SUMMARY: Industrial organization. Markets and strategies, Sintesi del corso di Economia Industriale

Scarica SUMMARY: Industrial organization. Markets and strategies e più Sintesi del corso in PDF di Economia Industriale solo su Docsity! CHAPTER 13 VERTICAL RELATIONS Overview Context:​ product sold to a retailer first, then by retailer to consumers. Effect on pricing and production decisions? Concepts:​ vertical integration, double marginalization, competition softening, hold-up and agency problems, vertical restraints. Terminology: - manufacturer/retailer - or, alternatively, producer/seller - or, alternatively, upstream/downstream firm - intermediate/final product - wholesale/final price - vertical relationship - vertical restraints (see below) - value chain Although we normally think of firms as selling products and services to consumers, the fact is that most firms sell to other firms, not to final consumers. There are several reasons why the relation between a manufacturer and a retailer is substantially different from the direct relation between a firm and the final consumer. 1. First, the demand faced by a manufacturer depends on the price it sets (the wholesale price) and on a host of other factors, most of which it does not directly control. 2. Secondly, retailers compete with each other (whereas consumers do not). In particular, each retailer cares about the wholesale price it has to pay the manufacturer as well as the wholesale price paid by other retailers. This is so because the wholesale price determines marginal cost (the retailer's marginal cost), and each firm's equilibrium profit is a function of all firms' marginal cost. 3. Third, is that the number of intermediate firms is small, whereas the number of final consumers is large; a firm that sells to the final consumer has more market power than a firm that sells to other firms. In fact, there are cases when most of the market power is on the buyer's side: large supermarket chains, for example, have a great degree of market power with respect to suppliers. These reasons justify the separate treatment of ​vertical relations​ between firms. "Vertical relations" → relations between two firms in sequence along the value chain​, as in the examples above. Normally, I will refer to the case of a manufacturer selling to one or several retailers. However, the analysis in this chapter applies more generally to cases when an​ upstream firm​ (e.g., cement producer, flour producer) sells to a ​downstream firm​ (e.g., concrete producer, bakery). 13.1 VERTICAL INTEGRATION (slides + book) Vertical relationship: new problems ● Vertical relationship causes many changes to the standard relationship between firm and consumers. ○ Many determinants of consumer demand beyond sellers’ control. – Retail price – Sales service – Local advertising ○ Retailers compete with each other ─ wholesale prices are (part of) the cost competing retailers, affecting downstream competition. ○ Number of customers (retailers) is relatively small. ─ Less market power for suppliers. ─ A new concept: buyer power. ─ Ex: Large supermarkets Selling to a retailer ● Retailer choose retail price ​p​ to maximize profit. ○ Demand for final product: ​Q = D(p) ○ Marginal cost: price charged per unit by manufacturer, ​w​ . ● Manufacturer chooses wholesale price ​w . ○ Demand is determined by the retailer (quantity it expects to sell). ○ Marginal cost of production, ​c​ . ● Double Marginalization ○ Manufacturer and retailer both choose price above (their own) marginal cost. Consider a structure consisting of an ​upstream firm (M)​ and a ​downstream firm (R). Firm R is best thought of as a retailer, while firm M could be the manufacturer of the product, a wholesaler, or the supplier of an important input to firm R. ​For example, M could be an oil refiner and R a gas station. Suppose there is a demand for the final product (supplied by ​R​), given by ​D(p). Regarding the production technology, let us assume the simplest possible structure: in order to produce one unit of output, ​R​ needs one unit of input. In fact, under the manufacturer/retailer interpretation this is the right assumption: in order to sell one TV set, a retailer must obtain exactly one TV set from the manufacturer. Suppose that ​R​ has no costs in addition to the wholesale price, ​w,​ that it pays its supplier. Finally, ​firm M​ has a constant marginal cost ​c​. (The situation is illustrated in Figure 13.1 below) Wanting to make a positive profit, the wholesaler sets w above marginal cost ​c,​ that is​ w > c​. The retailer, in turn, sets optimal output such that its marginal cost, ​w,​ equals its marginal revenue. This results in retail price ​p​R​. In this equilibrium, ​firm M​ earns a variable profit equal to area A, whereas ​firm R​ earns a variable profit equal to area B. Together, the "vertical chain" (​firms M ​and ​R​ combined) earn a profit A​ ​+​ ​B. Now suppose that ​firms M ​and ​R​ are integrated as ​firm M&R​ with the single goal of maximizing joint profit. This new monopolist inherits ​firm R'​s demand curve,​ D(p)​, and ​firm M'​s marginal cost, ​c​. Its