Master Thesis
MSc Business Administration - Strategic Innovation Management
In or out?
Assessing the impact of extrinsic and intrinsic rewards on franchisees’ intention to exit the system and the moderating
role of franchisees’ differentiated motivations.
Gloria Chan
Student ID: 2163365
E-mail: g.e.chan@student.rug.nl
Supervisor: Thijs Broekhuizen
Co-assessor: Rene van der Eijk
Word count: 10,993
2
PREFACE
Two years ago, I flew 5556 miles from Los Angeles to Amsterdam with the objective of achieving a Master Degree in Business Administration at the University of Groningen. After successfully completing the Pre-Master and the courses of the Master Program, today January 20
th, 2014, I am proud to present this Master Thesis which reflects my dedication and hard work as well as the support of all the people who accompanied me along this path: the unconditional support of my parents and brother in Los Angeles and my relatives in Colombia; the help and guidance of my supervisor Dr. Thijs Broekhuizen and co-assessor Dr.
Rene van der Eijk; and the encouragement my friends in the Netherlands who became my local family and made my international student life an unforgettable experience. Thanks to all of them, one of my dreams will become true, and a new chapter in my life will open.
~Gloria Chan
3
INDEX
1. INTRODUCTION ... 5
2. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT ... 8
2.1. CONCEPTUAL MODEL ... 8
2.2. DRIVERS: EXTRINSIC AND INTRINSIC REWARDS ... 10
2.3. MODERATORS: EXTRINSIC AND INTRINSIC MOTIVATIONS ... 11
3. METHODOLOGY ... 12
3.1. SAMPLE AND DATA COLLECTION PROCEDURES ... 12
3.2. MEASURES AND MEASUREMENTS VARIABLES ... 13
4. RESULTS AND ANALYSIS ... 20
4.1. DESCRIPTIVE STATISTICS ... 20
4.2. HYPHOTESES TESTING ... 20
4.2.1. Direct effects ... 20
4.2.2. Moderating effects ... 21
4.3. ADDITIONAL ANALYSIS ... 23
4.3.1. Objective performance rather than subjective performance as extrinsic rewards ... 23
4.3.2. Separate tests for independent variables ... 24
4.4. SUMMARY OF EMPIRICAL RESULTS ... 28
5. CONCLUSION AND DISCUSSION ... 29
5.1. CONCLUSION ... 29
5.2. DISCUSSION ... 30
5.2.1. Theoretical implications ... 30
5.2.2 Managerial implications ... 31
5.3. LIMITATIONS AND FUTURE LINES OF RESEARCH ... 32
6. REFERENCES ... 32
4
ABSTRACT
Several studies have analyzed franchise exit at the system and unit level, but only taking into account franchisor’s perspective. However, relatively little has been known about franchisees’ perception of rewards and their motivations. Therefore, this study takes into account franchisee’s perspective and aims to explore the impact of extrinsic and intrinsic rewards on franchisees’ intention to exit the system and the moderating role of franchisees’
differentiated motivations. The theoretical framework is based on Motivational and Self- Determination theories, and it is tested using Partial Least Squares (PLS) by means of survey data of 127 franchisees of a large Dutch franchise system. The empirical results indicate that even though franchisees are heterogeneous in how they are strongly extrinsic and intrinsic motivated, they all consider intrinsic rewards more important than extrinsic rewards.
Key words: Extrinsic rewards, extrinsic motivations, intrinsic rewards, intrinsic motivations.
5 1. INTRODUCTION
Current studies about franchise exit have been tried to explain it at the system and unit level. For instance, at the system level, system characteristic such as an inaccurate replication of the franchise template (Winter et al., 2012), franchisee business fraud, poor franchisee screening, and persistent franchisor-franchisee conflict could lead franchisees to exit the system (Stanworth et al., 1998). Franchise exit has been commonly examined through the lenses of the Agency Theory and the Resource-Based Theory by focusing on how franchisor’s policy decisions affect franchisee survival (Michael and Combs, 2008), the information environment in which the franchisor designed the franchise contract to counter the negative effects of adverse selection, moral hazard and free-riding (Shane, 2001; Lafontaine, 1992), and the decrease or increase of franchisor’s resources (Holmberg and Morgan, 2003). In addition, franchise exit has been explained at the unit level. For example, firm performance has been addressed by using an economic analysis focused only on financial factors such as sales, net income, total financial capitalization at start-up, and capital intensity (Bates, 1998;
Castrogiovanni et al., 1993). By conducting such an economic analysis, researchers only focus on explaining defection rates based on objective financial factors. Also, theories such as the Life-Cycle and the U-Curve have been frequently used to suggest that after the franchisee passes through the “honeymoon” phase, characterized by excitement and fascination, the relationship is deteriorated, and this causes the franchisee to consider other business alternatives (Blut et al., 2011).
Even though all these studies provide important insights into franchise exit, scholars have called for a greater need for theory and research from the franchisee’s perspective (Combs et al., 2004, 2010). Researchers have mainly focused on the franchisor’s perspective providing partial understanding about the behavior and motivations of franchisees (Meek et al., 2011) so there is a need for a more comprehensive view of franchise exit. For these reasons, the objective of this study is to contribute to the franchise exit discussion by taking into account franchisees’ point of view, their perceptions of rewards (beyond objective financial factors) and their differentiated motivations (intrinsic and extrinsic) based on Motivational and Self-Determination theories.
It is important to understand franchise exit as franchising has become an increasingly
popular way of entrepreneurship and a business format in different industries around the
world. Initiated in the United States, it rapidly expanded to Canada, United Kingdom, parts of
Western Europe and emerging markets such as Brazil (Welsh, et. al., 2006). It is estimated
6
that franchising in the United States generates 800 billion worth on business in gross sales and represents over 40% of retail sales (Michael, 1996). Franchising has captured most of the industry shares in tax preparation (67% of sales), printing and copying (71% of sales), and restaurants (46% of sales - Michael, 1996). Also, it is estimated that a chain organization opens a new outlet every eight minutes in the United States (Winter and Szulanski, 2001). For this reason, franchising is an important organizational form, and therefore franchise exit brings disadvantages for the franchisor and the entire system because it has a big influence on the system’s stability and performance. For instance, franchisors have high start-up costs because they have to develop a franchise system, and this is usually an expensive, long, and risky process. They have to make an up-front investment from their own resources for an extensive period of time before achieving a breakeven trading status (Stanworth et al., 1998).
