Effects of Contractual Characteristics on Franchisees’ Compliance and Opportunism in Franchise Relationships

Franchisors grant specific stores a license to use their intellectual property, such as the company's name. As a result, a principal-agent relationship is established. In this relationship, to ensure the franchisee’s cooperation, the franchisor might oversee their behavior or offer incentives based on their operational performance (Eisenhardt, 1989).

In the franchise business, the franchisor generates earnings from the royalties or fees they receive from the franchise system. However, as the franchisee makes a profit from the franchise store (Shane, 1998), they are usually incentivized to free-ride and maximize profits for their individual store, potentially harming the entire franchise system (Brickley & Dark, 1987). Thus, incentives and/or monitoring are necessary to prevent free-riding.

According to the principal-agent theory, the franchisor employs two control mechanisms: the franchise contract, which is used before the fact, and monitoring, which is used after the fact (Bergen et al., 1992; Jensen & Meckling, 1976).

However, in practice, uncertainty exists and it is not possible to precisely determine damages resulting from the franchisee’s breach of contract or deviant behavior. As a consequence, it is not feasible to draft a comprehensive franchise agreement (Anderson & Dekker, 2005; Klein, 1980). Therefore, in addition to the franchise contract, the franchisor carries out monitoring to ensure that the franchise store realigns and fits within the objectives (Davies et al., 2011; Kashyap et al., 2012).

The franchisor plays a crucial role in reducing the franchisee’s opportunistic behavior and increasing compliance by utilizing a suitable combination of two control mechanisms - the franchise contract and monitoring.

Existing research suggests that drafting a perfect franchise contract is costly and practically impossible (Klein, 1980). Additionally, monitoring can result in resistance or hostility from the franchisee. This can lead to a decrease in compliance or induce opportunistic behavior (Brehm, 1966). Moreover, franchisees often seek to preserve their autonomy through alternative means, especially when facing external pressures such as franchise contracts and monitoring. As a result, franchisors must identify the most effective approach to enhance compliance and reduce opportunistic behavior, taking into account the associated costs and potential consequences of each method. In this regard, preemptive control systems have an impact on the need for and extent of subsequent monitoring (Antia & Frazier, 2001).

Developing and implementing the details of the franchise contract requires significant effort and incurs substantial costs for the franchisor (Casson, 1991). Regarding monitoring, there are costs associated with constantly hiring and training employees, as well as conducting site inspections.

To address this issue, the franchisor must find avenues to increase compliance and reduce opportunistic behavior through the appropriate application of contracts and monitoring, without incurring excessive costs (Casson, 1991).

When drafting the contract, it is crucial to include content that enables the franchisee to comply with the franchisor’s policies and systems. The level of completeness and one-sidedness of the contract arises from these characteristics.

The completeness of the contract refers to the extent to which the relevant provisions are clearly stated (Gong et al., 2007). Contracts can take many different forms - from a simple document outlining the rights and obligations of the parties involved, to a comprehensive agreement that anticipates future emergency scenarios leading to a breach, and specifies the responsibilities that arise from such a breach (Wuyts & Geyskens, 2005). When there is a detailed contract that outlines the franchisee’s obligations, role ambiguity is reduced, and the chances of non-compliance or opportunistic behavior decrease (Argyres et al., 2007).

Contractual one-sidedness is defined as the extent to which some provisions favor the franchisor (Klein, 1980). In case the punishment for breaking or evading the contract results in more drawbacks than advantages, the franchisee is encouraged to avoid opportunistic behavior and conform to the agreement (Klein, 1980). Conversely, if the contract’s one-sidedness is significant, the franchisee’s opportunism and resistance may increase (Clee & Wicklund, 1980). Moreover, one-sidedness reveals the franchisee's mistrust, and the franchisee may question the franchisor's motives (Jap & Ganesan, 2000). Thus, in franchise agreements that require continuous transactional relationship and cooperation, one should be cautious of the negative impact of one-sidedness (Clee & Wicklund, 1980).

Compliance of the franchisee refers to the extent to which they adhere to the policy and procedures set out by the franchisor (Dalstrom & Nygaard, 1999). It is a crucial aspect of franchising as it requires consistency and standardization (Williamson, 1975).

