AGENCY THEORY FRAMEWORK
Agency theory explains how to best organize relationships in which one party (the principal) determines the work, which another party (the agent) undertakes (Eisenhardt, 1985). The theory argues that under conditions of incomplete information and uncertainty, which characterize most business settings, two agency problems arise: adverse selection and moral hazard. Adverse selection is the condition under which the principal cannot ascertain if the agent accurately represents his ability to do the work for which he is being paid. Moral hazard is the condition under which the principal cannot be sure if the agent has put forth maximal effort (Eisenhardt, 1989).
The problems of adverse selection and moral hazard mean that fixed wage contracts are not always the optimal way to organize relationships between principals and agents (Jensen and Meckling, 1976). A fixed wage might create an incentive for the agent to shirk since his compensation will be the same regardless of the quality of his work or his effort level (Eisenhardt, 1985). When agents have incentive to shirk, it is often more efficient to replace fixed wages with compensation based on residual claimancy on the profits of the firm (Alchian and Demsetz, 1972). The provision of ownership rights reduces the incentive for agents' adverse selection and moral hazard since it makes their compensation dependent on their performance (Jensen, 1983).
A number of scholars have shown that the problems of adverse selection and moral hazard exist in the management of retail outlets (Rubin, 1978; Mathewson and Winter, 1985; Brickley and Dark, 1987). Outlet managers have an incentive to shirk and to misrepresent their abilities since the owner of the firm cannot easily differentiate the effect of manager behavior on outlet performance from the effect of exogenous factors (Carney and Gedajlovic, 1991). Franchising scholars have found that one way that performance of retail outlets can be enhanced is through the provision of residual claimancy that comes from franchising (LaFontaine and Kauffman, 1994).
However, the establishment of a hybrid organizational form does not eliminate all agency costs. Rather, the sale of residual claimancies on the profits of retail outlets creates a number of new agency costs, which come from the management of hybrid organizational arrangements. This paper argues that the survival of new franchise systems depends on the ability of the franchiser to economize on these agency costs. In the paragraphs below, I examine these agency problems and generate specific hypotheses to predict the effect of specific agency economizing activities on franchise system survival.
Ensuring Ownership Incentives
The argument made above suggests that the provision of residual claimancy on the profits of a retail outlet provides a strong incentive for agents not to engage in moral hazard or adverse selection and so enhances the performance of a retail distribution system (Brickley and Dark, 1987). By turning individuals into residual claimants on the profits of retail outlets that they have purchased, franchising reduces the incentive to shirk (Alchian and Demsetz, 1972).
However, the argument that franchising provides a superior incentive to fixed wage employment (Williamson, 1991) assumes that the franchised outlets are run by owner-operators. If the franchised outlet is run by a passive owner who hires a manager to run the outlet, then the beneficial effects of franchising on adverse selection and moral hazard are lost. Passive ownership simply recreates the agency problems that franchising is designed to overcome (Bradach, forthcoming).
Moreover, passive ownership increases agency costs over those which exist when the owner hires outlet managers. In place of the cost of managing the agency relationship between the owner and the outlet manager, passive ownership imposes the cost of managing the agency relationship between the franchiser and the passive owner and the cost of managing the agency relationship between the passive owner and the outlet manager. Since the survival of new franchise systems depends on the franchisers' ability to economize on agency costs, those new franchise systems which allow passive ownership should be less likely to survive. This argument leads to the first hypothesis:
H1: New franchise systems which permit passive ownership of franchised outlets are less likely to survive than are other new franchise systems.
The performance of a franchise system depends on the ability of the franchiser to ensure that all members of the system work to develop and maintain its brand name. The brand name is a signal to customers of the quality of the retail outlet (Norton, 1988b). As such, it is the major strategic asset that differentiates outlets in the franchise system from outlets in other franchise systems or independent businesses in the same industry.
However, the development and maintenance of the brand name is an externality problem. Each franchisee has an incentive to free ride off of the efforts of other franchisees to develop and maintain the brand name. Each franchisee can internalize the entire value of the savings from shirking on quality maintenance or on advertising and promotion, but can externalize the cost of that shirking.
For example, if a franchisee degrades quality, he will still be perceived as having the same quality as the rest of the franchise chain and will receive the same amount of benefit from the brand name (Rubin, 1978). However, he will reap the entire savings of the unexpended cost of maintaining quality (Brickley and Dark, 1987; Carney and Gedajlovic, 1991).
Given the existence of a free rider problem, the survival of a new franchise system depends on the franchiserís ability to minimize the agency cost of free riding. Research has shown that information about agent behavior minimizes agent free riding (Brickley and Dark, 1987). Information raises the risk of detection and therefore the likelihood that the franchisee will be terminated. Even if the information does not allow the principal to detect all free riding, it deters agents as long as the agents are risk averse (Eisenhardt, 1989). In addition, the information allows the principal to better differentiate suboptimal performance that is due to the agent's shirking from exogenous environmental shifts; and any information that improves this differentiation reduces the agent's incentive to shirk (Holstrom, 1979).