They need to develop a franchise system infrastructure (providing for the recruitment, training and support of franchisees) before they reach a sustained royalty flow (Stanworth et al., 1998). Furthermore, when franchised outlets are closed (taken over by the franchisor via purchase or acquisition, or converted into an independent operation), the franchise system experiences a degree of disruption (Frazer, 2001). This situation could lead current franchisees to exit, and the system could be in danger of collapsing. Every franchise unit that exits has an impact on others in the system so the effects are far reaching and can be substantial (Frazer and Winzar, 2005). The closure of outlets also represents a negative signal to consumers suggesting that the system is unstable. This damages the brand and the reputation of other franchisees in the system, so it could be more difficult to attract new franchisees. Moreover, franchise exit also affects other parties beyond the franchisor and franchisees. For example, suppliers, financial institutions, customers, investors, and other stakeholders may have real costs/losses and/or opportunity costs due to franchisees exiting the system (Holmberg and Morgan, 2003). These costs could be avoided or minimized by understanding franchisees’ perception of rewards and motivations.
In business format franchising, the franchisor (owner of a tried-and-tested business
format) allows franchisees (usually aspiring entrepreneurs) to use it in exchange for an
entrance fee payment followed by an on-going royalty (Stanworth et al., 1998). In other
words, the franchisor sells the right to use the operating system, products/services, and trade
name to franchisees as well as provides training and continuous support in order to ensure a
good relationship and performance of each franchisee. The franchisees are allowed to offer
the franchisor’s product/service under the franchisor’s name within a specified region and
time period (Justis and Judd, 1998). Even though franchisees receive support from the
7
franchisor, there is a power imbalance: the franchisor controls the franchise system, brand name, intellectual property, and the business and marketing strategy at the system level; and the franchisees control the hands-on implementation and operation of the business at the unit level and own valuable local market knowledge because of the close interaction with customers (Weber, 2013). Therefore, most of the power is held by the franchisor because the franchisees are heavily dependent on the franchisor's intellectual property or know-how, mainly in the early stages where the learning curve is steep (Frazer et al., 2007). Franchisees may find themselves in a subordinate position because of their limited financial resources, organizational status and lack or experience (Oxenfeldt and Thompson, 1968). Consequently, they struggle to find energy, mobilize effort and be persistent at the tasks of life and work (Ryan and Deci, 2000).
In order to remain in the franchise system, franchisees are driven by their perceptions of rewards and their motivations. Some franchisees find extrinsic rewards (i.e. monetary outcomes) that result from performing the activity more valuable than intrinsic rewards, as they are extrinsically motivated because they are driven by the desire of acquiring a secure future or increase personal wealth. On the other hand, franchisees can also find intrinsic rewards (i.e. job satisfaction) while performing the task more valuable than extrinsic rewards, as they are intrinsically motivated because they are driven by the desire of acquiring personal growth and finding excitement to meet the challenge (Ryan and Deci, 2000). In order to explore the accuracy of these relationships, this research study will address the following questions:
How do extrinsic and intrinsic rewards impact franchisees’ intention to exit the system?
Do franchisees’ extrinsic and intrinsic motivations moderate the importance of
rewards?
The main objective of this study is to develop knowledge in the area of franchise exit
by taking into account franchisees’ heterogeneity based on their individual perception of
rewards and motivations. This study investigates how the degree of extrinsic and intrinsic
motivations impact franchisees’ needs to obtain extrinsic and intrinsic rewards. For instance,
franchisees may differ in the way they value certain rewards: (1) those who value extrinsic
rewards and are extrinsically motivated so they engage in job activities because of their
tangible value and (2) those who value intrinsic rewards and are intrinsically motivated so
they experience their work as inherently enjoyable and satisfying. By acknowledging the
differences in the value that franchisees attach to rewards, this study tries to open the “black
8
box” regarding franchise exit based on individual franchisees’ perception of rewards and their motivations.
This study contributes to current literature by not treating franchisees as one type: their perceptions of rewards and their motivations differ, and this may lead them to react differently. Furthermore, this provides managerial contributions to the area of entrepreneurship, franchising, and human resources. Since employee’s work motivation is one of the main determinants of an organization's development and success (Park and Word, 2012), motivating employees to be both effective and positive in performing their work remains an important challenge for franchisors and managers, as well as for the human resources department because they need to select, retain, and manage highly motivated people (Park and Word, 2012). By identifying the type of franchisee (strongly externally or internally motivated) present in the system, franchisors can potentially better serve them, and prevent dissatisfaction and subsequent exit.
The structure of the paper is organized as follows. The conceptual model, the literature background on Motivational Theory and Self-Determination Theory and the hypotheses are presented in section 2. The specifications of the data and the measurements are described in section 3. The results are presented and discussed in section 4. The final section addresses the conclusion and provides suggestions for future work.
2. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
2.1. CONCEPTUAL MODE L
This study has its foundation on two theories: Motivational Theory (Adams, 1965;
Porter and Lawler, 1968) and Self-Determination Theory (Ryan and Deci, 2000). The first
theory addresses entrepreneurial motivation as the process by which entrepreneurs
(prospective franchisees) decide whether or not to engage in entrepreneurial behavior by
taking into account intrinsic and extrinsic rewards (Kuratko et al., 1997). Based on this
theory, two hypotheses are developed which suggest a perceived expectation-outcome
relationship: the franchisee’s perception that the outcomes of the intended organization will
meet or exceed his/her expectations. For this reason, the prospective franchisee enters the
franchise system with expectations of extrinsic and intrinsic outcomes that will result from the
beginning and during the operation of the venture (Naffziger et al., 1994). When outcomes
fail to meet expectations, this will have a corresponding impact on the decision to continue to
act entrepreneurially within the franchise system (Naffziger et al., 1994). In other words, if
9
the franchisee’s perceived expectation-outcome is lower than expected, he/she will consider exiting the system (Figure 1: H1 and H2).
Secondly, Self-Determination Theory (SDT) accounts for differences in the need structure among franchisees by identifying extrinsic and intrinsic sources of motivation (Ryan and Deci, 2000). Based on this theory, two hypotheses are developed which suggest that franchisees may differ according to their motivational strength in the need to obtain intrinsic or extrinsic rewards (Figure 1: H3 and H4).
Furthermore, intention to exit refers to franchisee’s idea and propensity to terminate the franchise relationship in the near future by actively looking for alternatives or ways to exit the franchise system (Croonen and Brand, 2009). When a franchisee reaches this point, it is important that he/she set up an exit strategy (selling or closing down the business) in order to maximize the value of the business when it is time to walk away (Webber, 2013).
