Table 2 : Logit Analysis of Relational Competences
Paper presented at the EMNet conference
Budapest, Hungary, September 15-17, 2005
(
Market, hierarchy and knowledge transfers: what franchising networks tell us
Patrick RONDE
GRAICO , Université de Haute-Alsace, IUT de Mulhouse, 61, rue Albert Camus, 68093 Mulhouse Cedex, France,
e-mail :
Caroline HUSSLER
RECITS, Université de Technologie de Belfort-Montbéliard, 90010 Belfort Cedex, France
E mail:
Revised version 7/10/2005
A large part of the text has been modified (at least to improve the phrasing) starting with the title itself, but the major changes concern parts 1.1 and 2.1
Abstract:
Our paper aims at empirically vindicating the Knowledge Based View (KBV) of organisations. Concretely, we analyse knowledge transfers between franchisers and franchisees and compare the franchisees’ architecture of knowledge with the architecture of independent shops in the same sector. We examine whether franchisees (coordinated through a centralised franchising network) and independent shops (coordinated through the market) exhibit similar competences and whether the franchisees’ competences coincide with the knowledge held by franchisers.
The car maintenance industry in France is selected as our empirical ground. We find that sharing goals does not guarantee better knowledge transfers and that specific competences held by franchisees do not necessarily reflect franchisers’ view, the KBV of organisations being therefore only partially advocated.
Introduction
A number of recent contributions to the theory of the firm develop a knowledge perspective on organisations and present firms as loci of knowledge creation, combination and accumulation (Cohendet and Llerena, 2003; Grant, 1996). This new conceptualisation of firms and organisations breaks away from the contractual approach of firms based on notions of opportunism and incentives (Williamson, 1985; Demsetz, 1988), and puts the concepts of competence (Barney, 1996) and knowledge (Conner, 1991; Tsoukas, 1996) at the heart of the analysis of firms’ existence and growth. However seducing at the time of the knowledge based economy (Lundvall and Maskell, 2000), this new knowledge based theoretical framework still suffers from a lack of empirical validation (Poppo and Zenger, 1998). This paper precisely aims at filling this blank by empirically vindicating the Knowledge Based View (KBV) of the firm. As a way of supporting the validity of the KBV theory, we attempt to identify the systematic differences in architectures of competences between franchisees and independent shops in the car maintenance industry in France.
If we follow the KBV literature, franchising (as a specific form of hierarchical organisation) accounts for a large and growing share of business because of better knowledge transfers (Conner and Prahalad, 1996) and also because of shared identity and goals (Kogut and Zander, 1996) among the networking actors. One should therefore observe that franchisees master competences, which are not only different from the ones held by independent shops, but which also reflect the franchisers’ opinions about best organisational practices.
The paper begins with a more detailed development of the Knowledge Based View of organisations, insisting on its implications for knowledge and competences within franchising networks. In a second section, we present the empirical context of the study and the methodology we adopt. The results and interpretation sections are then provided, before concluding on the relevance of franchising networks for knowledge transfers.
- The knowledge based view applied to franchising networks: basic concepts and assumptions
1.1 Franchise as a centralised and partially hierarchical organisation
In our empirical analysis we want to compare the performances and the effectiveness of knowledge transfer within two different structures: the franchise structure and the market structure.
The franchise structure is composed by franchisers (Z) and franchisees (Y), who are viewed as a single entity (a specific type of hierarchical organisation) in the present analysis: the Y-Z entity. Indeed, for us, the franchise structure of governance is based on (at least partially) centralised relationships, which is even more specifically the case in the car maintenance industry. In this sector, franchisees have limited choices in terms of suppliers’ selection, price policy etc. Actually, franchisers prescribe a list of potential suppliers, and franchisees only have the possibility to select their own suppliers on that list. Moreover, the range of services offered by the shops are dictated by franchisers, and franchisees can only provide car repair according to the “authorized” tasks. In a word, if bottom-up flows of information (and more precisely of marketing information) are welcome and supported by franchisers, most management tools and operational choices are imposed by franchisers upon all franchisees within the network, as it is the case in hierarchical and centralised forms of organisation.
