Words: 2005
Price Discrimination There is strong evidence in th?? marketing literature that consumers are very different in terms of their willingness to pay for products. Empirical studies report high variability in consumers’ responsiveness to marketing-mix variables such as prices, in-store displays, and feature advertisements, as well as in their intrinsic preferences for brands. This empirical consistency suggests that firms have an opportunity to price discriminate profitably rather than charge ?? uniform price to all consumers (Gilbert, 2003, pp. 89).
There is ?? vast literature in economics on th?? theory of price discrimination , and ?? number of important papers in marketing have discussed different forms of price discrimination and how they might be implemented in practice. Whilst price discrimination by ?? monopolist always leads to profits that are at least as large as those under uniform pricing, th?? competitive implications of price discrimination in oligopoly markets are more subtle. Shaffer and Zhang (1995), for example, show in ?? theoretical model that targeted leads to ?? person’s dilemma in which all manufacturers issue coupons without profitably increasing their prices.
Competing firms may gain incremental profits when consumer target ability is very imprecise. However, as target ability improves, they also find th?? prisoner’s dilemma (Gilbert, 2003, pp. 89). Price discrimination by all competing firms may lead to “all-out competition,” which results in prices and profits getting reduced in all market segments (Gilbert, 2003, pp. 89). However, there are also counterexamples for which competitive price discrimination may lead to increased profits (Arrow, 1962, pp. 609???625).
Hence, in oligopoly markets it is an empirical question whether or not th?? particular market demand conditions are favorable for price discrimination. In spite of th?? quite extensive study of business firms’ pricing practices over th?? past twenty-five years, th?? application of th?? theory of th?? firm to th?? retailing case has been left at an unnecessarily primitive stage. For th?? most part th?? published discussions of th?? competitive equilibrium in retailing posit single-product firms, an assumption which can be viewed as unrealistic.
This paper is an attempt to establish ?? more adequate theoretical framework for explaining th?? price and product policies of retail firms. Primarily because empirical support for th?? thesis expounded here is most accessible in food retailing, th?? supporting evidence will be mostly in reference to supermarkets, however th?? conclusions would seem to be equally significant for other types of retailing(Arrow, 1962, pp. 609???625). Although it may be granted immediately that retail firms are almost always multi-product firms, there is evidence of some confusion over th?? nature of th?? product.
Many writers have noted that th?? output of retailing consists not of th?? goods sold to customers however really of th?? services of storage, selling, wrapping, delivery, credit extension and so forth. ‘ However this kind of statement makes retailing appear more unique than it really is. By th?? same logic we must admit that th?? automobile manufacturer’s output consists not of automobiles however rather of th?? services of milling, stamping, shaping, machining, assembling, painting, storing and so on.
For purposes of measuring value added in th?? industry it is these services which are relevant. However in ?? discussion of pricing policy, on th?? other hand, one must focus on th?? unit of sale (Dreze, Hoch, Purk, 1994, 301???326). Th?? grocer expresses his prices as so much per pound of coffee, and so much per pound of beef rather than as so much per dollar of credit extended or per pound-mile of delivery service. Example Price Discrimination in ?? Supermarket Chain Given ?? multi-product retail firm, profits are maximized only if ?? style of price discrimination is practiced.
Th?? reason for this is that retailers’ cost structures are much more like those of public utilities than has generally been recognised and profit maximization under conditions of high fixed costs and common costs is essentially ?? matter of operating as close to capacity as possible and setting prices so as to take into account th?? elasticities of th?? various demands which th?? producer faces (McFadden, 1981). However th?? public utility analogy provides only ?? starting point for th?? analysis of retail pricing. Like some utilities (railroads, for nstance, which can bid for traffic of differing demand elasticities) retailers are typically free to add or delete product lines. However many types of retailers and especially supermarkets are free to manipulate their product lines and prices in order to take into account not only th?? various demand elasticities nevertheless th?? interrelated demands. Because of th?? much-advertised convenience of ‘one-stop shopping’ th?? cross elasticities of demand are negative among countless combinations of goods in th?? supermarket, Profit-maximizing pricing must take these cross elasticities into account(McFadden, 1981).
There is considerable evidence that supermarket operators do indeed establish prices with not only price elasticities however cross elasticities in mind. Th?? staple foods, for which th?? demand functions faced by th?? individual store are quite elastic, probably carry very low margins over direct cost whereas th?? non-staples, for which th?? demand functions fitted by th?? individual store are more inelastic, are probably priced so as to make ?? much more substantial contribution to overhead and profit.
Thus th?? buyer of th?? coffee on th?? ‘special’ list is being favored whilst th?? consumer whose tastes run to caviar and canned onion soup is being discriminated against. Th?? material which follows first deals with th?? probable nature of th?? cost functions for th?? supermarket retailer. Then th?? theory of th?? price and product equilibrium of th?? firm, given differing demand elasticities among th?? products sold, interrelated demand functions and th?? option of adding products, is discussed.