The level of compliance exhibited by the franchisee has a positive correlation with the number of franchise stores owned by the franchisor and consistency in the system. This correlation is established through research by Bradach (1998) and Davies et al. (2011). On the other hand, a franchisee's lack of resources or poor understanding of the required behavior was found to have a negative correlation with compliance (Dickey, 2003).

Opportunistic behavior is characterized as actions that lack fairness and honesty, both of which should form an integral part of any transaction. This refers to behavior that appears to follow the contract from an external perspective, but in reality, it contradicts the contract or conceals crucial information. Additionally, opportunistic behavior is characterized by instances when parties refrain from undertaking mutually beneficial work within a transaction (Williamson, 1975). This phenomenon occurs due to incomplete contracts and arises when franchisees use factors that could not have been predicted or were not specifically described in the contract. Opportunistic behavior contributes to transactional costs (Williamson, 1975) and reduces the effectiveness of the franchise system (Shane, 1996).

To maintain the uniformity of the franchise system and achieve high performance, the franchisor should effectively control opportunistic behavior by franchisees (Dwyer et al., 1987).

Compliance and opportunism are not mutually exclusive behaviors, and they can coexist. However, despite following the policy or procedures of the franchisor, a franchisee may avoid responsibility for certain aspects that are not explicitly outlined (Bergen et al., 1992). In addition to opportunistic behavior, several reasons such as lack of resources, peculiarities of a certain location, and lack of understanding on required behavior contribute to why a franchisee may not follow the franchisor's policy (Williamson, 1975). Thus, it would not be appropriate to infer that an increase in compliance will always lead to a decrease in opportunism, or that a decrease in compliance will always lead to an increase in opportunism.

In order for franchise operations to succeed, franchisees must comply with the policy (Williamson, 1975). Given that opportunism can undermine the effectiveness of the franchise system (Shane, 1996), franchisors need to regulate both compliance and opportunistic behavior. Thus, understanding how the franchisor's regulation mechanism affects the franchisee's behavior requires considering both compliance and opportunistic behavior, not just one or the other (Kashyap et al., 2012).

The franchisor aims to regulate in advance to minimize or eliminate costs that may arise after contract closure (Bergen et al., 1992). As franchise contracts are incomplete, franchisors strive to achieve and calibrate objectives by supplementing monitoring after contract closure. Yet, as external pressures on the franchisor's regulatory system augment, franchisees attempt to realign to increase their autonomy (Clee & Wicklund 1980). Thus, infinitely increasing monitoring is not a solution. Consequently, when monitoring operations within the necessary parameters, the franchisor must consider, as stated earlier, the franchise contract’s particularities. In this sense, the franchise contract has an influence on dictating the necessity of monitoring after the contract has been closed (Antia & Frazier, 2001). If the contract is incomplete, the franchisee may exploit any loopholes or unforeseen circumstances at the contract's time to maximize profits and avoid complying with franchise contracts (Klein, 1980). Therefore, the franchisor should increase monitoring to address the aforementioned behaviors of franchisees.

In contrast, when the contract is highly comprehensive, franchisees understand the content and limitations of the projected actions, resulting in increased compliance (Casson, 1991). This, in turn, reduces the possibility of opportunistic behavior. As a result, incurred costs and monitoring requirements for franchisors decrease. Hence, if the franchise contract is more complete, the franchisor is expected to conduct less monitoring.

The one-sidedness of contracts is an attempt by the franchisor to increase franchisee compliance and reduce opportunistic behavior by specifying in the contract that breach or avoidance of the contract will be more detrimental than beneficial. Therefore, unilateral clauses are chosen to minimize transaction costs and are inherently unfair (Klein, 1980).

Consequently, there is a possibility that the franchisee may have a negative reaction to the clauses or there may be an increase in the franchisee’s opportunistic behavior. And it will be necessary for the franchisor to increase monitoring. Moreover, if it is a clause with high one-sidedness, it is more likely to be a clause that is essential for the unity of the franchise system (Huang & Sarigöllü, 2012; Yoo et al., 2000). It is probably essential for the franchisor to confirm and guide the franchisee whether the franchisee is properly following certain clauses in order to maintain the unity and awareness of the franchise brand. Therefore, if there is a high level of contractual one-sidedness, we expect a high level of monitoring.

