1
Management consultancy viewed from economic and sociological perspectives
Only since the 1990s has management consultancy prompted a great deal of attention in management research. Until then little had been written on this service sector, probably because it was not yet recognized as a mainstay in the economy. Management research, organization studies, and industrial sociology had primarily concerned themselves with larger industries and corporations, and the management consulting business was still too small to be recognized as an industry with considerable influence. Only a few authors, for example Hagedorn (1955), Higdon (1969), and Havelock and Guskin (1971), had begun to recognize the role of consultants in the transmission of business techniques. Other early publications on management consulting were concerned with organizational development, a consulting approach to help clients help themselves (Schein 1969; Argyris 1970).
Throughout the 1980s publications in the sociology of professions (Stanback 1979; Stanback et al. 1981; Noyelle and Dutka 1988; and later Tordoir 1995) referred to management consulting as one of the service sectors toward which industrialized economies shift. It became recognized as an emerging profession in which formal professional qualification has given way to professional work independent of a formal professional background (Abbott 1988; Brint 1994). At about the same time, Greiner and Metzger (1983) wrote a first advisory book for consultants, and the International Labour Organization (Kubr 1986) issued the second edition of a landmark book on best practices in management consulting, to which prominent management scholars and practitioners contributed and which aimed to cover a broad range of aspects from both consulting and client perspectives.
Despite these advances in the 1980s, the number of studies on management consulting remained low in comparison to the growth of the literature in the subsequent decade. Presumably it was assumed that not much could be added to the established view of consultants as transmitters of business techniques and carriers of organizational change methods. Not even the history of management consulting as a service sector and profession (McKenna 1995, 2001, 2006; Kipping 1996, 1997, 1999) was available to the scientific community before the 1990s. Only in the first half of the 1990s, following the rapid growth in the industry, did the significance and influence of management consulting become more recognized in the academic literature. Globally active consulting firms had achieved a high level of visibility, and management scholars could no longer ignore the influence of these firms on management knowledge, decisions, and practices. In the 1990s a large number of books appeared on the subject, oriented toward the markets for practitioners (e.g. Maister 1993; Kubr 1996), for MBA graduates applying to major consulting firms (e.g. Wet Feet Press 1996; Wickham 1999), or for those interested in starting their own consulting business (e.g. Kishel and Kishel 1996; Biech 1999).
At about the same time, a growing number of popular books on the potential dangers of hiring consultants appeared on the book market. These were mainly written by journalists or former consultants and had suggestive titles such as The Inside Story (Rassam and Oates 1992), Dangerous Company (O’Shea and Madigan 1997), or Consulting Demons (Pinault 2000). Even the Dilbert comics ridiculed consultants as shallow advisors. In this high tide of consulting bashing, well-known management scholars joined the ranks of those warning of Flawed Advice (Argyris 2000). Indeed, one of the salient characteristics of the consulting literature has been, and continues to be, that both journalists and academic commentators tend to have strong feelings about the business, considering consultants to be anywhere in a broad spectrum from shallow charlatans to modern carriers of economic growth.
Based on these images of the business, one can broadly distinguish between a functionalist and a critical view on consulting. The functionalist view sees consulting firms as carriers and transmitters of management knowledge. For example, Bessant and Rush (1995) distinguish between two knowledge-based roles for consultants: an intermediary one that supports clients’ acquisition of knowledge and technological developments; and a capability-building one that supports clients’ adoption and implementation of changes. Along this line, many authors have pointed out that consulting firms possess knowledge about analytical procedures which enables them to provide a variety of services and tasks that clients cannot perform on their own (Starbuck 1992; Moore and Birkinshaw 1998; Morris and Empson 1998; Sarvary 1999; Werr et al. 1997; Werr 1999, 2002; Armbrüster and Kipping 2002). Traditional organizations are assumed not to have the human resources, analytical skills, and procedural potential, with the result that taking management consultants into service has become a matter of course rather than an exceptional case, as it was some decades ago (Alvesson 1995; Faust 2002; Suddaby and Greenwood 2001). This perspective will be taken up in chapter 2 and integrated into a transaction cost perspective.