Obtaining information about franchisee behavior is a costly undertaking. Information on franchisee behavior is often gained through direct monitoring of franchisees. This activity includes verification that the franchisee is adhering to the contractual terms of the franchise agreement (Lal, 1990) through activities such as inspections of franchisee outlets by management personnel (Michael, forthcoming).
Efficiencies in outlet monitoring can reduce these monitoring costs in two ways. First, by increasing the ratio of outlets to monitoring personnel, per outlet monitoring costs are reduced. The primary mechanism for verifying franchisee compliance with franchiser terms is the unannounced audit. Since franchiser representatives can audit multiple outlets during a given work week, the cost per outlet of hiring a franchise consultant is lower if the franchiser has multiple outlets for each monitor on its payroll. Second, monitoring accuracy is improved by a high ratio of outlets per monitor. "Franchise systems with numerous units typically develop efficient routines for monitoring and measuring performance to assure profitability and uniform delivery of the franchise product or service. Such routines are enhanced by volume comparisons across units. The sheer volume of these comparisons can produce more educated routines for identifying and then managing the shirker (Huszagh et al, 1992: 8)." One would expect that the more monitoring efficiency a franchiser has, the more able the firm would be to monitor franchisees and so economize on agency costs. This argument leads to the second hypothesis:
H2: The greater the franchisers monitoring efficiency, the more likely the franchise system is to survive.
Monitoring costs increase with the distance between the principal and the agent (Norton, 1988b). The cost of monitoring franchised outlets is a function of the amount of time that monitors can spend on monitoring relative to other activities. The greater the distance between the monitor and the agent, the more travel costs must be added to the cost of monitoring (Rubin, 1978). Moreover, when distances are greater, monitors must spend a greater amount of time away from agents, increasing the latter's opportunity to free ride and raising monitoring costs (Martin, 1988). For these reasons, monitoring costs are lower when agents are geographically concentrated (Carney and Gedajlovic, 1991). Therefore, one would expect that the more geographically concentrated a franchise system's expansion is, the more the firm is able to economize on agency costs. This argument leads to the third hypothesis:
H3: The more geographically concentrated the expansion of the franchise system, the more likely it is to survive.
Franchiser Free Riding
Although much of the literature on franchising has focused on establishing franchisee incentives, franchisers must also be given incentives to motivate them to monitor franchisees against free riding (LaFontaine and Kauffman, 1994). As was described above, franchisees have an incentive to degrade the brand name and, in the absence of monitoring, will do so. The franchiser prevents franchisees from doing this by auditing the quality of franchised units, and by establishing and enforcing contractual provisions for advertising, training and outlet operations (Brickley and Dark, 1987).
However, these monitoring activities impose a cost on the franchiser. In the absence of a countervailing benefit, the franchiser has an incentive not to follow through on this monitoring obligation. Rubin (1978) has shown that if the franchiser were compensated entirely through these up-front franchise fees, he would have no incentive to monitor the system against agent shirking since he would receive no benefit from maintaining a high quality system.
The royalty that the franchiser receives provides an incentive to monitor franchisees against shirking (LaFontaine and Kauffman, 1994). To the extent that the franchiser receives ongoing royalties, he has an incentive not to default on his monitoring obligations. Defaulting on these obligations would lower the franchiserís return from establishing the franchise system since an unmonitored system would have greater franchisee free riding and lower sales.
Potential franchisees see the size of the royalty rate in the franchise contract as a measure of the franchiserís incentive to develop and uphold the brand name and the reputation of the franchise system. This reassures potential franchisees that the franchise system is organized in a way that will minimize agency problems (Lal, 1990). Therefore, one would expect that the higher the royalty rate in the franchise system, the more the firm is able to economize on agency costs. This argument leads to the fourth hypothesis:
H4: The higher the royalty rate, the more likely the new franchise system is to survive.
The use of master franchise agreements complicates the agency problems of managing a franchise system. Master franchise agreements are agreements "to grant the rights of development to an individual, who has no intention of operating an establishment himself or herself. Instead the intention is to recruit, train, and oversee the operations of individual franchisees in the area (Dandridge and Falbe, 1994: 41)."
One of the roles of the master franchisee is to enforce franchise agreements. The use of contractually employed monitors, as occurs with the use of master franchise agreements, increases agency costs. The introduction of a master franchise agreement requires that the franchiser be able to specify in the contract how master franchisees should enforce the franchise agreement. Given uncertainty and incomplete information about the ways in which franchisees could potentially shirk, it is difficult to identify the necessary enforcement behavior ex-ante. Without master franchise agreements, the codification of enforcement behavior is unnecessary. The franchiser can simply adopt appropriate monitoring routines as the situation dictates. However, with master franchise agreements, enforcement behavior must either be specified at the time of contracting or be foregone. Given the inability of franchisers to foresee all possible mechanisms for franchisee shirking, this requirement will reduce the ability to monitor franchisees and raise the opportunity for franchisee shirking. Therefore, the use of master franchise agreements should increase agency costs. This argument leads to the fifth hypothesis:
H5: Franchise systems which use master franchise agreements are less likely to survive.