On the one hand, monitoring provides important feedback and useful guidance to the franchisee (Heide et al., 2014). On the other hand, it is a double-edged sword in that it signals distrust to the franchisee or a violation of the franchisee's autonomy, thus increasing the franchisee's psychological resistance (Brehm, 1966; Heide et al., 2007). Furthermore, monitoring changes depending on the situation in which it occurs (Frey, 1993; Heide et al., 2014). If monitoring is perceived as a guidance function from the franchisee’s perspective, compliance may increase. On the other hand, if monitoring is perceived as a sign of distrust, compliance may decrease (Brehm, 1966; Heide et al., 2014).

However, according to existing research, behavioral monitoring increases opportunism (Heide et al., 2007), because when behavioral monitoring occurs, franchisees believe that their autonomy is being violated, which creates resistance and animosity toward the franchisor, which ultimately leads to opportunistic tendencies. Therefore, behavioral monitoring evaluates the behavior and management procedures of the franchisee and its employees. In this process, monitoring provides strict guidelines for what is considered to be within the franchisee’s autonomy, such as how work is performed. As a result, franchisees feel that their autonomy is being violated (Brehm, 1966; Heide et al., 2007).

Consequently, when behavioral monitoring is increased, the franchisee becomes resistant to the franchisor’s intrusion, and accordingly, it becomes more likely that the monitoring will be accepted as a violation of autonomy or a sign of mistrust rather than as a positive aspect (a function of providing guidelines). Thus, the stronger the behavioral monitoring, the more resistant the franchisee becomes to the monitoring. As a result, it is expected that the likelihood of compliance will decrease and the likelihood of opportunistic behavior will increase.

In the case of output monitoring, compared to behavioral monitoring, it is more likely to be accepted as a guideline that provides perspective for the following reasons.

First, output monitoring allows franchisees to perceive what goals to focus on in order to be successful (Kashyap et al., 2012). Second, output monitoring measures the results of franchise store management rather than the behaviors and procedures of franchisees and employees, allowing franchisees to proactively choose how to achieve results (Anderson & Oliver, 1987). As a result, there is almost no likelihood of intervention or interference that violates autonomy or the right to self-control, so there is little likelihood of franchisee resistance (Heide et al., 2007). And there is no reason for the franchisee to feel uncomfortable with output monitoring (Kashyap et al., 2012).

Consequently, in the case of output monitoring, the franchisee believes that one can receive the guidelines from the franchisor and operate the franchise business based on one's discretion. The franchisee follows the franchisor's policies or procedures. The possibility of opportunistic behavior is reduced (Heide et al., 2007).

Because each method has its own strengths and weaknesses, and because the two methods can complement each other, the two methods are used in parallel. First, output monitoring has the positive aspect of providing guidelines. However, in cases where it is not used concurrently with behavioral monitoring, it may give the franchisee the impression that the franchisor is not concerned with the franchisee. Because behavior monitoring is conducted face-to-face, there is a greater chance of interaction with the franchisee. There is the benefit of learning practical information, such as the operating conditions of the store. For this reason, franchisors use the two methods simultaneously to take advantage of the different control mechanisms of each method (Brown et al., 2000; Kashyap et al., 2012). Furthermore, when there is no meaningful metric for the results obtained through output monitoring, it is necessary to supplement output monitoring with behavior monitoring (Eisenhardt, 1985).

When behavior monitoring and output monitoring are conducted simultaneously, the franchisor can measure the franchisee's behavior and its outcome at the same time, and a consistent control system is possible (Anderson & Oliver, 1987). Therefore, if there is a high level of output and behavior monitoring carried out simultaneously, the franchisee believes that the franchisor removes uncertainty and instability factors and shows a proper behavioral direction to achieve appropriate goals (Anderson & Oliver, 1987), and it is expected that the negative perception that it is a violation of autonomy would be relatively weakened.

In addition, because there would be more detailed acquisition of the franchisee's management practices, the information asymmetry problem would be alleviated and the possibility of opportunistic behavior would be reduced.