The four proposed hypotheses and the control variables, which constitute some of the relevant characteristics of franchisees, are presented in the following conceptual model:
Figure 1: Conceptual Model
10
2.2. DRIVERS: EXTRINSIC AND INTRINSIC REWARDS
Motivational Theory (Adams, 1965; Porter and Lawler, 1968) addresses entrepreneurial motivation as the process by which entrepreneurs (prospective franchisees) decide whether or not to engage in entrepreneurial behavior (Kuratko et al., 1997). This theory moves beyond the issue of firm start-up to deal with the process that motivates an entrepreneur (i.e. franchisee) to stay with the entrepreneurship (i.e. franchising) as a career choice. Specifically, the motivational process of new venture creation and sustained ownership encompasses perceptions which include: (1) the impact of both intrinsic and extrinsic rewards on sustained behavior, (2) an expectation-outcome relationship, and (3) the value of rewards and their impact upon sustained behavior. This Motivational Theory describes the entrepreneurial path as primarily goal-oriented since entrepreneurs tend to direct their efforts to reach specific goals, and they value intrinsic and extrinsic rewards (Kuratko et al., 1997). In the case of franchise exit, it provides an insight of how franchisees value rewards attained from performing the job and how this affects their intention to exit.
According to the Motivational Theory, both intrinsic and extrinsic rewards have in general, a positive impact on sustained behavior. Extrinsic rewards refer to the tangible or verbal outcomes which are derived from performing an activity (Ryan and Deci, 2000). They include financial or other tangible rewards that are the result of the (financial) performance of the firm (Naffziger et al., 1994). Franchisees who value extrinsic rewards obtain gratification not from performing the activity itself but rather from the external outcomes that the activity leads to (Gagné and Deci, 2005). Extrinsic rewards can be measured by firm financial performance which encompasses two types: objective and subjective. The former refers to the actual monetary achievements which can be measured based on categories such as investor returns (change in price per share and dividend income) and accounting returns (earnings per share or price/earnings ratios) (Cochran and Wood, 1984); and the latter refers to the interpretation and perceptions of success. This study will focus on subjective financial performance rather than objective facts since it is considered to be better for analyzing perception and behavior. Based on this, we can infer that the lower a franchisee experiences subjective financial performance (i.e. extrinsic rewards), the more likely he/she will consider exiting the system:
H1: Higher extrinsic rewards reduce franchisee’s intention to exit the franchise system.
11
Apart from the extrinsic rewards, franchisees can value intrinsic rewards which encompass other means than simply money (Kuratko et al., 1997). They are based on employees getting a positively valued experience from doing their work (Spreitzer et al., 1997; Thomas and Tymon, 1994), the satisfaction of being their own boss, being more in control of their destiny, and having ultimate responsibility for the success of the venture (Naffziger et al., 1994). The value of personal satisfaction while performing the job helps predicting franchisees’ adjustment in the workplace (Deci et al., 2001). In other words, job satisfaction facilitates work engagement and psychological well-being of the franchisees.
Based on this, we can infer that the lower a franchisee experiences job satisfaction (i.e.
intrinsic rewards), the more likely he/she will consider exiting the system:
H2: Higher intrinsic rewards reduce franchisee’s intention to exit the franchise system.
2.3. MODERATORS: EXTRINSIC AND INTRINSIC MOTIVATIONS
Self-Determination Theory (SDT) provides a theoretical framework for a more in depth understanding of franchisees’ motivation that account differences in the need structure across franchisees. SDT identifies extrinsic and intrinsic sources of motivation and describes the respective roles of each motivation in individual differences (Ryan and Deci, 2000). SDT proposes three different categories of work motivation: (1) amotivation, (2) extrinsic motivation, and (3) intrinsic motivation (Ryan and Deci, 2000). This study will focus on the last two categories since the decision of becoming a franchisee is a highly involved decision which is not likely to be determined by an unmotivated behavior.
Franchisees who are extrinsically motivated perform an activity because of the
evaluations and opinions they fear others might have of them (Deci and Ryan, 2000), as well
as the desire to acquire wealth, increasing personal income and income opportunities (Kuratko
et al., 1997). They are motivated to work primarily in response to something apart from the
work itself (Achakul and Yolles, 2013). The activity becomes a means to an end rather than
an end in itself (Deci and Ryan, 1987). In other words, the behavior is carried out in pursuit of
external rewards and not because it is interesting or fun (Deci, 1975). Therefore, franchisees
who are strongly extrinsically motivated find extrinsic rewards more valuable than those who
are less extrinsically motivated. For instance, if a franchisee experiences low subjective
financial performance (i.e. extrinsic rewards), he/she will consider exiting the system, and this
decision is stronger if he/she is strongly extrinsically motivated:
12
H3: A higher level of a franchisee’s extrinsic motivation strengthens the relationship (i.e.
becomes more negative) between extrinsic rewards and franchisee’s intention to exit the franchise system.
In contrast, franchisees who are intrinsically motivated have an internal desire to do something because of passion and interest. They perform an activity and experience it as interesting, enjoyable, and satisfying (van Beek et al., 2012). They engage in their work for the challenge and enjoyment of it since the work itself is motivating. They have an inherent tendency to extend and exercise their capacities, to explore, and to learn (Ryan and Deci, 2000). They are driven to perform the job because they satisfy the need for control, and personal achievement (Timmons, 1976). They enjoy performing the activity and being entrepreneurs. For this reason, intrinsic motivations energize and sustain activities, and they emphasize experience-driven reasons which are part inherently in the activity, and are closely tied with individual interests (Ryan and Deci, 2000). Therefore, franchisees who are strongly intrinsically motivated find intrinsic rewards more valuable than those who are less intrinsically motivated. For instance, if a franchisee experiences low job satisfaction, he/she will consider exiting the system, and this decision is stronger if he/she is strongly intrinsically motivated:
H4: A higher level of a franchisee’s intrinsic motivation strengthens the relationship (i.e.
becomes more negative) between intrinsic rewards and franchisee’s intention to exit the franchise system.
3. METHODOLOGY
This study follows a quantitative research method since the research questions aim to explore whether specific determinants predict behaviors at a statistically significant level. In order words, how different levels of rewards and motivations affect franchisees’ intention to exit. In order to assess these relationships, a survey with closed-ended questions is designed.
The sample and data collection procedures as well as the measures and measurement variables are explained in the following sub-sections.
3.1. SAMPLE AND DATA COLLECTION PROCEDURES
For the purpose of anonymity, this study will refer to the franchise system as V. This
system has 209 franchisees and no company-owned units at the time of data collection. It
provides home repair services to its customers, and it is formed mostly of experienced
13
professionals who work alone or with very few personnel. The franchisor provides franchisees with a strong brand name, marketing, training, and coaching as well as it provides them access to financial services such as insurance and purchasing arrangements. There are strict obligations regarding a uniform presentation towards the customers (e.g. professional uniforms, company vehicles, and use of company logo on printed materials) and since customer acquisition and communication go largely through the internet, they share a common ICT-system.