Business format franchisers often require franchisees to pay for franchiser-specific investments such as materials, signs or building designs through an up-front franchise fee (Fladmoe-Lindquist, 1991). These franchise fees might have a positive effect on franchise system survival because they generate cash for the franchiser, which could mitigate cash flow constraints on survival.
However, agency theory suggests a stronger, countervailing effect of up-front franchise fees. Since franchiser-specific investments are worth more if the franchisee remains part of the franchise system than if he does not, these assets generate quasi-rents (Brickley and Dark, 1987; Carney and Gedajlovic, 1991). Mathewson and Winter (1985) explain that a franchiser can appropriate the value of these quasi-rents by precluding the franchisee from using these assets before the end of their useful lives. Moreover, the quasi-rents create an incentive for the franchiser to claim that the franchisee has violated the franchise agreement so as to permit the franchiser to appropriate these rents (Klein et al, 1978).
Quasi-rents place limits on the size of the up-front franchise fee that franchisees are willing to pay. Franchisees want to minimize these fees so as to recoup the cost of the up-front franchise fee during the life of the franchise contract and preclude the possibility of franchiser appropriation (Combs and Castrogiovanni, 1994). The greater the amount of the up-front franchise fee, the greater the divergence between the franchiser and franchisee over the required return on the outlet's assets during the period of the initial franchise agreement. For this reason, high franchise fees raise ex-ante bargaining costs (LaFontaine, 1992). Therefore, franchise fees should be positively associated with agency costs. This argument leads to the sixth hypothesis:
H6: The higher the initial franchise fee, the less likely the franchise system is to survive.
Length of the Agreement
Increasing the term of an agreement between the principal and agent reduces agency problems (Levinthal, 1988). Longer term agreements reduce agent shirking for three reasons. First, long time horizons provide an incentive for principals to invest in gathering information about agents' behavior (Eisenhardt, 1989). Increased information about the agent enhances the likelihood that the principal will detect shirking and reduces the incentive for the agent to shirk. It also helps to disclose the agent's type and thereby reduces the adverse selection problem. Second, Holstrom (1979:90) explains that over time the patterns of environmental effects on performance become clear, allowing the principal to more precisely separate exogenous environmental effects on performance from the agent's shirking behavior, making agent moral hazard more difficult. Third, the longer the time horizon of the agreement, the lower the agent's incentive to shirk or engage in perquisite-taking (Jensen and Meckling, 1976). Long time horizons increase the amount that the agent has to gain by proper behavior relative to the amount he has to gain from shirking (Williamson, 1991). Because agents will not shirk if the net present value of future earnings from not shirking exceeds any gains from shirking, long-term agreements reduce franchisee shirking (Klein et al, 1978). Therefore, longer term franchise agreements should reduce agency costs. This argument leads to the seventh hypothesis:
H7: The longer the initial term of the franchise contract, the more likely the franchise system is to survive.
Information transfer in a non-hierarchical setting is problematic because of the agency problem of moral hazard. If the principal has imperfect information about the agent's ability to perform the task that is demanded of the agent, the principal will have difficulty ensuring that the agent has performed that task (Barzel, 1989). This lack of measurement ability will provide the agent with an incentive to shirk on the proper performance of that task (Chi, 1994). In the context of franchising, this means that if the franchiser cannot be sure that the franchisee is performing the job of managing a local retail outlet, the franchisee will have an incentive to shirk on his efforts to manage the outlet. Michael (forthcoming) explains that the decisions that the franchisee makes about local demand, real estate, prices, and labor markets are important to the success of franchising; and it is difficult for the franchiser to ascertain whether a potential franchisee has made the right decisions about these things.
"Explicit long-term contracts can, in principle, solve [these] problems, but... they are often very costly solutions . . . since every contingency cannot be cheaply specified in a contract or even known and because legal redress is expensive (Klein et al, 1978: 303)." The cost of writing long-term contracts depends on the complexity of the franchise concept. Complex franchise concepts are ones that require both franchisers and franchisees to undertake a large number of different activities. The more complex the franchise concept, the more difficult and costly it is for the principal to specify he agent's required behavior under all contingencies (Eisenhardt, 1989). Moreover, the more complex the concept, the more costly it is for the principal to ensure that the agent is properly undertaking the appropriate activities (Anderson, 1985; Eisenhardt, 1985). Therefore, the more complex the franchise concept, the higher the agency cost of operating the business through a hybrid structure (Klein et al, 1978). This argument leads to the eighth hypothesis:
H8: The more complex the franchise concept, the less likely the franchise system is to survive.
Top of Page
Previous Page Main Page Next Page
Return to 1996 Topical Index
© 1997 Babson College All Rights Reserved
Last Updated 4/24/97 by Jo Ann Mathieu
To sign-up for the Center for Entrepreneurial
Studies' publication lists,
please register with the Entrepreneurship WebTeam.