optimal choice is given the intersection of marginal revenue and marginal cost. This corresponds to output level ​q ​M&R​ and retail price​ p​M&R​. In the vertical integration equilibrium, ​firm M&R​ makes a profit equal to areas A+ C in Figure 13.1. Since ​C​ is greater than ​B​, we conclude that ​firms M ​and ​R​ increase their joint profit by vertically integrating. The idea is that, under vertical integration, the wholesale price is simply a transfer price; therefore, it has no impact on retail price formation. For this reason,​ firm M&R​ is able to set the level of ​p​ that maximizes total value along the value chain. By contrast, under vertical separation, ​firm M ​sets wholesale price w above marginal cost ​c​; and ​firm R determines retail price ​p​ based on a cost​ (w)​ that is higher than the "real" marginal cost ​(c). We conclude that: If a manufacturer sets a wholesale price to a vertically separated retailer, then their joint profits are lower, and retail price is higher, than under vertical integration. This problem, known as the ​double marginalization problem​, provides a first justification for vertical integration. Under vertical separation, there are two monopoly pricing decisions in effect. If the only contractual instrument that ​M ​and ​R ​can use is the wholesale price, then two monopoly margins will be added to marginal cost, resulting in a price that is greater than monopoly price. As a consequence, the combined profits of ​M ​and ​R​ are lower than if they were integrated. Vertical Integration WITH DOWNSTREAM COMPETITION Things get a little more complicated if there is more than one downstream firm. The figure depicts the case when an ​upstream firm (manufacturer) M​ sells to two ​downstream firms (retailers) R​1​ and R​2 ​. Let ​w​i​ be the wholesale price paid by retailer ​R​i​ and let ​p​i​ be its retail price. Suppose that ​firm M​ merges with retailer ​R​i​ ; ​what impact do we expect this to have on prices? 1) The first effect we would expect is a ​decrease in ​w​1​ ​. This corresponds to the double-marginalization effect considered earlier: once ​M​ and ​R​1​ ​have a single profit maximization goal, one of the intermediation margins, ​w​1 ​, ceases to exist. There may still be a transfer price ​w​1​ used for accounting purposes; but as a decision problem, ​firm M&R​1 no longer considers two price margins. 2) The second effect we would expect is an ​increase in ​w​2​ . Firm R​2 ​, which was a customer of ​firm M​, is now also a rival of firm ​M&R​1​. An increase in wz has an effect on ​firm M'​s revenues, an effect that previously ​firm M​ was not taking into account. Now that​ firm M​ is united to ​firm R​1​, an increase in ​w​2​ also has an effect on downstream competition: it puts pressure on ​R​2​ to increase ​p​2​, which in turn helps ​R​1 ​, which in turn helps ​M&R​1 ​. In other words, an increase in ​w​2​ is a way of ​raising rivals' costs​, ​a strategy we discussed in Section 12.2. Finally, when it comes to ​p​1 ​ we find conflicting effects. Vertical integration eliminates double marginalization, which in turn provides a downward pressure on ​p​1​ ​(as we saw earlier in this section). However, vertical integration also has a ​competition softening effect​ that tends to push price up. When ​R​1​ considers an increase in price, it balances an increase in margin with a decrease in market share: the basic pricing trade-off we considered in Section 3.2. Now that ​R​1​ is part of ​M&R​1​, it should consider an additional effect: by increasing price and losing market share, ​R​1​ is giving ​R​2​ some market share. This in turn implies that ​R​2 ​ ​must order more input; and to the extent that ​w​2​ > c​, this extra input order increases the profits of the combined ​M&R​1​. Overall, we expect the profits of ​M&R​1​ to increase and the profits of ​R​2​ to decrease. Regarding consumers, we have conflicting effects: getting rid of double marginalization is good, but softening downstream competition is bad; in principle, the net effect can go either way. Recap from slides​: ☐ Effect of manufacturer integrating with one of the retailers? - Get rid of double marginalization. - Softens competition by raising rivals’ costs. ☐ Overall effect on final prices uncertain. INVESTMENT INCENTIVES So far we discussed pricing incentives along the value chain. In some cases investment incentives are also important. − Suppose that ​firm M​ comes up with a new electronic gadget worth ​v ​ to consumers, but consumers will only learn about it if ​firm R ​makes an investment​ ​(for example, training its sales force to sell the new gadget). Alternatively, − Suppose that ​firm R ​comes up with a new car model worth ​v​ to consumers, but production is only possible if firm M​ makes an investment​ (for example, designing and building a mould required to produce a car part). Suppose moreover that these investments are ​specific assets​ ​(Section 3.4)​: the sales force training is only useful for selling ​firm M'​s gadget; and the mould is only useful for making ​firm R'​ ​s car part. Let’s continue with the example of the car manufacturer ​R​ outsourcing a part to manufacturer​ M​. Frequently there are many contingencies which are difficult to predict and write a contract on​ (​for example​, how much it will cost to produce the part or how many cars will be demanded). Accordingly, at some point in the future the exact terms of the transaction between ​M ​and ​R ​will need to be determined; whereas ​M'​s investment decision must be made from the get-go. Then we have a problem. Specifically, suppose that the investment costs ​c = 5​, whereas the total value created by the investment is v = 8 ​(a value which is appropriated by ​firm R​ upon sale). Clearly, this is a worthwhile investment. However, once the investment is made it is effectively a sunk cost, which in turn induces a ​hold-up problem (cf Section 3.4)​: suppose that ​R ​makes a take-it-or-leave-it offer to ​M​, either ​p​H​ = 6 ​or​ p​ ​L​ = 3​. The situation is represented by the game in Figure 13.3. The only subgame-perfect equilibrium of this game ​(cf Section 7.2)​ is for ​R​ to offer ​p​ ​L​ and for ​M​ to choose ​a​. This is an inefficient equilibrium: total payoff is zero, whereas the choice of a would lead to a total payoff of 3. In this context, vertical integration may help solve the hold-up problem created by investments in specific assets: suppose that ​R ​acquires ​M​. Now decisions such as the investment level, the number of parts, etc., are made by ​R,​ whereas ​M'​s manager simply executes orders. Taking the ​M&R ​group together, ​R'​s manager correctly determines that the investment a is worthwhile (it costs 5 and generates benefits of 8); and accordingly, the efficient level of investment is chosen. If vertical integration solves some problems, it may also create new ones. Continuing with the same example: upon acquisition by the downstream firm,​ M'​s manager is at the orders of R'​s manager. As we saw in Section 3.3,​ this creates an agency problem: ​Firm R'​s manager wants ​Firm M​ (now a division of ​M&R​) to act in ​R'​s interest; but ​Firm R​ may not know exactly what to order, since ​Firm M'​s manager is better informed than ​Firm R'​s manager. In sum, When investments in specific assets are at stake, vertical integration alleviates the hold-up problem but increases the agency problem. The above analysis of the hold-up problem is based on a very stylized e.g.. However, the basic ideas are far more general. For example,​ we could assume that, instead of a take-it-or-leave-it offer by ​R,​ ​firms M ​and ​R​ negotiate over transfer payments and split the cake (that is, the value at stake). If ​M​ chooses ​a,​ then a value ​v = 8​ is at stake. If ​M ​and ​R​ split the cake, then ​R​ gets 4. This is still not sufficient to compensate for its initial investment of 5, so the unique subgame-perfect equilibrium is, again, for ​M​ not to invest. Recap from slides​: 一 Separation of production and retail can create hold-up problem. - New product by manufacturer, consumers only learn by training sales force at retailer. - Retailer design new product, but manufacturer’s investment required to produce it. 一 Example:​ retail firm comes up with a new car model worth ​v = 8​ to consumers. Production only possible if manufacturer makes investment​ c = 5​ required to produce a car part. - Specific asset: mold only useful for making retailer’s car part. - Problem: once the investment is made, effectively a sunk cost. - If ​R​ makes a take-it-or-leave-it offer to ​M,​ either ​p​H​ = 6 ​or​ p​ ​L​ = 3 一 Manufacturer’s problem:​ invest or not? 一 Vertical integration can solve ​hold-up problem​: achieve an efficient investment level. 一 But the resulting problem of agency: manufacturer and retailer have different information and different incentives. 一 Takeaway: vertical integration trades-off costs and benefits. 一 Other ways around the hold-up problem? ​YES. → write a long-term contract. Before making an investment, the upstream firm signs a contract with the downstream firm in which all the terms. Vertical Integration: the evidence and more theory In many industries, the last decades of the 20​th ​century witnessed a significant increase in the practice of outsourcing​, purchasing inputs from an outside supplier (even though those inputs are available internally). In fact, currently in the US almost one-half of upstream establishments make no internal shipments to their downstream integrated establishments. This evidence runs counter to the idea of vertical integration as a means to increase efficiency in vertical transactions (namely by reducing double marginalization). Much of the outsourcing shift was directed at foreign suppliers (that is, outsourcing took the form of offshoring). In this sense, one explanation for the trend is simply that firms were seeking cheaper suppliers. However, recently much of that outsourcing has shifted back to the US, and the lack of internal flows is still very noticeable. If vertical ownership is not associated with vertical flows, why then do firms vertically integrate? One explanation is that integration allows for the internal transfer of intangible inputs such as management