The data was collected in the fall of 2011 as part of a larger project. Secondary data was gathered from V’s website and internal documents. In addition, primary data was collected from semi-structured explorative interviews that were conducted with three franchisees and with the franchisor’s CEO. All these sources helped to develop the questionnaire and enhance its validity. The dataset compiled is composed of one single system with a large number of franchisees in order to control for industry and system differences and minimize the potential effects of differences in business formats and market conditions that may affect franchisees (Davies et al., 2011).
In order to enhance the willingness to cooperate in this study, the data collectors co- organized regional meetings with the franchisee council for all franchisees, and the objective and relevance of the larger project was explained. Franchisees who were present at these meetings also received a questionnaire in person. Weeks later after the meetings, several e- mail reminders were sent, and ultimately, all non-respondents were reminded by a telephone call. 209 questionnaires were sent of which 135 questionnaires were returned, and 127 were complete. The response rate was 60.7% and this is high for survey methods (Deutskens et al., 2004). All franchisees were male, which was mainly due to the industry context and not because of a response bias.
3.2. MEASURES AND ME ASUREMENTS VARIABLES
This study uses the statistical software application SmartPLS to conduct the PLS path
modeling to test the theoretical model. The sample contains 500 subsamples of 200, so this
technique is preferable because the sample is relatively small (Jakobowicz, 2007). Since
rewards and motivations have been studied in other organizational fields, the survey items of
this study are based on existing scales. Table 1 presents the list of items for each construct, the
score, the source where they were obtained and adapted from, and the scale.
Constructs’ items
Score Description Source 5-point likert scale
Extrinsic Rewards
SUBJPERF1 0.81 As a whole I consider my company as successful. Dant and
Gundlach (1999) and Cochet et al.
(2008).
1 = strongly agree 5 = strongly disagree SUBJPERF2 0.88 I am satisfied with my sales and profit trends.
SUBJPERF3 0.85 I have recently achieved my financial goals.
SUBJPERF4 0.71 Compared to my colleagues, I did a better job in the construction sector in the past year.
Intrinsic Rewards
INT_REWARDS1 0.74 I tell others that company V is a good formula.
Farndale et al.
(2011)
1 = strongly agree 5 = strongly disagree INT_REWARDS2 0.81 I am proud of being a V franchisee.
INT_REWARDS3* N/A There are more attractive alternatives compared to having a V membership.
INT_REWARDS4 0.69 I am prepared to do more than is expected from me Extrinsic Motivation
EXTRINSIC1 0.66 To obtain a pleasant and luxurious life.
Kuratko et al.
(1997)
1 = very important 5 = very unimportant EXTRINSIC2 0.90 To increase my personal income.
EXTRINSIC3 0.80 To increase my income opportunities.
EXTRINSIC4* N/A To secure my future.
EXTRINSIC5* N/A To make a lot of money in a short time.
EXTRINSIC6* N/A To gain public recognition.
Intrinsic Motivation
INTRINSIC1* N/A To meet the challenge.
Kuratko et al.
(1997)
1 = very important 5 = very unimportant INTRINSIC2 0.77 To obtain personal development.
INTRINSIC3 0.78 To prove to myself that I can do it.
INTRINSIC4 0.73 To deliver a high quality job.
Intention to Exit
INTENTTOLEAVE1 0.89 I do not think about leaving company V.
Kelloway et al.
(1999)
1 = strongly agree 5 = strongly disagree INTENTTOLEAVE2 0.85 I'm planning something else to do than being a V franchisee.
INTENTTOLEAVE3 0.85 I will no longer be at company V.
INTENTTOLEAVE4 0.88 I would want to terminate my franchise agreement.
INTENTTOLEAVE5 0.86 It is my intention to extend my franchise agreement when it expires.
INTENTTOLEAVE6R 0.67 I would recommend company V to potential new franchisees.
INTENTTOLEAVE7R 0.79 The likelihood of exiting the system when something (bad) happens.
15 Objective Performance
OBJPERF1 0.96 Gross margin annually
Fang et al.
(2008)
OBJPERF2 0.73 Annual sales
OBJPERF3 0.96 Total gross profit annually
Table 1: Constructs’ items
Notes: Items marked with (*) were dropped; (R) were reversed-scale item. Objective performance will be used for additional analysis in Section 4.3.
According to table 1, t
he independent variables are:
Extrinsic rewards (subjective financial performance): Franchisee’s perception of
working in a successful company, satisfaction with sales and profit trends, achievement of financial goals, and perception of doing better compared to colleagues in the same type of job and industry.
Intrinsic rewards: Telling others that the franchise system where they belong to is good
and being proud of it, being prepared to do more than it is expected, and recognizing that there are more attractive alternatives.
Moreover, the moderating variables are formed by:
Extrinsic motivations: The importance a franchisee attaches to obtaining a pleasant
and luxurious life, increasing personal income, and income opportunities.
Intrinsic motivations: The importance a franchisee attaches to entering into a
challenge, obtaining personal development, proving “I can do it”, and delivering a high quality job.
In addition, the dependent variable is:
Intention to exit: Thinking about leaving company V, planning something else to do
than being a V franchisee, no longer being at company V, wanting to terminate the franchise agreement, intention to extend the franchise agreement when it expires, recommending company V to potential new franchisees, and the likelihood of exiting the system when something (bad) happens.
Also, control variables are included to provide a stronger test of the hypotheses: franchisees’
age, education and branch experience. These control variables are related to franchising and they may impact a franchisee’s intention to exit. Table 2 presents the respondents’
characteristics.
Age Frequency Percentage
<30 years 3 2.3%
Between 30-40 years 29 22.5%
Between 41-50 years 65 50.4%
Between 51-60 years 30 23.3%
>60 years 2 1.6%
Education
Elementary school 5 3.9%
Secondary school 67 51.9%
Community College 48 37.2%
University 9 7.0%
17 Branch Experience
0 to 2 23 17.80%
3 to 5 34 26.40%
6 to 12 26 20.20%
13 to 20 24 18.60%
21+ 22 17.10%
Table 2: Respondents’ characteristics. Net response = 129
In order to construct the scales, the validity and reliability of each scale were assessed.
A confirmatory analysis was conducted by checking the reliability and validity, and when necessary items were removed. Items with low loading scores (lower than 0.50) were dropped and thereby, the scales were “purified” such that a larger number of measurement items were reduced into a smaller number of factors (Churchill, 1979).
