Abstract: This study examines reasons why multiple unit franchising arrangements are adopted from a franchisor perspective. Based on the resource constraints and agency literatures a set of seven hypothesis explaining the choice of multiple unit franchising from a franchisor’s perspective are derived and subjected to empirical assessment. The analysis is carried out using a sample of 114 Australian franchisors. Significant differences between prior explanations of the motivational incentives governing multiple unit adoption and practice within the Australian franchising sector were found. Results indicate that franchisors appear to approach the selection, recruitment and management of multiple unit franchisees in an ad hoc fashion that may have a deleterious effect upon the ongoing management of the franchising relationship. Implications for managerial action and future research direction are discussed.
INTRODUCTION
Multiple Unit Franchising (MUF) is an organizational arrangement in which franchisees are permitted to own and operate more than one unit within the same franchise system (Grunhagen and Mittelstaedt, 2005). Several negative and positive aspects of multiple unit franchising have been proposed. For example, multiple unit franchising necessitates the employment of unit level managers and resultant adverse selection and moral hazard effects may have a deleterious effect upon operational efficiency (Shane, 1995). However, multiple unit franchising has been described to have positive associations with system growth rates, system-wide adaptation to competition, channel member communication and system uniformity (Dant and Gundlach, 1999; Bradach, 1995; Kaufmann and Dant et al., 1996). This may explain, why over half of all franchise units are owned by multiple unit franchisees in the US (IFA Educational Foundation, 2002), but does not explain why similar rates of acceptance are not observable in Australia (Frazer et al., 2006). Therefore, the conventional wisdom explaining multiple unit adoption may not capture the full range of incentives governing multiple unit adoption in the Australian franchising sector.
This study adopts an agency-theoretic perspective to explain this franchising choice from the franchisor’s perspective. Following a review of the literature, an explicit set of seven hypothesis is presented. The hypotheses are empirically evaluated with survey data collected from a sample of franchisors. This research makes a significant contribution to theory and presents important implications for practice by integrating known concepts with newly identified motivational incentives that drive or inhibit the adoption of this organizational arrangement.
Company Owned Units Versus Franchising Choice
Two important theories have been presented to explain a franchisor’s
choice of franchising as a method of expansion rather than growth through company
owned units. The first, resource constraints theory, endorses the choice of
franchising as a method of extending capital (Combs and
Ketchen, 1999; Oxenfeldt and Kelly, 1968-1969),
labor (Norton, 1988a, b) managerial
(Combs and Castrogiovanni, 1994) and local market knowledge
(Minkler, 1990) limitations on future growth. The more
accepted theory, agency theory, explains the choice of the franchising form
as an effective response to agency and motivational problems associated with
geographically dispersed units (Brickley and Dark, 1987).
Under this theory franchisors judge franchisees, as residual claimants, to be
more highly motivated to follow their decisions and policies than would employee
managers (Michael, 1996; Wattel,
1968; McGuire and Matta, 2003). Given that multiple
unit franchising arrangements conceptually mirror company-owned operations,
most prior research has adopted an agency framework in explaining this organizational
arrangement (Garg et al., 2005).
Multiple Unit Franchising Versus Single Unit Choice
The advantages of multiple unit franchising has been assessed in relation
to the role of capital acquisition, agency cost minimization and upstream price
competition (Kaufmann and Dant, 1996; Kalnins
and Lafontaine, 2004). However, past studies may not capture a full range
of motivations and some results appear to be inconsistent with what is observed
in practice. For example, many larger and older franchisors favor single unit
growth strategies (Frazer and Weaven, 2004; Frazer
and McCosker, 1999), which contrasts with the findings of research conducted
in the United States (Grunhagen and Mittelstaedt, 2005;
Garg et al., 2005). Given this disparity, a thorough
investigation into the reasons that justify the existence of multiple unit franchising
forms from the perspective of the franchisor is warranted. Therefore, this research
investigates the research question:
| • | What are the motivational incentives and factors that influence franchisors to adopt MUF in Australia? |
Franchise System Maturity
Earlier research suggests that more experienced franchisors apportion less
risk to the development of mini-chains within their system than the managers
of systems in the early stages of their life-cycle (Kalnins
and Lafontaine, 2004; Bradach, 1995). Less experienced
franchisors should perceive multiple unit franchisees as entailing greater risk
than owner operators as the formation of a relationship with an unreliable party
could do more harm to the system than single unit failures (Dant
and Nasr, 1998; Kaufmann and Dant, 1998). Although
the size of franchisee-owned mini-chains in Australia may negate double marginalization
inefficiencies (Lafontainne and Slade, 2001), managers
of firms at an early stage of development may be inclined to conclude that the
advantages associated with rapid capital access and market penetration, are
outweighed by the risk of failure. Therefore, it is predicted that:
H1.1: There is a positive relationship between MUF and franchise system maturity.
