![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
Abstract
This paper develops an approach to analyzing an equilibrium in markets where firms can choose dual distribution to sell their products. Dual distribution involves a firm selling its product both through company-owned stores and through independently owned franchises. For a monopoly firm, the use of company-owned stores is assumed to play a number of roles. When the total number of markets is variable, an increase in company-owned stores can signal the quality of the product to potential franchisees, increasing the total number of markets served by the monopolist. Additional company-owned stores may also increase the royalty rate received by the franchisor, as well as increase demand in the local markets. There are limits, however, to the benefits of company ownership, called the “Penrose Effect.” For an equilibrium to exist, the monopoly firm must have no incentive to alter the the number of company-owned stores vis-à-vis franchised stores. The approach taken here yields a number of testable implications, which can form the basis of empirical tests of dual distribution.
Notes
I would like to thank Tom Ross for helpful comments on an earlier version of this paper. Helpful comments were also received from seminar participants at the Western Economic Association International 84th Annual Conference, Vancouver BC, and two anonymous referees. I am responsible for any errors or omissions.
1. In the management literature, for example Bradach (1997) and Dant, Perrigot, and Cliquet (2008), dual distribution is often described as a Plural Form of company operation, defined as the simultaneous use of company-owned and franchised units.
2. See Blair and Lafontaine (2005), Lafontaine and Shaw (2005), and Thompson (1994) for a discussion of the dynamics of dual distribution.
3. Two types of franchising have been identified. These have been called traditional franchising and business format franchising. Traditional franchising is observed when the manufacturer sells its product to distributors who resell to consumers with the franchisor earning profits from the markup charged to distributors. The second is business format franchising, in which the franchisor charges a royalty rate and a franchise fee to the franchisee to sell the product or service of the franchisor under a brand name. We focus on the business format franchise arrangement in this paper. For a further discussion of the differences in franchising arrangements, see Blair and Lafontaine (1995, 7–8).
4. Given the emphasis on the roles that brand name and quality issues play in much of the theoretical and empirical models of dual distribution, a model of dual distribution based on franchisor market power seems to be appropriate.
5. As outlined by Blair and Lafontaine (2005, 82–116), there is a close relationship between franchising, vertical integration, and vertical restraints. The models developed in the latter part of the paper can be viewed as resale price maintenance applied in a dual distribution setting.
6. See Caves and Murphy (1976), Mathewson and Winter (1985a) and Brickley and Dark (1987) for pioneering articles in this area.
7. They also outline a mixed strategy outcome when contracts are incomplete but self-enforcing in monitoring, where the strategy sets are the probability that the franchisee free-rides and the probability that the franchisor inspects the site.
8. A nice overview is found in Lafontaine and Shaw (2005).
9. Related to the issue of franchising is the literature that examines vertical integration and firm boundaries, for example Lafontaine and Slade (2007, 2013). These papers provide a summary of studies that examine the factors that lead to increased vertical integration on the part of firms. It is possible to interpret the move toward greater use of company-owned stores as increasing vertical integration. However, given the business format framework, I feel it is instructive to develop explicit models of the dual distribution system as well.
10. For earlier empirical work on franchising, see Lafontaine (1992), Lafontaine and Kaufmann (1994), and Minkler (1990).
11. Earlier work in the management literature suggested that successful franchise systems would ultimately become wholly owned. Oxenfeldt and Kelly (1968) argue that while older successful franchise systems will eventually become wholly company owned, the plural form (or dual distribution) is advantageous to a successful franchisor mainly during the infancy and adolescence of the enterprise and for exploiting selective markets. This has been termed the resource dependency theory by Windsperger and Dant (2006), who argue in favor of the property rights approach to the franchise decision, in contrast to the agency theoretic or transaction cost theories. Oxenfeldt and Kelly (1968) outline a number of factors leading franchisors to buy out franchisees and why some franchisees sell good businesses. For a nice overview of the management literature as well as a discussion of some results from the economics literature, see Dant, Perrigot, and Cliquet (2008).
12. See Edith Penrose (1959) for a presentation of this argument. Thompson (1994, 211) provides a number of reasons why franchising may allow a firm to externalize part of the management function, circumventing the Penrose constraint.
13. An interesting issue related to the Penrose Effect is the difference between the use of company ownership, which involves labor contracts with managers, and franchised outlets, which involves the use of contracts between firms. While not discussed in this paper, this issue is worth exploring. I thank an anonymous referee for raising this issue.