Th?? empirical evidence that th?? theory is an explanation of actual practice is reviewed and finally some of th?? implications of th?? theory are set forth (McFadden, 1981). The Theory and Model We will first review Clemens’s model as it applies to th?? supermarket. Let us posit th?? retail firm in th?? large numbers case selling ?? single product. Th?? Chamberlinian tangency solution follows with its proof of excess capacity. However if it is simple for th?? single-product firm to become ?? multi-product firm, as it is in retailing, we must drop th?? ingle-product assumption, for th?? retailer will cast about for another product which can be sold to utilize th?? excess capacity. Here Clemens introduces th?? assumption that th?? marginal cost function is not affected by th?? choice of th?? alternative products, i. e. it matters not whether output is increased by adding product B or product C, th?? marginal cost is th?? same in both instances. Thus in Fig. I there is however ?? single marginal cost function shown. In order to increase profit by th?? greatest possible amount, th?? retailer will add those products for which marginal revenue overlaps marginal cost by th?? greatest amounts.
It follows that th?? firm will adjust prices, and hence output, so that th?? marginal revenues will be equated to each other and to marginal cost paralleling Mrs. Robinson’s model(Johnson, Myatt , 2003, 748???774). Th?? nature of Clemens’s equilibrium is shown in Fig. I. Given th?? marginal cost function MC, th?? retailer will first invade th?? product market for which th?? demand is least elastic. If he were to remain ?? single-product retailer, he would expand output of th?? initial product to ???? where marginal cost and marginal revenue are equated.
However if he can enter th?? market for th?? second product, it is clear that profits can be increased if he were to divert some of his capacity from th?? first product to th?? second since at any price lower than OP1 th?? marginal profit on th?? first product is less than th?? marginal profit on th?? first units of th?? second product. Similarly th?? retailer will proceed to invade all those product markets in which marginal revenue overlaps marginal cost, so adjusting prices as to equate to marginal cost th?? marginal revenues for all items handled. In Fig. EMR is th?? line formed by th?? points of equal marginal revenue(Johnson, Myatt , 2003, 748???774). Given this equilibrium of th?? multi-product firm, price discrimination is practiced in th?? sense that various percentages of profit are earned on th?? many products. Clemens points out that price discrimination of this sort is essentially th?? same as price discrimination among identical goods (or services) sold in markets of differing demand elasticities. Th?? firm reaches many markets by producing many products or many variations of th?? same product.
In order to be in long-run equilibrium, then, th?? firm must have invaded all th?? product markets in which th?? price overlaps th?? marginal cost(Johnson, Myatt , 2003, 748???774). Clemens proceeds to spell out various limitations to product diversification, however we do not need to review them here. We can move on to adapt Clemens’s analysis to th?? supermarket case. We can immediately free ourselves of Clemens’s assumption that th?? marginal cost function is unaffected by th?? product mix. more widely applicable version of his model would assume only that at any time th?? marginal cost of expanding output (by expanding th?? output of any of th?? many products) can be identified ??? it need not be th?? same marginal cost for each product. In th?? multi-product case th?? marginal cost of increasing th?? output of ?? given product is, of course, more accurately considered ?? function of th?? output of all other products. So we should drop th?? single marginal cost function of Fig. 1.
We must then abandon Clemens’s formulation of th?? short- run equilibrium (marginal revenues equated to each other and to ‘th??’ marginal cost) in favor of th?? more general statement that marginal profits for th?? various products must be equated. In th?? long run, equilibrium requires that all products on which marginal revenue overlaps marginal cost be added to th?? product line, and output of each expanded to th?? point where marginal profits are all zero(Johnson, Myatt , 2003, 748???774).
This model differs from th?? familiar Robinsonian model primarily in that it substitutes th?? multi-product assumption for th?? single- product, multi-market assumption, and it assumes that th?? number of products is variable. Th?? same model can be derived from Brems’s multi-product case by dropping th?? criteria (characteristics), other than price, which he includes as variables. ‘ Brems’s analysis runs in terms of th?? marginal cost of increasing revenue, rather than output.
Revenue can be increased not only by altering th?? price nevertheless by altering th?? various characteristics of th?? product or by increasing th?? number of products. For any given increase in revenue there exists ?? least-cost change in criteria (including price) for achieving that increase in revenue, By increasing revenue from zero by ?? succession of least- cost steps, ?? minimum total cost function, with revenue as th?? independent variable, can be constructed.
Equilibrium can then be described in terms of th?? total cost and total revenue functions(Johnson, Myatt , 2003, 748???774). If th?? criteria other than price are taken as data in th?? Brems model, profit maximization in both Brems and our adapted Clemens model results in price discrimination (in th?? broader sense) if th?? products face demand functions of different elasticities. Even without specific evidence on th?? question, one would expect this brand of price discrimination to be practiced in th?? supermarket.
Th?? fact that supermarkets handle ?? multitude of products is enough to indicate that th?? opportunity to discriminate exists. For any given firm th?? demand functions for th?? staples are probably considerably more elastic than th?? demand functions for th?? non-staples. Th?? term ‘staple’ in this context implies that th?? good is one bought frequently and in considerable quantity by th?? housewife. She will surely be more aware of price differences on such items than on th?? items which are bought less frequently.
Since all supermarkets carry staples, price comparisons are relatively simple to make. Price comparisons on non-staples, on th?? other hand, are less likely not only because th?? housewife buys any given non-staple less frequently nevertheless because th?? money spent on any single non-staple is too small relative to th?? total food budget to warrant price-consciousness with respect to that item(Leslie, 2004, 520???41).