On the other hand, we distinguish a market structure, which is characterised by independent shops connected with some important “trade names”, the latter sub-contracting “some small and well established” technical operations to the former. In our study, independent shops and trade names constitute two distinct entities interacting through the market: Y + Z.
Our final objective is then to compare and hence to be able to measure knowledge within two types of organisations: Y-Z (the hierarchical organisation[1]) and Y+Z (the market structure). But why should there be any difference between Y’s knowledge and competences when Y belongs to an (Y-Z) structure or to a (Y+Z) structure? If we follow the theoretical arguments developed by knowledge based (KB) scholars since the early 1990s, one possible explanation of those differences in competences lies in differences in terms of knowledge transfers among partners. To investigate this argument in more depth, we decide to concentrate our analysis on the concepts developed in two main contributions of the KB stream of research, namely Conner and Prahalad’s work (1996) and the papers by Kogut and Zander (1992, 1996).
1.2. Conner and Prahalad (1996)’s hypothesis applied to franchising networks
The starting point of Conner and Prahalad is the concept of bounded rationality popularised by Simon’s seminal work, which suggests that different agents have different but limited cognitive abilities. This implies that the stock of knowledge of each agent is different from the one of his/her neighbours. Therefore different agents do not react in the same way. For Conner and Prahalad, one of the advantages of hierarchical organisation precisely lies in what they call the knowledge substitution effect. “Knowledge substitution effect takes place when Z tells Y to do something that Y, relying on its own knowledge and that which it has managed to absorb from Z, otherwise would not do” (Conner and Prahalad, 1996, p 485). As the hierarchical organisation creates the necessary authority for knowledge substitution, two costs are saved in this organisational form: the cost of negotiating Y-Z market contracts and (at least partially) the cost of Y internalising the knowledge of Z, which can explain the existence and the growth of firms. If one considers franchise as a hierarchical organisation, applying Conner and Prahald hypothesis entails that franchisees (Y) should adopt behaviours and develop competences that they would not have developed had they been independent. Hence French law (judgement of the"Cour d'Appel de Paris" of 29/9/1992) stipulates that the franchiser has to provide a “specific,substantial, original, secret, conceptualisedandtransposableknowledge” to the franchisee. It can be a question of technical know-how (changing a car muffler) and/or of know-how of exploitation (commercial, organisational or managerial knowledge). But the same judgement specifies that it is possible to speak about “specific knowledge” only if it is special compared to the knowledge traditionally used, if it supplies a competence, a specific and original skill, which appeals to the customers. The existence of those specific competences constitutes the core of the franchiser’s legal duties, the notoriety of the trade name being another incentive to enter a franchising network (Willams, 1998). Consequently, entrepreneurs taking part in franchising networks save the cost of identifying and developing those specific competences, contrary to their independent counterparts which have to discover and learn the tricks of the trade on their own. Finally, Conner and Prahalad ‘s knowledge substitution effect applied to franchising networks raises a first hypothesis which needs to be tested:
Hyp 1: Because of the knowledge substitution effect, the organisational competences developed by franchisees should differ from the ones mastered by independent shops, the former benefiting from the knowledge and experience of the franchisers
1.3. Kogut and Zander ‘s analysis of franchising
If we look at Kogut and Zander’s point of view , we find another knowledge based explanation of the efficiency of hierarchical organisations and firms. Their analysis is grounded on more sociological, psychological and philosophical arguments. The basic idea is that a firm can be understood as a social community specialised in the creation and transfer of knowledge. As summed up by Foss (1996, p 519), Kogut and Zander define firms as “normative communities that embodied shared identities. Because such identities are symbolically coded, they may significantly lower the costs of communication, co-ordination, and learning”. Hence firms’ superiority in terms of efficiency is a direct consequence of the presence of some “higher order organising principles” within the boundaries of the firm. Those principles include such elements as shared identity and shared cognitive models. In other words, in a social community with shared identity, one produces in a better way because one learns easily thanks to a better transmission of knowledge within the community and thanks to a convergence of objectives and methods among the members. To test Kogut and Zander’s idea of High Order Organising Principles, our empirical analysis should then bring out differences in knowledge stockbetween organisations where goals and understanding are shared between members, and organisations in which this community of goals does not prevail. According to the knowledge based view, such differences reflect quicker knowledge transmission and better learning mechanisms in the first category of organisational structures.
