entiants out, long run profits may yet becdfhe zero (of worse)7

GOVERNMENT SUPPORTED CARTELS

One attractive solution to the problem of cheating and entry faced by privately organized cartels is direct government sponsorship of the collusive activity. For example, sellers can sometimes induce the

Fig. Ill Joint Profit Maximizing Cartel

MC,=MCj=MC,----./......

Other Prirr-Sitnrhrr.\ Alor/u/s

government, not only to set super-competitive prices, but also to keep potential rivals out of the cartelized business. To the extent that sellers can achieve such protection from the rigors of the marketplace, they shield themselves from the two greatest threats faced by any cartel. We can only present a few examples of government supported cartels here, but be assured that a complete list would prove quite daunting.

Consider the California trucking industry. With approximately ten thousand firms registered as common carriers, i.e., available for hire on an "as needed" basis, we would have an almost perfect example of a compeddve industry, if government policy toward the industry was neutral. Policy by the California Public Udlides Commission (PUC), however, has been far from neutral. In the past, the PUC not only set minimum common carrier rates - and took severe sancdons against any company daring to charge less - it also shielded exisdng carriers from new compeddon by retarding or outright blocking compedtion by new entrants. To enter the trucking business as a common carrier, an applicant needed a so-called certificate of pubhc convenience and necessity. This meant that the appUcant needed to demonstrate to the PUC that a "public need" existed for a new carrier Inevitably, the PUC saw litde need for new compeddon.

While the PUC today no longer blocks entry the way it used to, it sdll publishes standard trucking rates, which it calls tariffs. To charge less than these tariffs or to modify them in any way once they are set requires permission of the PUC. At present, about a third of all applicants are denied permission to cut fares below the regular tariff. Others are discouraged from e\en trying by the expense and delay of establishing before the PUC that they can afford to operate under low tariffs. Although these procedures are supposed to protect consumers from the prospect of losing goods in transit through the bankruptcy of a carrier, the real impact is to transform state trucking from an almost perfecdy competidve industry into a government supported cartel. The expense of supporting and maintaining this cartel is met by the taxpayers. The taxicab business in many California cities is likewise a government supported cartel. In San Francisco, for example, the city government not only sets high taxi rates, it also licenses and restiicts the number of cabs by issuing so-called taxi medaUions. Until 1978, a medallion once issued could be sold or transferred, and since the number of medallions issued was small in relation to the desiire for them, they sold for prices up to $30,0004 Conversely, if die San Francisco had

not severely restricted licenses and set up a cartel in the first place, the medallions could never have fetched such prices.

Federal agricultural policy also supports cartels. An example occurs in the case of tobacco allotments. Basically, these allotments are legal permits to grow tobacco on a restricted number of acres of land Land with a tobacco allotment sells for about $5,000 more per acre than land without. In short, the right grow tobacco is valuable because the government has made it scarce by setting up a cartel. Conversely, the legal right to grow corn does not sell for anything, because the government has not made this right artificially scarce.

KINKED DEMAND OUGOPOLY

When government policy is neutral and no explicit collusion, such as a direct agreement to fw prices or divide markets, occurs in an oligopolisdc maiket, we are back with a conjectural marginal revenue curve for each firm. If, howe\er, we are prepared to make a strong assumption about how firms imagine rivals would react to price changes, we can make some predictions about oligopoUstic behavior An exampl occurs in the kinked demand curve model, develo to explain the apparent rigidity of prices in some olij gopolistic markets. Suppose in Figure IV an individual oligopolist is producing output q and selling it at price! P. If the management believes that competitors would not follow a price rise, but would match a reduction,] then the irnagined demand curve will be kinked at di; existing price/output combination. For example, prices above P, even a small increase in price causes large reduction in units, because the firm alone sell above this price. The others wouldnt match i Similarly, even large price cuts below P produce fi added units sold, because competitors would mat( such cuts. As such, the firm cannot take much busin fi-om rivals. Hence, the demaiid curve is kuiked at P.

The marginal revenue curve follows first flat portion, and then die steep portion. This pi duces a discontinuity (jump) in the associated margij al revenue curve at the kink. In Figure fV, margin! revenue is positive up to output q and negative the: after However, marginal revenue could be posi beyond q as in Figure V.

If the marginal cost curve also passes throu] the discontinuity, the profit maximizing output-wh MC is as close to MR as possible - occurs at q. The sequence is that shifts in the marginal cost curve h no effect on the profit maximizing price/output

Fig. IV

Kinked Demand Oligopoly

bination, provided die new marginal cost curve passes jvthrough the discondnuity. Thus, in Figure IV, the ini-idal equilibrium given by the marginal cost curve J labeled MC is idendcal to the new equilibrium given by the new maiginzd cost curve labeled MC. Hence the stability of oUgopoly prices is explained. Can you see why market shares, the fracdon of total industry \ output or sales captured by pardcular firms, are stable in this model?

Note, however, that the theory does not" explain the price itself. The model explains only why the price may be invariant to cost changes in certain tioligopolisdc markets. Moreover, the empirical rele->ivance of the model is debatable. Price wars, for example, are frequent in some oligopolistic markets. Market shares also often vary. In any case, insofar as Soligopoly is not characterized by price compeddon, I non-price compeddon may be intense. For example, advertising, sales drives, and product differendation are all ways of potendally increasing sales at the expense of rivals. As Figure V illustrates, the kinked demand oligopoly equilibrium may also include eco-nomic profit. Of course, entry barriers of some type, perhaps economies of scale or government created obstacles, are necessary to mamtaui such profits.

OUGOPOLY IN THE AMERICAN ECONOMY

We have already seen in Secdon VI that poten-dal compeddon can take the fun and profit out of being a monopolist. Even in this extreme case, selling as the sole firm in an industry without the blessing of entry barriers does not guarantee the life of easy secure profit associated with classical monopoly. (We suspect this fact was appreciated by monopolists in

open-entry industries long before it was appreciated by economists.) Entry barriers are required for that.

But if classical monopoly behavior in the sense of price much greater than marginal cost cannot be inferred solely because a pardcular firm is, in fact, a monopolist, sdll less can it be inferred when the reactions of actual as well as potential rivals have to be considered. All of this is to say that the joint profit maximizing and kinked demand curve models are only two in a whole bewildering~sferie"s of 6ligopoly~models. Whether these two models or any of their numerous competitors are particularly useful ways of viewing oligopoly is at least partially an empirical question. In brief, the competitiveness of any one real world oligopoly cannot be resolved on the basis of theoretical considerations alone. As Professor Demsetz of UCLA puts it:

[We] have no theory that allows us to deduce from the observable degree of concentration [market share of sellers] whether or not the market price and output are competitive.

Nonetheless, considerable empirical evidence and theoretical scraps suggest that real world oligopolies who are not explicidy meeting to fix prices behave more as pure competitors than as monopolists.*

First of all, Lf real world oligopolists already achieve monopoly prices, then price should be litde affected when a monopolist becomes an oligopolist through the entry of a second firm. Yet in such situations, entry by as few as even one competitor often causes large reductions in established prices. Moreover, the prices charged by the same firm when it