Franchise System Corporatization
Given that maturing and therefore increasingly professional systems engage
in multiple unit franchising (Robicheaux et al., 1994;
Kaufmann and Kim, 1995), it is reasonable to assume that
the standardization of functional aspects of the firm’s operation would
be both necessary and advantageous to a franchisor adopting this growth strategy
(Floyd and Fenwick, 1999). Franchise system corporatization
refers to the process of turning a proprietor-led business into one that is
governed by a formal management structure, defined by skill sets, job roles
and strategic planning (Weaven and Frazer, 2007a). It
follows that these organizational characteristics will influence the perceived
attractiveness of the franchise system to potential incumbents and govern the
strategic management of the firm and the selection, recruitment, training and
ongoing support given to chain franchisees (Weaven and Frazer,
2007b). Thus, the above arguments give rise to the following hypothesis:
H1.2: There is a positive relationship between MUF and the degree of franchise system corporatisation.
Plurality of Distribution
Prior research shows that plural distribution (a combination of franchised
and company-owned units) is effective in providing additional information to
franchisors (Bradach and Eccles, 1989), maintaining
control and consistency (Lafontaine, 1992) and ensuring
product and service quality (Scott, 1995). Given that
multiple unit structures use employee managers within the subsystem mini-chains
(Garg et al., 2005), the transference from company
to multiple unit ownership should encompass fewer transitional costs. In addition,
company-owned units promote the refinement of existing routines and franchisor
management practices (Yin and Zajac, 2004; Sorenson
and Sorensen, 2001). As multiple unit franchisee operations approximate
franchisor operations (Kaufmann and Dant, 1996; Dant
and Gundlach, 1999), it may be beneficial for franchisors to maintain company
operations to inform mini-chain operations in areas such as spans of control,
staff recruitment and performance evaluation processes and operations reporting
schemes (Garg et al., 2005). However, there is
some evidence that company ownership by Australian-based franchisors is significantly
lower than reported overseas (Pricewaterhouse Coopers, 2004;
Frazer et al., 2006). Therefore, an examination
of the relationship between dual distribution arrangements and multiple unit
ownership is warranted:
H1.3: There is a positive relationship between MUF and franchise systems characterized by plural forms of distribution.
Intra-Firm Conflict
Informal provisions within the franchising contract, together with the existence
of bargaining, imply that conflict exists in franchising arrangements (Dant
and Schul, 1992). As multiple unit franchisees tend to strictly replicate
franchisor management guidelines and processes (Kaufmann
and Dant, 1996; Bradach, 1995), this informality may
be negatively perceived as a weakness in the investment opportunity. In addition,
the existence of conflict may indicate a franchisor’s ineffectual management
of the franchising relationship (Baucus et al., 1996;
Hunt and Nevin, 1974; Lewis and
Lambert, 1991). Should franchisors be unable to enlist the cooperation of
existing franchisees to operationalize their goals then it would be unlikely
that the franchisor would allow them to increase their holdings (and power)
within the franchise system.
Furthermore, franchisors that were unsuccessful in realizing small scale economies at the unit level would be less willing to subjugate administrative power to new multiple unit franchisees, unless they could confidently predict the future success of these franchisees. As the existence of conflict may be indicative of poor selection and recruitment processes (Jambulingam and Nevin, 1999), it would be reasonable to conjecture that these processes would be ineffective in choosing suitable multiple unit candidates. Thus, it is hypothesized that:
H1.4: There is a negative relationship between MUF and the level of conflict within the franchise system.