The second hypothesis we want to test is thus the following:
Hyp 2: Organisational structures characterised by an efficient knowledge transmission among members are also those in which members share goals and identity.
To sum up, the first objective of the paper is to show whether or not we observe different competences in Y-Z organisations and in Y+Z organisations. Then, we test whether those differences in terms of competences are the result of better learning processes and knowledge transmission, ie whether Y’s competences the reflect Z’s knowledge base in the franchise case. Finally, we will try and investigate whether we can findshared identity and goals in Y-Z organisations that we do not find in Y+Z organisations.
- Empirical analysis
2.1. Car maintenance sector
We choose to test our conceptual framework on the car repair sector in France. This choice is motivated by several reasons. The most important one is that franchise in this sector is highly hierarchical and centralised, allowing us to use it as our proxy of firms’ structure in opposition to market structure, as explained in the previous section. Moreover, this activity is characterised by a similar proportion of franchising networks reflecting the firm (Y-Z) solution and Trade Name / Independent shops reflecting the market (Y+Z) solution. Indeed, we recorded 62 franchisees and 75 independent car repair shops in France. This constitutes a second significant motivation for selecting this sector as empirical ground. This sector also sounds interesting for our study in the sense that it is not a knowledge intensive activity, which simplifies the analysis of knowledge transfers. Lastly, this sector has a good mix of codified and tacit knowledge. Indeed, all the mechanics do have access to technical description provided by car makers (the codified part) but they have an interpretative margin in choosing the process of failure identification and then the repairing technique itself. But for all those reasons, we also do not take the risk to extend our conclusions to other sectors.
2.2 Measurement problems: from knowledge to competencies
If following Saviotti (1998), we define the knowledge base (KB) of a firm as the collective knowledge that the firm uses for its productive purpose, then the performances of firms directly depend on their capacity to identify and manage their KBs. So what is the KB of the firm? This question is important because we know (Spender, 1996) that different types of knowledge lead to different types of economic rents (“Pareto rents”, “Penrose rents”), and then different strategies. Nevertheless, despite many interesting typologies of knowledge (Lundvall and Johnson, 1994; Blacker, 1995; Spender, 1996), we face a problem of practical applicability and measure of such typologies. For instance, in the car maintenance industry, being able to change a defective car muffler clearly corresponds to Liebeskind’s (1996) “practice knowledge” (i.e. knowledge defined as information whose validity has been established through test or proof). It is also a problem of know-how in Lundvall’s typology of knowledge, which refers to skills. But there can be some difficulties associated with the car muffler replacement, if for example the bolts are rusted. Solving such difficulties requires other types of knowledge. First of all, one can use explicit and individual knowledge, “about facts” (know-what): iron can rust. Second, other explicit but collective knowledge, referring to the physics law of degradation of iron under specific conditions can also be called upon (know-why). Third, if the problem is too complex to solve, the worker might call on specific and selective social relations in order to know who knows what to do and who can do it. This corresponds to collective but tacit knowledge, built on routines or rules of conduct.