Therefore, in cases where behavioral and output monitoring occur simultaneously, it is expected that compliance would increase and opportunistic behavior would decrease (Kashyap et al., 2012).

In the case where the objective between franchisee and franchisor are in discord with each other, the franchisee tries to behave according to his interest (Bergen et al., 1992), and as a result, the franchisee tries to benefit through opportunistic behavior by abusing on details that are not explicitly written in the contract (Klein et al., 1990). Therefore, the franchisor increases monitoring to prevent or reduce opportunism. However, the franchisee's response to increased monitoring may differ depending on the purpose and nature of the monitoring.

For example, from the franchisee’s perspective, if the franchisee desperately needs guidance from the franchisor, the focus will be more on providing information and compliance will increase. On the other hand, if the franchisee does not need guidance from the franchisor, he will most likely perceive the franchisor’s monitoring as a signal of mistrust or a violation of his autonomy.

But monitoring is done to realign the franchisee’s goals with the franchisor’s goals (Jensen & Meckling, 1976), and in cases where Goal Incongruity is high, the franchisee and the franchisor have fundamentally different goals (Song et al., 2000) and there is little motivation or enticement to cooperate with each other (Dyer & Song, 1997), so monitoring becomes an unnecessary and disruptive behavior to achieve the franchisor’s goals, not the franchisee’s own goals. Therefore, in cases where goal incongruence is high, it is expected that the decrease in franchisee compliance due to franchisor monitoring will be higher.

Next, we consider the moderating effects of goal incongruence on monitoring and opportunistic behavior.

The reason for franchisees to avoid franchise contract obligations is that the benefits of avoiding obligations are greater than the benefits of fulfilling obligations (Klein, 1980), and fulfilling obligations requires additional expenditures of time, effort, and other resources (Bergen et al., 1992).

Therefore, in cases where goal incongruence is high, the franchisor's goals are not beneficial to the franchisee and become a matter that only requires additional effort on the franchisee's part, thus reinforcing the franchisee's impulse to act for his or her own benefit and avoid achieving the goals set by the franchisor. Thus, the franchisee has a negative reaction to the monitoring conducted to achieve the franchisor's goals, and opportunistic behavior to avoid the monitoring is strengthened.

To measure the completeness of the contracts, we followed the method presented in the study of Ghosh and John (2005) and Kashyap et al. (2012). At this point, in order to account for the different characteristics of each industry, we collected the clauses for contract binding, contract maintenance, contract termination, and post-termination obligations in each contract from each industry, and we created the superset list for each industry. Then, two people evaluated the existence of each clause in each contract and counted the number of clauses and found the total sum, and we calculated the ratio of the superset list of each industry to the number of clauses in each brand.

The one-sidedness of the contracts went through the same procedures as the completeness. At this point, in order to take into account the different characteristics of each industry, we collected the clauses that were one-sided in favor of the franchisor and the clauses that imposed costs beyond the usual standards in the contracts, and we prepared the superset list. Then, two people evaluated the presence of each clause in each contract and counted the number of clauses and found the total sum, and we calculated the ratio of the superset list of each industry to the number of clauses in each brand.

This category was based on Niehoff and Moorman’s (1993) research. We considered the metrics of “checking to see if the franchise business is being run efficiently,” “meeting to discuss the actions that need to be taken to achieve the goals of the franchise business,” and “meeting to discuss the customer base of the franchise business” to be included in a total of three questions and measured on a seven-point Likert scale.

This category was based on the research of Niehoff and Moorman (1993), but adapted for the purpose of this research. The metrics were conceived as the franchisor monitoring the franchisee for “consumer response to new merchandise,” “expansion of customer base,” and “sales performance,” and were incorporated into a total of three questions and measured on a seven-point Likert scale.

The compliance metrics were based on Tyler and Blader’s (2000) research, but adapted for this study. The metrics were based on the franchisee’s perception of whether the franchisee “follows the procedures and applicable rules of the franchise business,” “follows the policies of the franchisor,” “tries his or her best to follow the instructions of the franchisor,” and “follows the rules of the franchisor in detail,” incorporated into a total of four questions and measured on a seven-point Likert scale.