Geographic Dispersion of Units
From an agency perspective, one would assume that even if a multiple unit
franchisee’s goals were perfectly aligned with that of the franchisor,
at some level of organizational complexity (Hall, 1977),
monitoring information would become inaccurate (Brickley
and Dark, 1987). Therefore, a franchisor would most likely favor sales of
contiguous units to a multiple unit franchisee. This action would promote monitoring
economies within the chain organization by reducing the geographical dispersion
of franchisee units. Although, empirical research in the United States supports
this contention (Kalnins and Lafontaine, 2004), some
franchisors in Australia appear reluctant to sell units in close proximity (especially
within urban localities) due to quality benefits associated with intra-system
competition issues (Bolton, 2002). This has some support
in the organizational governance literature examining the relationship between
service quality and intra-system competition in asymmetrical exchange relationships
(Sorenson and Sorensen, 2001), but to date has not yet
been examined in the Australian context. In an attempt to clarify this apparent
contradiction in franchising theory and practice it is hypothesized that:
H1.5: There is a positive relationship between MUF and the availability of geographically contiguous franchisee units.
Reward Strategy
It is possible that multiple unit arrangements enable franchisors to exert
motivational control over sequential franchisees wishing to expand. This may
give them leverage in control of quality deviations that could result in performance
decay and concept deterioration (Bercovitz, 2004; Kaufmann,
1990; Kaufmann and Kim, 1995). Bercovitz
(2004) suggested that multiple unit franchising strategies are effective
in reducing opportunistic franchisee behaviours by raising ex-post rent expectations
within self-enforcing franchising contracts (Williamson, 1985).
However, Australian franchise systems do not favor sequential modes of expansion
due to issues surrounding managerial and administrative capability and experience
(Weaven and Frazer, 2007a). However, it appears plausible
that the lure of multiple unit ownership would be beneficial in manipulating
franchisee commitment to the system resulting in the minimization of shirking
and free-riding behaviours. Thus, it is hypothesized that:
H1.6: There is a positive relationship between MUF and system reward strategies.
Franchise System Growth
The chief advantage associated with multiple unit expansion is its association
with accelerated system growth, increases in profitability and long term survival
(Bradach, 1995; Kaufmann and Kim, 1995;
Kaufmann and Dant, 1996). However, the literature on
multiple unit franchising and system growth is sometimes unclear. For example,
research by Kaufmann and Dant (1996) suggested that
multiple unit franchising arrangements do not handle agency related expenses
(such as adverse selection and moral hazard) as well as single unit expansionary
strategies. In contrast, recent research by Garg et al.
(2005) suggested that relative growth advantages exist for multiple unit
franchising over single forms of franchise governance in terms of generating
closer incentives alignment between agent and principal. However, most prior
research has tended to examine multiple unit growth strategies through the assessment
of a single derivation of master franchising (that is, area development agreements)
(Kaufmann and Dant, 1996) which is only receiving limited
acceptance in the Australian market. Thus, previous analyses overseas may not
fully accord with the reasons governing growth strategies and methods in the
Australian franchising sector. Thus, it is hypothesized that:
H1.7: There is a positive relationship between MUF and firm growth.
MATERIALS AND METHODS
The Franchise Council of Australia (FCA) granted permission to include questions within the Franchising Australia survey in 2004 (administered in April/May 2004 by Griffith University), which is a biennial survey conducted throughout all states and territories of Australia. The resultant data were analyzed to test a set of hypothesis of franchisor motivations to adopt multiple unit franchising arrangements. Inferential statistical tests were used to test the hypothesized relationships. From 540 questionnaires, 114 useable responses were received giving a response rate of 21%, which is similar to those obtained in earlier surveys (for example, Frazer and McCosker, 1999).
Operationalization of VariablesMultiple Unit Versus Single Unit Grouping
and Franchisee Density
Two dependent variables were developed to reflect the choice of multiple
unit systems versus single unit systems. A simple dichotomous variable was used
which simply measured whether a system contained no multiple units (hereinafter
termed the single unit group), or contained any number of multiple units (hereinafter
termed the multiple unit group). Just under a third of the sample (35 systems)
did not allow multiple unit franchising.