ability, marketing know-how, intellectual property, and R&D capital. This would then suggest that the effects of vertical integration have less to do with strategic interaction in setting prices and output levels than with organizational efficiency issues. However, before discarding strategic behavior as a driver of vertical integration, we should note that, even if vertical integration is not associated with internal flows, it does imply the option of internal flows. In other words, the advantage of being vertically integrated with a potential supplier is that it forces outsource suppliers to be more efficient and bid more aggressively or else the buyer switches to the internal source. ─ Why? Does it create or solve inefficiencies? ─ Imagine an industry with 1 manufacturer and 2 retailers. ─ A retailer may make an investment in sales services that increases sales. This investment: ○ benefits the retailer 1. ○ benefits for the manufacturer, because of the sales increase. So, this investment makes the profit of the manufacturer go up ​(π​ ​M​ ↑)​ and the profit of retailer 1 go up​ (π​1​ ↑). We assume that these new profits generated are larger than the costs. ─ However, it may be that the retailer 1 is not able to appropriate the benefits from the investments. ─ Example: - retailer 1 trains sales people to assist consumers in assessing the value of a new electronic gadget. - demand for the gadget increases. - consumers buy from retailer 2, which charges lower prices since it didn’t not invest in training. ─ ─ → So, it is not clear whether ​π​1​ ↑ ​is going to happen, because it is possible that some consumers will go to retailer 1, get explanation about the new gadget, but then go buy from retailer 2 which hasn’t make the investment and so charges a lower price ​(p​2​ < p​1​)​. ─ ─ Anticipating the non-appropriability of the investment, ​the retailer does not make it​. ─ Loss of a profit opportunity for both the retailer and the manufacturer. ─ A ​RPM may prevent​ this situation: ○ if retailer 2 cannot lower its price, it cannot appropriate the benefit of retailer 1’s investment. ○ retailer 1 thus chooses to make the efficient investments. ⇒ Remember that looking for complex contracts which are able to solve inefficiency in vertical industries. In this vertical industry the inefficiency is not a pricing inefficiency, is not an hold-up problem, but it is an investment inefficiency, which derives from the fact that this investment is not fully appropriable by the firm who makes it. Then the resale price maintenance is a complex contract which is able to solve this inefficiency. (book) RESALE PRICE MAINTENANCE Consider the market for consumer electronics: personal computers, smartphones, and so on. These are complex products for which sales effort is important: consumers gain a good deal from the service provided at the point of sale. Now suppose that one retailer makes a large investment in sales effort, so that it attracts a large number of consumers to its store; and suppose also that a second retailer makes no investment in sales effort but sells at a lower price than the first retailer. One possible outcome is that many price-conscious consumers will visit the first retailer to learn about the available products; and then visit the second retailer to purchase the preferred product at a low price. This situation entails an important ​externality​ ​(cf Section 5.1)​: the investment in sales effort by the first retailer benefits both retailers. In fact, if most consumers are price conscious, then the investment benefits mostly the second retailer, who free rides on the first one. As a consequence, the incentives for investment in retail service quality are very low. Finally, the manufacturer suffers from this, for final demand depends to a great extent on sales effort. Resale price maintenance​ (RPM)​,​ the practice whereby the manufacturer imposes a minimum price on retailers, provides a way out of this dilemma. If the minimum price is effectively enforced, then every retailer sticks to it, that is, every retailer prices at the uniform minimum level; and then the free-riding problem disappears: not even price conscious consumers gain from purchasing from a different retailer than the first one they visit. Therefore, the benefits from increased sales accrue to the retailer who makes the corresponding investment . 