In this study, we built on the research of Brown et al. (2000) to determine the metrics of franchisee responses in situations where franchisees perceive the franchisor’s orders to be unfair. The metrics were defined in three categories - whether the franchisee “ignores the franchisor’s policies to achieve his goals,” “violates the franchise agreement to further his interests,” and “makes outward promises but does not behave as promised”. And in cases where franchisor support was poor, the metrics were defined into four categories - the franchisee “exaggerated their difficulties to get more support,” “distorted the truth,” and “inevitably lied to further their interests”. The total of seven categories were measured on a seven-point Likert scale.

In this research, we followed the research of Xie et al. (2003), we defined goal incongruity as differences in goals between franchisor and franchisee, standards of decision making, urgency of issue, and process of decision making. The metrics were defined in four categories - among different franchise stores, “the long and short term goals are different,” “the standards of decision making for franchise store operations are different,” “the priority for urgency of issues,” and “the operational decision-making process is different”. These four categories were measured using a seven-point Likert scale.

From the franchisee’s perspective, the initial start-up costs and the franchisor's proprietary assets prevent attrition and become a factor that makes the franchisee more dependent on the franchisor. At this point, brand equity increases as the store size increases. Therefore, the initial start-up costs and the size of the store will function as a conversion barrier that continues the preexisting relationship (Dant, 1996), and because of this, the franchisee becomes more dependent on the franchisor. We can predict that compliance increases for franchisors and opportunistic behavior decreases.

Furthermore, for franchisees with high sales, they may have a positive attitude toward the franchisor’s policy, and there's a high likelihood of following the policy. And because the importance of the franchise outlet (or the dependence on the franchise outlet) increases, we can predict that the compliance for the franchisor increases and the opportunism decreases (Dant & Gregory, 1998). For this reason, in the franchise store aspect, the cost of starting the business, the size of the store, and the number of sales were included in the control variables.

Also, in the case of franchisees who also operate a store of another franchise brand, the share of the surveyed brand in their total asset portfolio will be much lower than that of a franchisee with only one brand, so we expect the dependence on a particular brand to be much lower. Therefore, we included the number of other brand stores owned in the control variables.

On the other hand, assets, sales, total number of stores, and age of the company are measures of the size of the franchisor. As shown below, it may influence the compliance and opportunistic behavior of franchisees.

As the franchisor becomes larger, the amount of resources that can be disposed of increases, and based on the economy of scale, the ability to control the franchise outlet per unit increases, compliance will increase, and it will influence to decrease opportunism (Brickley et al., 1991). On the other hand, the size of the franchisor signifies the value of the franchise brand, and based on several previous researches, as the size of the business becomes more significant and the age of the firm increases, the franchisor charges a higher royalty (Lafontaine, 1992). However, in the case of the franchise industry in the Republic of Korea, the franchisor's main source of income is not the royalty, but he profits from the item margins of essential items that the franchisor supplies to the franchisee. Therefore, just as the royalty ratio increases when the reputation capital increases due to the increase in brand scale, we can expect that the ratio of the item margins of essential items will be higher when the reputation capital increases. However, based on the reality that many conflicts stem from the disagreement between franchisor and franchisee on essential items, the increase in item margins due to the increase in franchisor scale may not necessarily have only positive effects on franchisee behavior, but also negative effects. In addition, according to resource dependence theory, as the size of the franchisor increases and the age of the firm increases beyond a certain point, there is a tendency to reduce the proportion of franchised stores and become more dependent on company-owned stores (Dant et al., 1996). The large franchisor is likely to reduce the level of resources invested in existing franchised outlets. For these reasons, we judged that the size of the franchisor is likely to affect the compliance and opportunistic behavior of the franchisee, and we conducted an analysis of capital, sales, total number of stores, and firm age based on the size of the franchisor.

For the variables in this research that ask about the level of perception, which are behavior monitoring, output monitoring, compliance, opportunistic behavior, and goal incongruity, we considered previous research. We constructed the questions so that franchisees would not feel alienated when answering the questions. To do this, we conducted intensive interviews with franchisees in eight different industries who were actual respondents for one month starting June 1, 2019, to validate and strengthen the survey content and structure. If there were any questions that were requested to be changed during the intensive interview, we reflected on the requests and refined the questions in the survey form.