In order to assess convergent validity, the Cronbach’s Alphas were assessed because it is an internal consistency test that measures the degree of which the items (measurements) consistently measure the underlying latent construct (Fornell and Larcker, 1981). The Cronbach’s Alpha is expressed as a number between 0 and 1, and it can be assumed that the items have good internal reliability or are measuring the same thing consistently when the Cronbach’s Alpha is equal or greater than 0.7. In addition, when the communality measures for each item are standardized, the average communality of a block of indicators is referred to as Average Variance Extracted (AVE) (Fornell and Larcker, 1981). If the AVE measures exceed 0.5, the latent variable explains at least half of its measures. Table 3 presents the values of AVEs, Composite Reliability, and Cronbach’s Alphas values, and the total items per construct for tested model. It can be seen that the AVE values vary from 0.54 to 0.81, and they are all greater than 0.5. Similarly, constructs are also reliable as most of the Composite Reliabilities and Cronbach’s Alpha values are above the recommended 0.7 level (Bagozzi and Yi, 1988). Thus, the constructs are considered to be satisfactory with evidence for reliability.
Constructs AVE Composite Reliability
Cronbach’s Alpha
Number of items
Extrinsic Rewards 0.66 0.89 0.83 4
Intrinsic Rewards 0.54 0.56 0.26 3
Extrinsic Motivations 0.63 0.83 0.74 3
Intrinsic Motivations 0.59 0.81 0.66 3
Age N/A N/A N/A N/A
Branch Experience N/A N/A N/A N/A
Education N/A N/A N/A N/A
Objective Performance 0.81 0.93 0.94 3
Intention to Exit 0.69 0.94 0.92 7
Table 3: Construct reliability and validity
18
Moreover, discriminant validity was assessed to ensure that the items’ measure of each variable actually measured it and was not measuring another variable (Fornell & Larcker, 1981) (i.e. if they are sufficiently different). In fact, it estimates the extent to which a given construct is different from other constructs (Hulland, 1999). In order to assess the discriminant validity, two things need to be considered: First, the correlation of the latent variable scores with the measurement items needs to show an appropriate pattern of loadings, one in which the measurement items load highly on their theoretically assigned factor and not highly on other factors (Gefen and Straub, 2005). Table 4 shows the pattern loadings of each measurement item, and it can be seen that the ones higher than 0.6 (highlighted in red) correspond to the assigned factor.
Constructs Extrinsic
Motivations
Extrinsic Rewards
Intention to Exit
Intrinsic Motivations
Intrinsic Rewards
EXTRINSIC1 0.66 0.24 0.02 0.34 0.09
EXTRINSIC2 0.90 0.16 0.10 0.34 -0.03
EXTRINSIC3 0.80 0.06 0.07 0.59 0.09
SUBJPERF1 0.24 0.81 -0.27 0.23 0.38
SUBJPERF2 0.06 0.88 -0.30 0.16 0.28
SUBJPERF3 0.14 0.85 -0.33 0.17 0.30
SUBJPERF4 0.10 0.71 -0.32 0.14 0.39
INTENTTOLEAVE1 0.09 -0.24 0.89 -0.08 -0.59
INTENTTOLEAVE2 0.08 -0.35 0.85 -0.22 -0.68
INTENTTOLEAVE4 0.06 -0.38 0.85 -0.17 -0.58
INTENTTOLEAVE5 0.03 -0.36 0.88 -0.19 -0.67
INTENTTOLEAVE6 0.13 -0.35 0.86 -0.16 -0.64
INTENTTOLEAVE7 0.14 -0.21 0.67 0.00 -0.59
INTENTTOLEAVE8 0.02 -0.27 0.79 -0.12 -0.59
INTRINSIC2 0.47 0.17 -0.12 0.77 0.15
INTRINSIC3 0.33 0.14 -0.15 0.80 0.15
INTRINSIC4 0.40 0.20 -0.09 0.72 0.16
INT_REWARDS1 0.08 0.30 -0.53 0.02 0.74
INT_REWARDS2 0.16 0.41 -0.54 0.15 0.81
INT_REWARDS3 0.09 0.33 -0.45 0.23 0.69
Table 4: Loadings and cross-loadings
Second, an analysis of the AVE was done by comparing the latent variable
correlations. As a rule for valid discriminant validity, the square root of the AVE of each
construct must exceed the correlations with all other constructs (Chin, 1998). The variable
correlations are shown in table 5. It can be seen that the square root of AVE in the diagonals
of matrix is higher than the correlation of the other values, and the highest correlation is
19
between intention to exit and intrinsic rewards (r = -0.75). This confirms that there is evidence for discriminant validity.
AGE BE EDU EM ER IE IM IR OP
AGE N/A
BE -0.09 N/A
EDU -0.08 -0.04 N/A
EM -0.17 0.06 0.11 0.79
ER -0.09 -0.03 -0.03 0.16 0.81 IE -0.03 0.04 -0.03 0.09 -0.38 0.83 IM 0.02 -0.07 -0.1 0.51 0.21 -0.17 0.77 IR -0.08 -0.09 0.15 0.04 0.42 -0.75 0.2 0.74 OP 0.08 0.07 -0.18 0.08 0.36 -0.07 0.18 0.05 0.65
Table 5: Correlations
Note: Square root of AVEs on the diagonal. AGE = Age; BE = Branch Experience; EDU = Education; EM = Extrinsic Motivation; ER = Extrinsic Rewards; IE = Intention to Exit; IM = Intrinsic Motivation; IR = Intrinsic Rewards; OP = Objective Performance.
Since PLS does not have formal testing procedures to assess the validity of the overall model, an alternative is measuring the Goodness-of-Fit (GoF) which can serve as a diagnostic value (Wetzels et al., 2009). The GoF index for PLS path model is defined as the geometric mean of the average communality and the average R
2for endogenous constructs (Tenenhaus et al., 2005) and it is represented in the following equation:
* R
2y Communalit GoF
This index varies between 0 and 1, and even though a minimum threshold does not exist, a value >0.36 is recommended (Wetzels et al., 2009). Table 6 presents the R
2, the communality of the constructs, and the GoF index:
Model R2 Communality GoF Intention to Exit 0.48 0.69 0.57
Table 6: Model validity
According to table 6, the GoF index of this study is 0.57 which is the above the value
recommended and suggests that the model performs favorably against baseline models.