14. See Blair and Lafontaine (2005, 84–85).
15. Kalnins and Lafontaine (2004) in examining multi-unit ownership in franchising within the fast-food industry in Texas find that franchisees are more likely to be allocated ownership of a new unit closer to their existing units. They also find that franchisors use similar criteria when retaining units as company-owned as when choosing among franchisees.
16. In effect, I am assuming both types of outlets are assigned exclusive territories by the monopolist.
17. Given the assumption of local monopolies, the question of the optimal number of stores in relation to fixed costs is not an issue.
18. A number of franchisors also charge fees to cover costs incurred by the franchisor to advertise at the regional or national level. In order to highlight the central roles of the market penetration and the Penrose Effect, I abstract from this issue.
19. For a discussion of alternative royalty systems, see Blair and Lafontaine (2005, 100–101).
20. I examine the issue of an endogenous franchise fee in Section 4.
21. For a discussion of resale price maintenance in relation to franchising, see Blair and Lafontaine (2005, 174–201). In Section 4, I examine two models of resale price maintenance in a model of dual distribution. One model assumes that the monopolist must set a single price for both company-owned and franchised stores. This would be based on the franchise operation having a uniform price for its product. The second assumes that the monopolist is able to set separate retail prices for company-owned and franchised outlets. This price discrimination case can be based on the observation that one would expect prices for company-owned and franchised outlets to be similar but not necessarily identical, given the respective differences in objectives and management. In both cases, the monopolist also sets the number of company-owned stores.
22. See Bond (2009, 23) for a discussion of the role of franchise fees. According to Bond, “the franchise fee is a payment to reimburse the franchisor for the incurred costs of setting the franchisee up in business. It is function of the competitive franchise fees and the actual out-of pocket costs incurred by the franchisor.”
23. It is also possible to make the franchise fee endogenous by making it a function of the expected profits of the franchisee. Endogenous franchise fees are examined in Section 4, which considers franchisor price setting for both company-owned and franchise outlets in a dual distribution framework.
24. However, this is not the case for a single retail price, as shown in Section 4, where the issue of resale price maintenance is examined in the context of the dual distribution decision.
25. The distortion is due to the fact that the royalty is levied on the revenues of the franchise rather than its profits.
26. It is important to note that (5) can be also be obtained by choosing the optimal number of company-owned stores, ni, and the number of franchised outlets, nf, subject to a constraint on the total number of stores, n.
27. For example, the effect of an increase in the costs of franchised outlets, cf, on the price of company-owned stores is , which is negative in the fixed-market case, since
, and positive in the variable-market case,
, given
, which is the second-order condition corresponding to the first-order conditions given by equations (3) to (5).
28. In other words, depending on the characteristics of the firm and industry, the royalty rate is determined by competition among the set of franchise chains.
29. In addition, company ownership is also a commitment by the firm to the extent that the operation of company-owned stores is a sunk investment in the event of firm exit.
30. The individual market demands are functions of the number of stores operated by the monopolist, that is, and
, assuming
and
with the second derivatives,
,
.
31. See Mathewson and Winter (1985b, 1998) for nice overviews of the resale price maintenance literature.
32. Sometimes called the Availability Hypothesis, see Gallini and Winter (1983) for details of their Outlets model of RPM.
33. For an excellent overview of the RPM literature, including the Outlets Hypothesis, see Mathewson and Winter (1998).
34. The conditions governing the optimal number of company-owned stores for the case of endogenous local demands, as outlined in (10) to (13), are derived in Appendix II.
35. Resale price maintenance (RPM) is alternatively viewed as the upstream firm setting a price ceiling, a price floor, or the downstream price. I assume the monopolist sets the price for franchise outlets, which can be interpreted as setting the downstream price.
36. Note that when , the royalty rate drops out of (19), since it no longer plays a role in pricing.
37. It is easy to show that when ,
, that is, the uniform price is decreasing in the royalty rate if
, which is the most likely case.
38. Given the relative complexity of the first-order conditions (17) and (18), full comparative statics become unwieldy and not very definitive. Instead, for Propositions 4 and 5, I hold the distribution system constant, focusing solely on comparative statics with respect to prices.
39. This is clearly different from the single retail price case (19), where the retail price is a function of the number and distribution of company-owned stores versus franchised outlets.