EFMD EQUIS CREDITED Industrial Organization Lecture 8 Price Dispersion and Search Theory 學火旦 于 udan university
Binglin Gong Fudan University Industrial Organization Lecture 8 Price Dispersion and Search Theory
Role of information Recall bertrand paradox Incomplete information is one resolution of Bertrand paradox Acquiring information on prices is costly to consumers, and consumers al ways weigh the cost of searching against the expected price reduction associated with the search process
Role of Information • Recall Bertrand paradox • Incomplete information is one resolution of Bertrand paradox. • Acquiring information on prices is costly to consumers, and consumers always weigh the cost of searching against the expected price reduction associated with the search process
Tourist-trap model · Many souvenir shops Guidebook tells distribution of prices (i.e. broadly something like this 5% of stores charge price =10, 95% charge price =20 and so on Costs tourist c in time and expenses to visit a shop and check price or buy(this is what defines the lack of information hence for c =o all the results of standard Bertrand should hold). Lack of information is same thing as saying it is costly to obtain and or process information If price=p, costs p +c if tourist buys from first store and p+ 2c if tourist buys from second store
Tourist-trap Model • Many souvenir shops • Guidebook tells distribution of prices (i.e. broadly something like this 5% of stores charge price =10, 95% charge price =20 and so on). • Costs tourist c in time and expenses to visit a shop and check price or buy (this is what defines the lack of information. Hence for c =0 all the results of standard Bertrand should hold). Lack of information is same thing as saying it is costly to obtain and or process information. • If price = p, costs p + c if tourist buys from first store and p + 2c if tourist buys from second store
Tourist-trap model The big question: Is a competitive price charged?(Recall in standard Bertrand where there is full information competitive price is charged- this is what the text book and in lass we call as the full information price equilibrium) Suppose all stores charge full-information competitive price, Pc This price is equilibrium price only if no seller wants to charge a different price No firm wants to sell for less than pe =marginal cost as it will result in a loss Suppose one firm charges pl-p+8, where a=small positive number just less than c In this case, a consumer still buys from it, so store makes a higher profit. Thus, competitive price cannot be equilibrium price
Tourist-trap Model
Tourist-trap model Monopoly price is pl an equilibrium price? No (Just repeat the previous argument) A firm wants to charge p2Pl+8=P+28 Repeating argument: only possible single price equilibrium is monopoly price where no firm wants to charge more and if it does not pay for a firm to cut price, monopoly price will indeed be an equilibrium price. a deviant firm which with all firms charging the monopoly price will like to lower price only if it can induce new consumers to search for this low price shop Hence it has to lower price by more than the amount of the search
Tourist-trap Model
Tourist-trap model cost which is c. Now there is an additional point which we state here only heuristically (we do not do the maths for it)and i.e. even if this firm which starts charging a lower price, lower by an amount greater than c, the consumers would search for this low price store only if it is not one in many. In that case consumer have a very low chance (probability)of hitting the low price store and will not search. With a large number of firms then no firm will have an incentive to lower the price when other firms are charging a monopoly price. The maths of this is below but you are not required to know this part i.e. the mathematical derivation of this result
Tourist-trap Model
Tourist-trap model For a customer who initially arrives at another store, to even consider looking for th cheaper store, the cheaper store must charge p"-c(n-1)-1c. Why? Let d be the discount the deviating store offers, i. e. its price is p-d. Since there are n stores, and the tourist is already in one of the expensive ones, the chance that the tourist will find the other store is 1/(n-1). I she does find the cheaper store she saves d-c(i.e, makes an expected gain of (d-c)/(n-1). If she does not find the cheaper store, she pays c more than she would have if she bought at the first store. The probability of this is(n-2)/(n 1). The expected loss then is c(n-2 )/(n-1). Thus the consumer will look for the cheaper store on/vi (d-c)(n-1)>c(n-2)/(n-1) d-c>c(n-2), or d>c(n-1)
Tourist-trap Model
Tourist-trap model e. g. if the cheaper store chooses a discount d=c(n-1)+ 1c. From this you can see that if n is small, the deviating store does not have to lower price as much to attract customers. And the less the store needs to lower price the less profits it will for ego by doing so. This means the smaller n the less likely the monopoly price can be an equilibr rIum This is contrary to what we see under full information and we need to appreciate this result both from the point of view of learning and exams. recall that our argument did not hinge on the size of c. Reducing search costs does not help! The existence of infinitesimal search costs make the competitive equilibrium break down and only possible single price equilibrium features the monopoly price
Tourist-trap Model
Tourist-trap model In the light of what we have learnt we should be able to comment on the followin henomenon (1) Advertising and price-. Federal Trade Commission(FTC)opposes groups wanting to forbid price advertising (2) In an empirical study it was found that the price of eyeglasses is 28% higher in states that forbade advertising than in those that permitted it (Benham 1972) ()On a cynical note- Why some doctors and lawyers organization prevent advertising?
Tourist-trap Model
Price Dispersion The commonly agreed upon"law of one price stating that identical products sold at the same location at a given time period must be sold for Identical prices is actually rarely observed in any market. Most retail markets are instead characterized by a rather large degree of price dispersion The cost of obtaining price information might be one reason for price dispersion in homogeneous product market
Price Dispersion • The commonly agreed upon "law of one price" stating that identical products sold at the same location at a given time period must be sold for identical prices is actually rarely observed in any market. Most retail markets are instead characterized by a rather large degree of price dispersion. • The cost of obtaining price information might be one reason for price dispersion in homogeneous product market