20 4. RESULTS AND ANALYSIS
4.1. DESCRIPTIVE STATISTICS
The total observations are 129. The minimum value for extrinsic rewards, intrinsic rewards, intention to exit is 1 and for extrinsic motivation and intrinsic motivations is 2. The maximum value for all five constructs is 5. As for the control variables, the minimum value for age is 27 years and the maximum is 62 years; the minimum value for education is 1 year and the maximum is 4 years; and the minimum value for branch experience is 1 year and the maximum is 5 years. In addition, the minimum value for objective performance is 57,637 and the maximum value is 721,096. This information along with the mean and standard deviation for each construct are summarized in Table 7.
Variables N Minimum Maximum Mean
Std.
Deviation
ER 129 1 5 3.51 0.84
IR 129 1 5 3.50 0.96
EM 129 2 5 4.04 0.62
IM 129 2 5 4.46 0.51
IE 129 1 5 2.55 1.13
AGE 129 27 62 44.83 7.82
EDU 129 1 4 2.47 0.69
BE 129 1 5 2.91 1.36
OP 129 57,637 721,096 188,813.78 90,074.28 Table 7: Descriptive statistics
Notes: ER = Extrinsic Rewards; IR = Intrinsic Rewards; EM = Extrinsic Motivation; IM = Intrinsic Motivation;
IE = Intention to Exit; AGE = Age; EDU = Education; BE = Branch Experience; OP = Objective Performance.
4.2. HYPHOTESES TEST ING
4.2.1. DIRECT EFFECTSThe statistical significance of each path coefficient is tested using a bootstrapping
procedure in Smart PLS 2.0. Three models are presented in Table 8: M
1presents the results of
the direct effects and it explains 41% of the variance of the construct; M
2presents the results
of the direct and moderating effects and it explains 51% of the variance of the construct; and
M
3presents the results of the direct and indirect effects and the control variables and it
explains 52% of the variance of the construct. It can be seen that there is an increase in the
percentage of the variance from M
1to M
3(
M2ΔR2 = 0.10; M3ΔR2 = 0.02)which suggests a
21
satisfactory level of predictive performance. In addition, one of the four hypotheses is significant at the 5% level.
Model 1 (M1) Model 2 (M2) Model 3 (M3) Structural
Relationships β t-
value β t-
value β t-
value
H1: ER → IE -0.13 1.49 -0.11 1.23 -0.11 1.19
H2: IR → IE -0.58** 6.65 -0.56** 5.90 -0.59** 6.66
H3: ER * EM → IE 0.17 1.01 0.17 0.99
H4: IR * IM → IE 0.08 0.62 0.08 0.55
Control variables:
Age -0.07 0.89
Branch exp. -0.06 0.78
Education 0.06 0.76
R2 0.41 0.51 0.52
ΔR2 0.10 0.02
Table 8: Structural model results
Based on one-tailed tests, t-values greater than 1.65 are significant at p*< 0.05; t-values greater than 2.33 are significant at p**< 0.01.
Notes: M1= Direct effects; M2= Direct and moderating effects; M3= Direct and moderating effects and including control variables; ER = Extrinsic Rewards; IR = Intrinsic Rewards; EM = Extrinsic Motivation; IM = Intrinsic Motivation; IE = Intention to Exit.
According to table 8, extrinsic rewards have a negative but not significant effect on intention to exit throughout all three models (H1:
M1β = -0.13;
M2β = -0.11;
M3β = -0.11, all p’s >0.05) so there is no support for H1. As predicted in H2, intrinsic rewards have a significant negative effect on intention to exit in all 3 models (H2: M
1β = -0.58; M
2β = -0.56; M
3β = -0.59, all p’s
<0.01). Overall, support is found mainly for the influence of the intrinsic rewards on franchisee’s intention to exit.
4.2.2. MODERATING EFFECTS
Two methods are used to test the moderating effects: Product indicator approach in SmartPLS and a linear regression in SPSS. Based the PLS results (See Table 8), extrinsic motivations have a positive but not significant effect on the negative relationship between extrinsic rewards and intention to exit (H3: M
2β = 0.17; M
3β = 0.17, both p’s > 0.05) so H3 is not supported. In addition, intrinsic motivations have a positive but not significant effect on the negative relationship between intrinsic rewards and intention to exit (H4: M
2β = 0.08; M
3β
= 0.08, both p’s > 0.05) so H4 is not supported. Regarding the control variables, age has a
negative but not significant effect (M
3β = -0.07, p>0.05), branch experiences has a negative
22
but not significant effect (M
3β = -0.06, p>0.05) and education has a positive but not significant effect (M
3β = 0.06, p > 0.05).
The second method for measuring the moderating effects is performing a linear regression in SPSS. This approach does not use bootstrapping since the multiplication occurs for the complete construct rather than at the individual item level. Table 9 presents the results of this test:
Coefficientsa
Model β t Sig.
1 (Constant) 1.89 0.06
ER -0.71 -1.05 0.30
IR -0.06 -0.09 0.93
EM 0.07 0.23 0.82
IM -0.05 -0.20 0.84
ER x EM 2.10** 2.77 0.01 ER x IM -1.27 -1.62 0.11 IR x EM -1.80** -2.84 0.01
IR x IM 1.10 1.47 0.14
Table 9: Linear regression results
Note: Dependent Variable: Intention to Exit. Based on one-tailed tests, t-values greater than 1.65 are significant at p*< 0.05; t-values greater than 2.33 are significant at p**< 0.01.
Contrary to the PLS results, table 9 shows that two moderating effects are significant:
The interaction term between extrinsic rewards and extrinsic motivations is statistically significant at 1% level of significance level, and the interaction between intrinsic rewards and extrinsic motivations is statistically significant at 1% significance level and with a negative coefficient. Figures 2 and 3 represent these interactions (standard deviation ±1).
Figure 2: Interaction between Extrinsic Rewards and Extrinsic Motivations
23 Figure 3: Interaction between Intrinsic Rewards and Extrinsic Motivations
Based on figure 2, intention to exit declines more rapidly for franchisees with weak extrinsic motivations as they receive more extrinsic rewards than for franchisees with strong extrinsic motivations. This could be explained through the lenses of franchisees’ expectations.
For instance, franchisees with weak extrinsic motivations have no expectations of obtaining extrinsic rewards. Once they receive high extrinsic rewards, these benefits take them by surprise and are so unexpected that they feel pleased with the system and their intention to exit declines to a greater extent.
According to figure 3, the intention to exit the system for franchisees with strong extrinsic motivations declines as they receive more intrinsic rewards. This is could be also due to the unexpected benefits they receive which make them feel content with the system. As for franchisees with weak extrinsic motivations, the finding is counterintuitive because their intention to exit increases as they receive more intrinsic rewards. These two significant moderating effects based on SPSS results will be further discussed in the conclusion section.