The functionalist view also points out other features of large consulting firms: the worldwide representation, the familiarity with a wide variety of industrial sectors, and the “one-firm” governance concept (for details, see chapter 8). These features ensure that consulting firms can obtain knowledge from a large variety of sources and, potentially, apply experiences gained in other industrial sectors or parts of the world. From this perspective, the methods to generate data and information outside and within the client organization constitute the primary driver of the consulting business and its growth. The recruitment of talented personnel, an extraordinary work ethic, and the strong commitment to an achievement culture represent a fundamental aspect of their performance and of the demand for their services. From the functionalist perspective, systematic knowledge management allows consulting firms to stay up to date with industry practices and market information, and it also enables them to distribute knowledge resources in a manner unequaled by conventional organizations (Larsen 2001; Hansen 1999, 2002; Hansen et al. 1999; Hansen and Haas 2001). I shall come back to these arguments in the transaction cost approach to consulting in chapters 2 and 8.
The critical literature on consulting does not necessarily doubt the usefulness of consulting for clients, but argues that the view that “consultants are experts and provide knowledge and analyses to clients for a fee” is too narrow to grasp what is going on in consulting projects (Clark and Fincham 2002). For example, Abrahamson (1996), Kieser (2002) and Ernst and Kieser (2002) refer to the faddish character of many management activities and argue that, among others, consulting firms have an economic interest in the up- and downswings of management concepts and substantially contribute to fashion setting. Berglund and Werr (2000) point to consultants’ communicative flexibility, for example in their use of rationality and pragmatism myths to legitimate their approaches. Benders et al. (1998) have done empirical work in this context, finding that consultants use the term “business process reengineering” for a large variety of services that have often little to do with Hammer and Champy's (1993) original call for radical changes. Benders et al. (1998) argue that consultants separate the label from the contents of this management concept and create a sense of urgency by using a particular term without relating project contents to it. Similarly, Fincham (1995) argues that, in particular, business reengineering is constructed and marketed as a saleable commodity in order to meet the needs of the “managerial consumer.” Ernst and Kieser (2002) and Kieser (2002) draw on these ideas to suggest that the circulation of management concepts and fashions contributes to managerial insecurity and fuels the demand for consulting services.
In a micropolitical view of consulting, Jackall (1988: 140–4) argues that consultants often trade in the troubles between the internal factions of a client organization, and that consultants often have to work on the problem as defined rather than develop a solution autonomously. As in an earlier approach by Moore (1984), client firms are not conceptualized as organizations as a whole, but as consisting of competing actors and groups. Using IT consulting as an example, Bloomfield and Danieli (1995) argue that the socio-political skills of consultants are indissoluble from their technical expertise, because technology cannot be separated from its communicative representation and thus from vested interests within a client firm. During the elaboration and implementation of advice, consultants and clients mobilize discursive and symbolic resources, which render it impossible to conduct consulting without any micropolitical involvement (see also Bloomfield and Best 1992). As with the other approaches, the micropolitical view draws on the insight that consultancy services are intangible and that their commercial impact is difficult to evaluate. But, rather than focusing on the consequential market mechanisms, the critical perspective on consultancy looks at the ways in which consulting assignments and client–consultant interactions are open to distortion.
In this context, Czarniawska-Joerges (1990) holds that the use of metaphors and labels that are new to the client organization can give meaning to situations and engender action through sense making. Seen from this perspective, the communicative resources of consultants provide some potential to obfuscate issues, to interpret situations for vested interests, or to manipulate definitions of success and failure. For Alvesson (1993), the point of departure is the uncertain character of all types of knowledge, even scientific knowledge. He argues that knowledge work needs to be viewed in the context of institutionalized myths of rationality, since there is no objectively determinable knowledge. Claims of knowledge, and therefore of communicative performance, may move into the foreground of this business, as credible stories about the world need to be delivered. The work of Clark (1995) has been influential in this respect. Given the lack of objective criteria for quality assessment, he argues, convincing clients of consulting quality requires considerable communicative skills and thus promotes consultants’ impression management and rhetorical abilities. Along these lines, Clark and Salaman suggest viewing management consultants as “systems of persuasion creating compelling images which persuade clients of their quality and work” (Clark and Salaman 1998: 18).