In this research, in order to improve the representativeness of the research sample and the efficiency of the survey, we tried to select eight industries from the 43 industries that the Korean Fair Trade Commission considers as Level II specifications.

Accordingly, we selected eight industries (education, laundry services, automobiles, hair salons/beauty salons, fried chicken restaurants, pizza restaurants, coffee shops, convenience stores) that currently have many existing conflicts or have a high proportion of franchise stores within the entire industries. According to the Korean Fair Trade Commission's Level I specification of industries in 2017, the number of stores per industry is as follows: the restaurant sector is 118,780, the service sector is 71,971, and the wholesale and retail sector is 57,339. Among them, the proportion and the number of stores held by the eight industries within this Level-I specification are as follows.

Within the restaurant sector, fried chicken restaurants were 20.9% (24,803 stores), coffee shops 12.6% (14,924 stores), pizza restaurants 5.6% (6,699 stores). Within the service sector, education is 54.9% (39,537 stores), automotive 11.0%(7,899 stores), hair salon/beauty salon 6.3% (4,542 stores), laundry 6.1% (4,372 stores). Finally, we selected only the convenience store business for the wholesale and retail sector, and the numbers were 57.1% (57,339 stores).

In this research, a total of 892 surveys were used, and considering that the sample is divided into two dimensions, franchisor and franchisee, we distinguished each sample by dimension and showed the general characteristics by franchisee and franchise brands.

First, in the case of the franchisee sample used in this research, the industry with the most samples was convenience stores with 198 people (22.2%), followed by coffee shops with 142 people (15.9%), education with 138 people (15.5%), and chicken restaurants with 121 people (13.6%).

In terms of the size of the franchise store, it was counted that the less than 10 pyeong was 96 people (10.8%), larger than 10 pyeong but smaller than 20 pyeong was 400 people (44.8%), larger than 20 pyeong but smaller than 50 pyeong was 298 people (33.4%), larger than 50 pyeong was 98 people (11.0%), so we found that we had equitable distribution of the sample of small to medium-sized stores and medium to large stores.

In addition, the length of operation, the total cost to start the business, the work before starting the business, the size of annual sales, and the success or failure to achieve the target profit number of the current affiliated brand are as shown in <Table 1>.

On the other hand, the annual sales of the franchise stores, the initial franchise amount, the number of stores, the age of the company, and the sample brands of the franchisor's industry that were examined in the disclosure document information are shown in <Table 2>.

As seen above, this research attempted to contribute to the generalization of research by selecting industries that held more than the majority of the particular level-II specification or were representative of the particular level-II industry.

The collection of franchise contracts can distinguish the data collection of this research to measure contractual completeness and one-sidedness, the collection of disclosure documents and surveys to extract control variables, and the collection of surveys to derive perception variables.

First, we obtained the franchise contract materials and disclosure documents with the help of the Korea Fair Trade Commission. As for the franchise contracts, we were able to obtain a total of 634 franchise contracts that were registered between 2018 and 2019. And we were able to obtain disclosure documents for 6,053 brands that were registered in 2018.

On the other hand, the survey collection was obtained for 75 days from July 1, 2019, taking into account the industry share and the share of medium and large brands. The survey was conducted in 1,004 franchise stores across the Republic of Korea. Additional surveys were conducted for industries that did not meet the proportional allocation, and a total of 1,108 surveys were collected. After thoroughly explaining the research contents and objectives to the respondents, we conducted the surveys through a professional research company. The survey methods were conducted through personal visits and telephone surveys, and the survey agents conducted the survey by checking whether the respondent understood the content of the survey when answering the survey.

The final sample selection of this survey was conducted with a population parameter of 110 sample brands collected from the survey. First, from the survey samples of the 110 brands, we selected 58 brands as the final research object, which coincided with the franchise contracts of the 634 brands provided by the Korean Fair Trade Commission. We then coded the variable corresponding to the control variable within the same brand sample from the disclosure document data. Accordingly, we were able to obtain 892 samples with intact data for the affiliated brand's franchise contract and disclosure document among the survey sample of 1,108 franchisees. The control variables were extracted in the standards shown in <Table 3>, except for industries, which were dummy variables.