4.3. ADDITIONAL ANALYSIS
4.3.1. OBJECTIVE PERFORMANCE RATHER THAN SUBJECTIVE PERFORMANCE AS EXTRINSIC REWARDS
An additional test is performed in SmartPLS to check if the results are different if extrinsic rewards are measured by objective performance instead of subjective performance.
Two models are presented in Table 10: M
4presents the results of the direct effects and it
explains 40% of the variance of the construct and M
5presents the results of the direct effects
including the control variables, and it explains 43% of the variance of the construct. It can be
seen that there is an increase in the percentage of the variance from M
4to M
5(ΔR
2= 0.03)
which suggests a satisfactory level of predictive performance.
24 Model 4 (M4) Model 5 (M5)
Structural
Relationships
β
t-valueβ
t-value H1: Obj Perf → IE -0.04 0.51 -0.01 0.09H2: IR → IE -0.63** 8.99 -0.68** 9.31
Control variables:
Age -0.13* 1.77
Branch exp. 0.10 0.61
Education -0.05 1.23
R2 0.40 0.43
ΔR2 0.03
Table 10: Results for direct effects with objectives performance as extrinsic rewards
Based on one-tailed tests, t-values greater than 1.65 are significant at p*< 0.05; t-values greater than 2.33 are significant at p**< 0.01.
According to table 10, extrinsic rewards have a non-significant effect on intention to exit (H1: M
4β = -0.04; M
5β = -0.01, both p’s > 0.05) so H1 is again not supported. Intrinsic rewards remain to have a significant negative effect on the intention to exit (H2: M
4β = -0.63, M
5β = -0.68, both p’s < 0.01) so H2 is again supported. These results are consistent with the previous results in models 1 to 3 (See Table 8). There are no substantial differences between objective and subjective financial performance.
4.3.2. SEPARATE TESTS FOR INDEPENDENT VARIABLES
Since extrinsic and intrinsic rewards are positively correlated at 0.42 (See Table 5),
additional tests are performed in SmartPLS to verify if the independent variables in isolation
have an influence on intention to exit. Table 11 presents the results of the isolated test
performed for extrinsic rewards: M
6presents the results of the direct effect of extrinsic
rewards on intention to exit; M
7presents the results of the direct and moderating effect; and
M
8presents the results of the direct and indirect effect including the control variables.
25 Model 6 (M6) Model 7 (M7) Model 8 (M8) Structural
Relationships
β
t-valueβ
t-valueβ
t-valueH1: ER → IE -0.38** 3.78 -0.39** 3.83 -0.39** 4.29
H3: ER*EM → IE 0.15 0.68 0.14 0.69
Control variables:
Age -0.04 0.36
Branch exp. 0.02 0.26
Education -0.05 0.46
R2 0.15 0.19 0.20
ΔR2 0.04 0.01
Table 11: Results for isolated tests of extrinsic rewards
Notes: Based on one-tailed tests, t-values greater than 1.65 are significant at p*< 0.05; t-values greater than 2.33 are significant at p**< 0.01. M6= Direct effects; M7= Direct and moderating effect; M8= Direct and moderating effect including control variables.
Based on table 11, it can be seen that there is an increase in the percentage of the variance from M
6to M
8(M
7ΔR
2= 0.04; M
8ΔR
2= 0.01), but the highest R
2(M
8= 0.20) only explains 20% of the variance of the construct so the level of predictive performance is low.
Also, extrinsic rewards have a significant negative effect on the intention to exit in all three models (H1:
M6β = -0.38,
M7β = -0.39,
M8β = -0.39, all p’s < 0.01), thus contrary to the results in the earlier test (See Table 7), H1 is supported. On the other hand, extrinsic motivations have a non-significant effect on the negative relationship between extrinsic rewards and intention to exit (H3:
M7β = 0.15;
M8β = 0.14, both p’s > 0.05) so H3 is again not supported.
Moreover, an isolated test is performed for intrinsic rewards. Table 12 presents the results of: M
9direct effect of intrinsic rewards on intention to exit; M
10direct and moderating effects; and M
11presents the results of the direct and indirect effect including the control variables.
Model 9 (M9) Model 10 (M10) Model 11 (M11) Structural
Relationships
β
t-valueβ
t-valueβ
t-value H2: IR → IE -0.63** 9.81 -0.61** 7.84 -0.66** 8.23H4: IR*IM →IE 0.11 0.54 -0.09 0.59
Control variables:
Age -0.13 1.56
Branch exp. -0.04 0.59
Education 0.10 1.35
R2 0.40 0.41 0.43
ΔR2 0.01 0.02
Table 12: Results for isolated tests of intrinsic rewards
26 Notes: Based on one-tailed tests, t-values greater than 1.65 are significant at p*< 0.05; t-values greater than 2.33 are significant at p**< 0.01. M9= Direct effects; M10= Direct and moderating effect; M11= Direct and moderating effect including control variables.
According to table 12, there is an increase in the percentage of the variance from M
9to M
11(M
10ΔR
2= 0.01; M
11ΔR
2= 0.02), and the highest R
2(M
11= 0.43) explains 43% of the variance of the construct so the level of predictive performance is high. Also, intrinsic rewards have a significant negative effect on the intention to exit in all three models (H2:
M9β
= -0.63;
M10β = -0.61;
M11β
= -0.66, all p’s< 0.01) so H2 is again supported. On the other hand, intrinsic motivations have a non-significant effect on the negative relationship between extrinsic rewards and intention to exit (H4:
M10β = 0.11;
M11β = -0.09, both p’s > 0.05) so H4 is again not supported. Regarding the control variables, they are non-significant negative effect (Age β = -0.13; branch experience β = -0.04; education β = 0.10, all p’s < 0.05).
The results of these additional tests find support for H1 and H2 if they are analyzed separately. In first place, extrinsic rewards and intrinsic rewards are positively correlated at 0.42 (See Table 5). Besides, when taken in individually they both significantly explain the intention to exit. This suggests that intrinsic rewards are contaminated by extrinsic rewards and vice versa. In order to control for the correlation between these variables when studying their effect on the intention to exit, a linear regression model is performed in SPSS between the two independent variables. In this way, we can check how much the variable intrinsic rewards are explained by the variable extrinsic rewards and vice versa. The figures below show the linear regression model, first using intrinsic rewards as the dependent variable (Figure 4), and then extrinsic rewards as the dependent variable (Figure 5).
Figure 4: Intrinsic rewards as the dependent variable Figure 5: Extrinsic rewards as the dependent variable
27
Next, a partial correlation is performed by taking the residuals of the linear regression model in order to determine the effect of each of the independent variables on intention to exit by controlling for the other variable (Figure 6 and 7).