A second dependent measure was also developed for those systems which contained multiple units-The Franchisee Density Index. This provided an index of the degree to which multiple unit franchising had been adopted within systems containing multiple units. The Franchisee Density Index was calculated as the ratio of the number of franchisees within the system divided by the number of franchise units within the system. Sixty five systems within the sample had one or more multiple units. For the multiple unit group, the index ranged in value from 0.10 to 0.99 (Mean = 0.78, SD = 0.23). Both skew and kurtosis were found to significantly deviate from normality. Procedures described by Tabachnick and Fidell (2001) were followed. The variable was first reverse scored and then a square-root transformation applied. Skew was still significant so a log transformation was applied. Following this transformation the value for both skew (-0.16, p>0.01) and kurtosis (-0.42, p>0.01) were found to not significantly deviate from normality. The variable was then again reverse scored to maintain the rank order of the original untransformed variable.
Franchisor Experience
This variable was operationalized as the number of years the franchisor
had been franchising. The tests of normality on the group data showed no significant
deviation from normality.
Degree of Corporatisation
Respondents were asked to indicate if they had a company board of at least
five members, had two or more appointed board members who were external to the
firm, had a marketing department, or had at least one franchise on the board.
A composite variable was produced as the sum of responses across these four
dichotomous variables with a possible range in values from 0 to 4.
Plurality of Distribution
This variable was operationalized as a dichotomous variable by grouping
systems as those that either did or did not contain company owned units.
Intra-Firm Conflict
This was a dichotomous variable indicating whether or not the franchisors
were in conflict with any franchise in the system.
Geographical Dispersion of Units
Units were coded on an ordinal scale: 0 = Neighboring suburbs, 1 = same
city and 2 = not same city.
Reward Strategy
Dichotomous variable indicating whether or not the franchisors viewed rewards
as important.
System Growth
This index was calculated by taking the difference in the number of franchise
units in the system between the survey year and the value for two years earlier
divided by the base level of the survey year. Due to significant levels of both
skew and kurtosis, the variable was reverse scored, the log taken and the variable
reverse scored again (Tabachnick and Fidell, 2001). This
produced a transformed variable with acceptable levels of skew (skew = -0.676,
p>0.01) but which still exhibited a significant departure from normality
with respect to kurtosis (kurtosis = 7.29, p<0.01).
RESULTS
Descriptive Statistics
Slightly more than half of the respondents (57%) reported that they encouraged
suitable franchisees to own multiple units within their system and 41 (36%)
reported that they granted area development rights to franchisees. Nearly one
third of the franchisor respondents (32%) reported that their franchisees had
between one and three years’ franchising experience before they were granted
additional units within the franchise system. Overall, more than half of the
franchisors (52%) felt that multiple unit franchisees required the same level
of support as single unit franchisees.
Inferential Statistics
The results of the inferential tests of hypotheses are provided in Table
1. For each hypothesis inferential tests were performed in two ways. First,
comparisons were made of single unit systems versus systems that contained multiple
unit franchisees. These comparisons allowed tests of whether the hypothesized
independent variables had any influence on the choice by franchisors to allow
multiple unit franchising within their system. Second, for the systems containing
multiple unit franchisees, the relationship between Franchisee Density Index
scores could be made with other variables. These comparisons allowed tests to
evaluate whether the hypothesized independent variables had any influence on
the extent to which multiple unit franchising was adopted once the decision
had been made by the franchisor to allow multiple unit franchising within their
system.
| Table 1: | Results of Inferrential Tests of Hypotheses a |
aTwo comparisons are made for each hypothesis using one of two independent variables. (a) The dependent variable is a dichotomous variable which groups franchisors as either single unit indicating there to be no multiple units in their system, or as multiple unit, (b) The dependent variable is metric, the Franchise Density Index (FDI) and is calculated for franchisors in the multiple unit group only, bDifferent results are shown in the pattern of results as each variable utilized different measurement scales (as specified in the materials and method section). N/A: Not applicable | |
The specific inferential test used to test each hypothesis varied as a function of the scale of measurement of the two variables in a specific analysis. When both variables were non-metric, a contingency table (Chi-square) analysis was conducted. When one variable was metric and the other non-metric, an Analysis of Variance (ANOVA) was conducted. When both variables were metric a Pearson product-moment correlation was conducted. The results of these analyses are provided in Table 1.