For this reason, more than two coders coded and derived the franchise contract variables (one-sidedness, completeness), disclosure document variables used as control variables, and variables that included the franchisee's perceived responses. Of these, the variables related to the franchise contract or disclosure documents are secondary sources, but the variables measuring the franchisee’s perception require reliability and validity certification.

In this research, we used five constructs of behavior monitoring, output monitoring, franchisee compliance, opportunistic behavior, and goal incongruence. To test the internal consistency of these concepts, we used the SPSS 24.0 reliability certification program.

In the initial reliability analysis, we administered all the observed variables that correspond to the constructs and we limited the administration of variables until we had a Cronbach's a value that was greater than 0.7 and we secured a coefficient greater than 859 for all the constructs. The results are presented in <Table 4>. The reliability coefficient was shown to be between .859 and .910, so we can say that we have a reasonable reliability for each variable.

In order to analyze the validity of the constructs, we conducted analyses of Convergent validity and Discriminant validity through confirmatory factor analysis. The results are presented in <Table 5>.

The CR and AVE values were used to confirm the convergent validity of the variables used in this research. Normally, if the AVE is greater than 0.5, if the CR values is greater than 0.7, it is considered as acceptable levels (Fornell & Larcker, 1981).

According to the analysis, the load level of each factor was more than 0.5, and the AVE value, which corresponds to opportunism, meant the minimum value of 0.525, so we can say that all variables followed the AVE value. In addition, the CR value must be greater than 0.7. But in the case of output monitoring, it had a level of 0.695, which does not meet the CR value standard, but we confirmed that they meet the standards for the other variables.

Next, judgment validity is defined as the analysis of the differences between different constructs, if each independent variable has a low correlation with different variables, then we can say that there is high judgment validity. At the same time, we confirmed the judgment validity by comparing the square root of the highest value and the lowest value of the correlation between variables. According to <Table 5>, the variables with the highest square root values of correlation were behavior and output monitoring with values of 0.576. On the other hand, the lowest AVE was 0.525, and it was close to the value of 0.576, so we could accept the judgment validity to some extent.

The results of regression analysis, using the CMP (conditional mixed process) Roodman (2009) designed within the Stata 14.0 statistical software are presented in the <Table 6>.

First, it showed that the higher the contractual completeness, the lower the level of franchisor's output monitoring perceived by the franchise store, but for behavior monitoring it does not have statistically significant result. Therefore, Hypothesis 1(b) was supported, but Hypothesis 1(a) was rejected. This is different from Kashyap et al. (2012), who found that the completeness of franchise contracts decreases behavior monitoring and has no significant effect on output monitoring. This result may be due to the fact that in Korea, the franchisor’s main source of income is the distribution margin rather than royalties, and behavior monitoring was considered to be more than just managing the franchisees, but also including guidance and support for the franchisees. Suppose the franchisee purchases essential products through channels other than the franchisor. In this case, the franchisor’s profit is reduced, so even though it involves additional expenses for the franchisor, it is considered more beneficial for the franchisor to prevent the purchase of essential products through other channels, so in this sense, the franchisor could conduct behavior monitoring. In the case of Republic of Korea, the franchisors who have the margin of providing goods and raw materials as their main source of income is 80.6%, and the tendency to do so increases with the size of the franchisor, so we can understand this phenomenon in this context (Ministry of Trade Industry and Energy, Korea Chamber of Commerce and Industry, Korea Franchise Association, 2018). In addition, the monitoring of franchise stores has the aspect of control and the aspect of supporting the franchise store. Therefore, in order to function properly as a franchisor, they cannot reduce the monitoring just based on the cost.

Second, when the contractual one-sidedness was high, the franchisee perceived the franchisor's behavior monitoring to be high. However, there was no statistically significant effect on output monitoring. Therefore, Hypothesis 2(a) was supported, but Hypothesis 2(b) was rejected. This is consistent with the research of Kashyap et al. (2012), which states that one-sidedness of contracts increases behavioral monitoring. The franchisor defines certain content in a one-sided way because the said content is important, and it is more important to prevent the outcome than to punish the outcome. Thus, by writing into the contract the prohibition of certain behaviors and the guidelines for punishment when the behavior is performed, and by conducting behavioral monitoring of contract compliance at the appropriate time, the franchisor can prevent the behavioral outcomes from occurring.