In other words, we face a major problem: how can we measure and identify knowledge within firms? To solve this problem, our analysis switches to the concept of competences defined as a specific way of packaging knowledge (Malerba, Orsenigo, 2000) which has more powerful and practical implications. Between the franchisees’ competences and the franchiser’s knowledge, a process of conceptualisation, codification, transmission and knowledge sharing takes place, which must be secret due to the definition of the franchise-networking concept. By focusing our analysis on competences directly involved in the development and application of the knowledge base, we follow Coombs and Hull (1998) who think that knowledge bases are more difficult to access than competences. In fact, the franchiser possesses different types of knowledge, but what we observe in the franchisees is their competences i.e. different packagings of different types of knowledge (essentially know-how but also know-what, know-who and know why) in order to satisfy the customers’ wishes.
To sum up, our analysis will consist in comparing the knowledge held by franchisers and trade names with the competencies mastered by respectively franchisees and independent shops.
2.3. Data collection
Franchisers’ and Trade Names’ knowledge
The original conceiving of a successful business comes from the franchiser who seeks to share his/her experience and his/her knowledge, and from Trade Names who have market contracts with independent shops for small technical operations. Therefore, we ask franchisers and trade namesfor their representation of knowledge required to be successful in their sectors (what Z wants to transmit to Y). The interviews of franchisers and trade names took place during the years 1997-1998. The sample was constituted of 50 franchisers involved in various sectors, the 10 Trade names and the 5 franchisers (Axto, Feu Vert, Maxauto, Midas and Speedy) of car maintenance in France.
Franchisees’ and independent shops’ competences
The objective is to find out whether franchisees (respectively independent shops) have specific competencies that reflect the franchiser’s (respectively Trade name’s) knowledge. We built a questionnaire on competences (cf Appendix 1) that we believe to be good proxies for all pieces of knowledge identified as crucial by the car maintenance franchisers and trade names in the first step. We administrated it to the franchisees and the independent shops in order to test the specificities of franchised shops as compared with independent ones. It was sent by post to all French car maintenance franchisees and independent car repair shops (respectively 62 and 75 shops). The final sample of respondents was composed of 33 franchisees and 20 independent companies (38 % returns).
2.4. Data processing
Determination of the knowledge franchisers and trade names consider as crucial for a successful activity
We used a two step procedure. First, we run a lexical analysis on all interviews in order to highlight the most frequently cited pieces of knowledge. Indeed, lexical analyses are regularly used in the literature when looking for the basic structure and the most important terms of a text (for a survey, see Aho, Sethi & Ullman, 1986 or Lesk, & Schmidt, 1975). In a second step, we asked the franchisers and trade names of the car repairing sector to ordinate the pieces of knowledge (identified as significant in the first step) according to their degree of importance in the success of business. They were also required to exclude the items they considered as useless and to add pieces of knowledge they considered determinant in the specific case of the car repairing industry.
Determination of the specific competences of each category of shops
The questionnaire on competences administrated to franchisees and independent shops in the car maintenance sector gathers both quantitative and qualitative answers, each category of answers being processed in a specific way.
Quantitative answers. Firms (franchisees and independent shops) are asked to evaluate the degree of importance (on a scale from 1=low to 10=high) of each quantitative competence for promoting their shops. Then, to test the difference between quantitative competences held by franchisees and those mastered by independent shops, we use a multivariate t-test with the null hypothesis that there is no mean difference between groups. The significant differences we find allow us to characterise the specific competences of franchising network members.
Qualitative answers. For other competencies, the firms surveyed only have to answer whether or not they possess them. We choose to disaggregate those competences into a set of “elementary” competences (corresponding to alternative ways of developing the aggregated competence) in order to help firms to evaluate whether or not they master the competences. A specific (aggregated) competence is then said to be mastered by a shop if the shop possesses at least one of the elementary competencies the aggregated competence is composed of (cf. Tables 3, 4, 5 for the decomposition of aggregated competences into elementary ones). Hence, firm i holds the specific competence “price policy” if it uses either “fidelity card” or “regular special prices”.