Organizations within which multiple unit franchising was occurring had been operating as franchisors for a significantly longer period (Mean = 11.13 years) than had organizations where only single unit franchising was reported (Mean = 6.35 years), F (1,96) = 13.53, p<0.0005. This result supports hypothesis H1.1. Partial support was also found for this hypothesis within the multiple unit group with a correlation between years of franchising and the franchise density index approaching significance (r = 0.243, p = 0.05).
Significant relationships were also found between franchising and intra-firm conflict as predicted by hypothesis H1.4. The effects, however, were in the reverse direction to that hypothesized. A contingency table analysis found there to be a significant relationship between intra-firm conflict (in dispute/not in dispute) and system grouping (single/multiple), χ2 (1, n = 100) = 5.51, p>0.05. Only 6% of franchisors from single-unit systems reported there to be conflict between the franchisor and one or more of the franchisees, whereas for multi-unit systems, 25% of franchisors reported the existence of conflict. This pattern was also found with respect to variation in the Franchise Density Index across m ultiple unit systems. A significant effect was found for intra-firm conflict, F(1,63) = 7.70, p<0.01, with higher mean density scores being displayed for systems which are in conflict (Mean = 2.71) compared with systems which were not in conflict (Mean = 1.87). No other significant relationships were found.
DISCUSSION
This study explored franchisor incentives for multiple unit franchising and the implications arising from the study are discussed below. The results from the quantitative analysis found support for the proposed link between franchise system maturity and multiple unit franchising (H1.1) and is consistent with previous research (Robicheaux et al., 1994; Kaufmann and Dant, 1996; Wadsworth, 2002). Unlike systems in the early stages of their life-cycle, more experienced franchisors are likely to appeal to new franchisees as they have an established reputation and brand, proven concept and track record of performance and managerial acumen. In addition, multiple unit franchising may be more popular in older franchise systems due to the needs of incumbent franchisees. In practice, franchisors may attempt to promote franchisee satisfaction and maximize unit level performance through allowing multiple unit ownership.
The proposed relationship between the degree of franchise system corporatisation and multiple unit franchising was not supported (H1.2). This lack of support may indicate that franchisors that encourage multiple unit ownership do not necessarily perceive managerial and functional structural differentiation as important in attracting investment from prospective and existing franchisees. Most franchisors continue to favor sequential modes of multiple unit expansion (Frazer and Weaven, 2002; Frazer et al., 2006) which contrasts markedly with United States franchisors (Kaufmann and Dant, 1996). The founders of many franchise systems may be reluctant to encourage a process of corporatisation as it may reduce their power and levels of personal control within the system and may believe that an autocratic management style is necessary to maintain decision-making flexibility, especially in an environment characterized by high demand uncertainty.
There was a similar lack of support for the agency costs argument (H1.3). One possible explanation for this lack of support would be that franchisors are unaware of the complementary benefits of higher innovation and coordination gains associated with dual-distribution networks (Michael, 1996; Bradach, 1998; Cliquet, 2001; Lafontaine and Shaw, 2001). However, franchisors that encourage multiple unit ownership may feel that the maintenance of system-wide standards are less important in companies with multiple unit ownership as franchisees are less likely to free-ride off other units within their own subsystems (Gal-Or, 1995; Dant and Gundlach, 1999).
Although, a negative association between intra-firm conflict and multiple unit franchising was proposed (H1.4), a significant positive relationship was found. Franchisors find it difficult to attract suitable franchisee candidates (Frazer et al., 2006) and may find it easier to encourage existing single unit franchisees to become multiple unit holders, even though these franchisees may not have the managerial skills, knowledge or financial resources to commit to additional units. This may minimize ex-ante costs associated with the recruitment and selection of candidates, but may have a deleterious impact upon the future operational efficiency of franchisee-owned subsystems.