Third, when behavioral and output monitoring were conducted separately, we found that they did not have a meaningful impact on franchisee behavior, thus rejecting all of Hypotheses 3 and 4. These results are consistent with the findings of Kashyap et al. (2012).

Finally, when both outputs and behaviors were high, compliance increased and opportunistic behavior decreased. Thus, hypotheses 5(a) and 5(b) were both supported. These results are consistent with Kashyap et al. (2012).

The results of hypotheses 3 to 5, as shown above, indicate that the context of monitoring is important, and in the reality of Republic of Korea, monitoring should be more active. The fact that franchisee compliance did not increase or opportunism did not decrease due to output monitoring, and the fact that franchisee compliance did not decrease or opportunism did not increase due to behavioral monitoring does not show that franchisee had a negative reaction to behavioral monitoring. Rather, it shows that the franchisee wants a more active type of monitoring beyond that seen in output monitoring. Thus, output monitoring alone is not enough for the franchisee. It can be assumed that the franchisee wants the franchisor to pay more attention to the franchised stores through behavioral monitoring. We can confirm this point by the fact that the franchisee does not have a negative reaction to the behavior monitoring, but is not satisfied with the output monitoring. When both monitoring is done simultaneously, compliance increases and opportunism decreases.

In the case of high goal incongruence, compared to low goal incongruence, franchisee compliance with franchisor behavior monitoring was shown to decrease more drastically, and the increase in compliance due to output monitoring was shown to be drastically attenuated. Hypotheses 6(a)(b) were both supported (<Table 7>).

When goal incongruence is high, compared to when it is low, opportunistic behavior was drastically increased by franchisor behavior monitoring, and the decrease in opportunistic behavior due to output monitoring is rastically attenuated. Hypotheses 7(a)(b) were both supported (<Table 8>).

The overall results of hypothesis testing are summarized in <Table 9>.

Regarding the characteristics of franchise contracts (completeness, one-sidedness), this research coded each clause of the franchise contract into numbers. Based on the contracts of 58 brands from eight industries, two coders analyzed and quantified the characteristics of the contract. Methodological implications were provided.

In addition, by studying eight industries with many conflict issues and a high number of franchisors and franchisees, including restaurants, services, wholesale and retail, and automobile sales, it increased the universality of the research. And given that goal incongruence increases the likelihood of conflict, we analyzed the influence of goal incongruence on the relationship between monitoring and franchisee behavior.

Through the following research, we were able to identify the following realistic implications.

First, there is a need to reduce monitoring by increasing the finality of contracts. Since behavior monitoring is a face-to-face control method with the franchisee, there is a possibility of conflict with the franchisee. The franchisor will incur more costs based on the number of face-to-face interactions. But the one-sidedness of contracts leads to an increase in behavioral monitoring, while high completeness can reduce monitoring. Thus, by reducing contractual one-sidedness and increasing completeness, a franchisor can reduce the potential for conflict and cost.

Second, working with the merchant screen is important. The franchisor must create mechanisms that reduce goal incongruence between the franchisee and the franchisor. In recruiting for franchisees, the franchisor must have a policy mechanism that could discern people with attitudes compliant with the headquarters policy of franchise business uniformity and quality assurance.

Third, we need to change the way we do monitoring. According to our research results, in cases where both behavior monitoring and output monitoring were conducted, it showed that it has a positive influence on the franchisee's behavior. Therefore, the franchisor can conduct output monitoring by introducing non-face-to-face methods (order quantity, inventory check through POS records, etc.) where the franchisee is not aware of the monitoring. In this way, the franchisor can reduce the cost of monitoring and share the monitoring results with the franchisee to show that the franchisor is diligently practicing output monitoring. In addition, if behavioral monitoring is strengthened with information provision and consultation function, we expect that the impact of monitoring can be maximized.