3. EXCLUSIVE TERRITORIES AND EXCLUSIVE DEALING Advertising provides a similar instance of an inter-retailer externality. Suppose that a car dealer pays for local TV advertising. Advertising expenditures increase demand, which in turn benefits all the car dealers selling the same car model, not just the dealer who paid for the advertising costs. That is, if there is a second dealer selling the same car in the same area, then part of the increase in sales goes to that second dealer. In this case, RPM would probably not do the job of correcting the externality, but other vertical restraints might. A specific alternative is to award ​exclusive territories​. This is a vertical restraint whereby each retailer is assigned a given territory that other retailers have no access to. For example, ​car manufacturers have an exclusive dealer in each European country. The German Fiat dealer, for instance, is not allowed to sell cars in France. Exclusive territories do the job because, if the dealer's advertising campaign is confined to its exclusive territory, then there is no longer an inter-retailer externality. Externalities may also occur between manufacturers. Continuing with the car dealerships example: it is often the case that manufacturers invest resources in training salespeople working in dealerships. Some of this training is specific to the manufacturer's cars, but some is more general ​(e.g., the art of car selling). If dealers were to work with more than one manufacturer, then there would be an inter-manufacturer externality: part of the training investment made by one manufacturer benefits the rival manufacturer. One way of solving this externality is to impose on retailers the vertical restraint of ​exclusive dealing​, whereby the retailer cannot work but with one manufacturer. Vertical restraints such as resale price maintenance, exclusive territories, and exclusive dealing allow upstream and downstream firms to internalize the effects of demand-increasing investments. Recap from slides​: ─ Sometimes upstream firms use another practice called ​exclusive territories​ → the intermediate good is sold at a given price, with a clause that the retailer can only operate in a given territory. ─ A variant, called ​exclusive dealing → ​the clause requires that the retailer sells exclusively that good. ─ Exclusive territories/dealing create local monopolies. (And local monopolies typically reduce output → manufacturers shouldn’t be happy). ─ Why these practices?​ ​Do they create or solve inefficiencies? ─ Similar situation as before: an industry with 1 manufacturer and 2 retailers. A retailer may make an investment that increases sales - benefits for the retailer. - benefits for the manufacturer, because of the sales increase. However, it may be that the retailer is not able to ​fully​ appropriate the benefits from the investments. ─ Example: - retailer 1 invests in advertising for the product sold by the manufacturer. - demand for the product increases, both for retailer 1 and 2. - consumers also buy from retailer 2; part of the benefits of the advertising investment are lost to retailer 1. ─ Anticipating the partial appropriability of the investment, retailer 1 does not make it. ─ Loss of a profit opportunity for both retailer 1 and the manufacturer. ─ An exclusive territory may prevent this situation: - if retailer 1 is the only one in the territory where it advertises the product, it fully appropriate the benefit of its investment. - retailer 1 thus chooses to make the efficient investment. Takeaways ⇢ Separation of manufacturer and retailer can lead to problems such as double marginalization and hold-up. ⇢ Vertical integration can alleviate these problems in some situations, but can also raise issues: - Competition softening. - Collusive behavior. - Agency problems (like the holdup). ⇢ Separation of manufacturer and retailer can lead to inefficiencies due to externalities. → Vertical restraints can alleviate these problems in some situations, but can also raise issues: - Limiting the benefits of competition. ⇢ No clear cut effect of firms’ policies in vertical industries, need a case-by-case approach. 13.3 PUBLIC POLICY In the previous sections, I suggested a variety of reasons why vertically-related firms may want to merge or establish contracts that regulate transactions between upstream and downstream players. Are vertical mergers and vertical restraints good for consumers? Should policy makers care about it? Unlike horizontal mergers, where the presumption is that the transaction leads to higher equilibrium prices, when it comes to vertical integration the effect is ambiguous. As we saw in Section 13.1, ​the double marginalization effect suggests that prices may decrease as a result of a vertical merger; but the competition softening effect suggests that they may increase. Moreover, a vertical merger may lead to foreclosure, which in turn may lead to higher prices in the long run. Whichever effect dominates is likely to depend on each industry's particulars. Consistent with this view, the US and EU non horizontal policies are generally more permissive than the corresponding horizontal merger policies. Similar to vertical integration, the effect of vertical restraints on consumer welfare is ambiguous. To summarize: Vertical restraints may facilitate collusion or lead to competitor foreclosure. To conclude: The number of effects involved in vertical relations is large (double marginalization, investment externalities, foreclosure, collusion, etc.). Moreover, the number of possible settings is also very large (one or many upstream or downstream firms; linear or nonlinear contracts; vertical integration or vertical separation; etc.). Given all this variety, it's likely that public policy with regard to vertical relations will continue to incorporate an important case-by-case component.