The positive relationship between the geographic proximity of franchised units and the use of multiple unit franchising strategies was not supported (H1.5). This finding is inconsistent with prior research conducted overseas (Kalnins and Lafontaine, 2004) suggested that Australian franchisors’ operational strategies are unique. Franchisors appear to actively discourage ownership of geographically close units, perhaps as it negates the benefits associated with the maintenance of beneficial levels of benchmarking (Sorenson and Sorensen, 2001) and competition between incumbent franchisees (Brown, 1998; Bolton, 2002).
Similarly, the proposed relationship between system reward strategies and multiple unit franchising (H1.6) was not significant which is in contrast to previous research reported overseas (Bercovitz, 2004). Franchisors may be reactive rather than proactive in allowing successful franchisees to increase their unit holdings in the system. Although, the granting of additional units may be viewed by franchisees as rewards for good performance, franchisors may not explicitly offer them as rewards.
Whereas earlier research has confirmed the relationship between multiple unit franchising and system growth (Kaufmann and Kim, 1995; Kaufmann and Dant, 1996), hypothesis H1.7 was unsupported in this research. Whilst some multiple unit arrangements in the United States encompass many hundreds of outlets (Grunhagen and Mittelstaedt, 2002), Australian systems are typically smaller (typically 1-2 units, Frazer and Weaven, 2004) and may be unable to satisfy size requirements in attaining operational efficiencies. Earlier research suggests that franchisee-owned mini-chains incorporating only two or three outlets may suffer from performance decay as the potential for operational synergies (in purchasing, management and administrative costs) are not realizable in the short term (Kaufmann, 1992). Future research should examine optimum unit allocation strategies within multiple unit franchising arrangements (and within the context of different strategic priorities (Garg et al., 2005; Bradach, 1995) so, as to identify methods of maximizing mini-chain efficiency.
One limitation of this study arises as a consequence of bias resulting from self-complete surveys. In particular, measurement error may result from inaccuracies in a respondent’s reporting of data. In addition as data were collected from Australia, the generalizability of results may not extend beyond this region. Replication in other global settings is recommended to resolve this limitation. Furthermore, non-response bias could not be tested due to the anonymous nature of the survey.
This study has important implications for practitioners. The results suggest that Australian franchisors have unique motivations for encouraging franchisee subsystem development. Australian franchisors appear to utilize ad hoc methods of expansion n an attempt to grow system units as their compensation is directly linked to growth in sales. More mature systems appear to encourage multiple unit ownership, although, the results suggest that this may not be due to their ability to offer structural arrangements conducive to attracting external investment, but may be due to a willingness to apportion less risk to this form of expansion. While, this research hypothesized strategic motives for this endeavor, the real reasons why multiple unit arrangements exist and continue to grow, may be due to a franchisor’s desire to grow the size of their system, or more realistically, just survive. However, this may prove to be a non-sustaining strategy due to the increasing level of substantial disputation in systems characterized by multiple unit ownership.
Although, there was minimal support for the hypothesized relationships, this research may be useful in providing some preliminary answers as to when to engage in multiple unit franchising, who should engage in multiple unit franchising and why franchisors should encourage the growth of franchisee-owned subsystems. This should help minimize franchisee and franchisor dissatisfaction associated with the adoption of hybridized franchising forms based upon operational and structural incompatibility and competing expectations. This, in turn, should minimize failure and result in the growth of workable and efficient franchise systems. In addition, the results could be used to inform managers of the expansionary outcomes associated with a given proportion of multiple unit franchising within the context of factors moderating acceptance. Managers can use this information to determine how compatible these determinants are within the context of firm-level strategic priorities.
CONCLUSION
The research findings infer that the conventional wisdom explaining multiple unit franchising may not adequately explain the reasons influencing the growth of franchisee-owned mini-chains Australian franchising systems. In particular, the inconsistency in the results of this research and other studies suggests that sectoral and firm-level differences between countries may impact upon the popularity and successful management of different types of franchisee-owned mini-chains. Although, prior research posits positive agency and growth - related outcomes with multiple unit adoption, Australian franchisors’ growing endorsement of this strategy may reflect current economic conditions characterized by difficulties in attracting suitable franchisee candidates and realizing minimum efficient scales of operation.
ACKNOWLEDGMENT
The author would like to thank Griffith University and the Franchise Council of Australia for their assistance in this research project.