Figure 6: Residual effects of Intrinsic Rewards residuals on intention to exit.
Table 13: Results for Intrinsic Rewards residuals
Figure 7: Residual effects of Extrinsic Rewards residuals on intention to exit.
Table 14: Results for Extrinsic Rewards residuals
Estimate Std. Error t-value
Intercept 2.55 0.08 29.99
Intrinsic Rewards residuals -0.67 0.10 -6.86
Estimate Std. Error t-value
Intercept 2.55 0.10 25.80
Extrinsic Rewards residuals -0.18 0.13 -1.36
28
According to table 13, we find that even after controlling for the effect of extrinsic rewards, intrinsic rewards significantly explain intention to exit (t-value: -6.86, p<0.01). On the other hand, when controlling for the effect of intrinsic rewards, extrinsic rewards is not significant anymore (Table 14, t-value: -1.36, p>0.05). This explains why when both independent variables are considered at the same time, only intrinsic rewards has a significant effect, while if each variable is considered independently both are found to have a significant effect on intention to exit (See Table 8). This means that intrinsic rewards have a higher predictive value on intention to exit than extrinsic rewards. In other words, intrinsic rewards are sufficient to explain intention to exit whereas if only extrinsic rewards are taking into consideration, intrinsic rewards can still provide information for predicting intention to exit.
4.4. SUMMARY OF EMPIRICAL RESULTS
Overall, support is mainly found for H2 which suggests that higher intrinsic rewards reduce franchisee’s intention to exit the system. Table 15 provides an overview of the empirical results based on the various tests performed.
Hypotheses
Empirical Results
PLS SPSS Isolated test
Partial Correlation
H1: Higher extrinsic rewards reduce franchisee’s
intention to exit the franchise system. No No Yes No
H2: Higher intrinsic rewards reduce franchisee’s
intention to exit the franchise system. Yes No Yes Yes
H3: A higher level of a franchisee’s extrinsic
motivation strengthens the relationship (i.e. becomes more negative) between subjective financial
performance and franchisee’s intention to exit the franchise system.
No No No No
H4: A higher level of a franchisee’s intrinsic
motivation strengthens the relationship (i.e. becomes more negative) between job satisfaction and
franchisee’s intention to exit the franchise system.
No No No No
Table 15: Summary of results
29 5. CONCLUSION AND DISCUSSION
5.1. CONCLUSION
The main objective of this study is to develop knowledge in the area of franchise exit by taking into account franchisees’ heterogeneity based on their individual perception of rewards and their motivations. The main finding suggests that higher intrinsic rewards reduce franchisee’s intention to exit the franchise system. In other words, the lower a franchisee experiences job satisfaction, the more likely he/she will consider exiting the system. This means that intrinsic rewards (i.e. job satisfaction) is the only factor that reduces franchisees’
intention to exit the system. Even though franchisees are heterogeneous in how they are strongly extrinsic and intrinsic motivated, they all consider intrinsic rewards more important than extrinsic rewards. Moreover, there is no support for the effect of extrinsic rewards (i.e.
subjective and objective financial performance) on reducing franchisees’ intention to exit which suggests that they do not matter over and above intrinsic rewards. However, additional analysis shows that when intrinsic rewards are excluded, extrinsic rewards have a negative and significant effect on intention to exit. This suggests that when they are considered individually, each one can explain intention to exit, but intrinsic rewards have a stronger effect on predicting intention to exit than extrinsic rewards.
Concerning the moderating effects, when using product indicator approach in SmartPLS, none of the proposed moderating effects are significant. This finding suggests that motivations do not moderate the relationship between rewards and intention to exit. This finding is in line with current studies which also highlight that intrinsic rewards are more important than extrinsic rewards (Mottaz, 1985; Huang and Van der Vliert, 2002; Stumpf et al., 2013), but surprisingly this is the same across franchisees with different level of intrinsic and extrinsic motivations. On the other hand, when performing a linear regression in SPSS to test the moderating effects, two of them appear significant, but they are in the opposite direction of what is expected. Franchisees with weak extrinsic motivations are more strongly affected by an increase in extrinsic rewards than franchisees with strong extrinsic motivations.
This finding can be analyzed through the lenses of franchisees’ expectations. For example, a franchisee, who does not care about money and does not expect it, receives a bonus of
€1000.
This is an unexpected benefit that makes the franchisee feels content with the system and very
strongly lowers his/her intention to exit. This surprisingly unexpected reward makes him/her
value it more than a franchisee who cares about money (strong extrinsic motivation) and is
30
expecting a bonus of €3000, but instead he/she obtains €1000. This franchisee has higher expectations and since they are not met, he/she gets disappointed. Similarly, franchisees with strong extrinsic motivations are affected by an increase in intrinsic rewards. For instance, they find money very important, but they realize that they enjoy their work and obtain job satisfaction. This unexpected reward makes them to continue performing their job and their intention to exit declines. These results need to be interpreted with caution, but they do show interesting things.
This study is the first to consider how the degree of extrinsic and intrinsic motivations impact franchisees’ needs to obtain extrinsic and intrinsic rewards. In addition, the most significant streams of research have examined franchise exit through the lenses of Agency and Resource-Based, and Life Cycle theories, so this study moves away from those theories by instead applying the Motivational and Self-Determination theories which provide a deeper insight of franchisees’ characteristics. Moreover, both objective and subjective performance are used to measure extrinsic rewards. Even though subjective performance has a higher negative correlation with intention to exit than objective performance (See Table 5), no clear difference were found between the effects of the two on intention to exit.
5.2. DISCUSSION
5.2.1. THEORETICAL IMPLICAT IONS
This study contributes to current literature by stressing the importance of intrinsic
rewards to explain franchise exit. Research regarding franchise exit needs to be focused on
intrinsic rewards (i.e. experience of work as meaningful, the ability to exercise some degree of
choice, the experience of progress, and the development of a greater sense of competence)
since they contribute to franchisees getting a positively valued experience from doing their
work (Spreitzer et al., 1997; Thomas and Tymon, 1994). A focus on intrinsic rewards
highlights that work can be its own rewards contributing to a sense of growth, fulfillment, and
empowerment. Therefore, this finding is in line with current studies which also stress that
increasing intrinsic rewards can reduce employee dissatisfaction and withdrawal (Stumpf et
al., 2013). In addition, the way of measuring the moderating effects has an influence in the
results. When using PLS approach, the moderating effects are not significant whereas when
using SPSS approach, two moderating relationship are significant, but they are in the opposite
direction of what it was proposed. This suggests that it is important to perform various checks
in order to verify results and before drawing conclusions.
31 5.2.2 MANAGERIAL IMPLICATIONS