In summary, the critical view argues that consulting results and project achievements are too problematic to be sufficiently theorized in terms of knowledge transfer. Authors in this paradigm point to the contestable nature of consulting knowledge, to the involvement of consultants in vested interests in client organizations, and to the potentially flexible mode of “consultancy speak.” In so doing, they are expressing much of the concern, or even distaste, of an academic research community regarding consultants (March 1991), contributing to a more emancipated comprehension of the business. This critical take on consultancy will be taken up in chapter 4.
Publications of the above two types, the functionalist and the critical views, today characterize the literature on consultancy and have considerably advanced our knowledge of the industry and its mechanisms. Nevertheless, to date both are beset with limitations. The functionalist view lacks a systematic outline of why clients have increasingly externalized management services and continue to do so, and the critical view lacks an acknowledgment of economic processes and clients’ rational deliberations. More precisely, the functionalist view presents useful lists or outlines of the economic role of consulting firms, but it lacks an analytical grounding. Neither theoretically nor empirically does it engage with the question of why client firms do not perform the services themselves or hire experts as employees rather than making use of external consultancies. It has not delved into the question of how clients gain quality certainty or why they hire a particular consultancy in preference to another, and a more theoretical analysis and elaboration suggests itself.
For its part, the critical view exhibits a limitation that is at least equally serious. As Salaman (2002) points out, it is preoccupied with consultants’ truth claims, with consultants’ supposedly unscientific approaches, and with an ostensibly dark side to consultancy. It either focuses on management fashions that clients supposedly fall for – which represents an oversocialized conception of the consulting market, to use Granovetter's (1985) term – or it portrays consultants as opportunistic agents who exploit clients’ lack of quality certainty – which represents an undersocialized conception of management consulting. In some cases, the critical approach mixes over- and undersocialized views by portraying clients as somewhat retarded victims of both opportunistic consultants and mesmerizing management fads. This way, it has no concept of situations in which clients know exactly what they are doing when they hire consultants, and of conditions in which social ties and reputation effects preclude opportunistic action by consultants. Much of the literature from the critical camp seems based on an anti-consulting attitude, and scholars reproduce and reinforce their attitude in their research. The neglect, or even denial, of client prudence and economic deliberations is reminiscent of what W. O. Coleman (2002) has recently pointed out as anti-economics. I shall take up this discussion in chapter 4 and in the conclusion.
The only theory that the previous literature on consultancy has systematically drawn on is sociological neoinstitutionalism. For example, many articles in the volumes edited by Sahlin-Andersson and Engwall (2002) and Kipping and Engwall (2002) draw on Meyer and Rowan (1977), DiMaggio and Powell (1983), Powell and DiMaggio (1991), or Tolbert and Zucker (1996). Sociological neoinstitutionalism is based on the argument that it is belief in the efficiency of particular practices or solutions, rather than any proven efficiency, that determines or influences economic action. According to this view, legitimacy toward the organizational environment rather than technical efficiency represents the core of organizing. If the efficiency or efficacy of organizational innovations or management ideas cannot be objectively evaluated, then they are oriented toward what the environment or decision-makers themselves believe to be efficient or effective. This leads to a number of effects – such as the institutionalization of management ideas – that are deemed efficient but are not necessarily so, or to pressure on organizations to adopt the same practices or structures as other firms (isomorphism) in order to gain legitimacy. Issues such as the legitimacy of organizational structures, the enforceability of change processes, and the validation of management decisions have taken center stage in the literature on consultancy (Sahlin-Andersson and Engwall 2002; Kipping and Engwall 2002; Alvesson 1993, 2004). The large and renowned consultancies in particular have duly been described as carriers not only of knowledge but also of legitimacy, as their analyses and reputation validate management decisions.
The diffusion of management concepts and innovations also touches upon elements of isomorphism in the neoinstitutional sense. If the efficiency and effectiveness of change initiatives or innovations often remain uncertain, then organizational decisions are frequently – on a normative or mimetic basis – oriented toward the behavior of other organizations. If a number of firms adopt a particular practice or innovation, then this is taken as signifying that these practices or innovations generate improvements. Even if it remains impossible to determine with certainty whether an innovation triggers progress or more efficient operations, a firm at least puts itself on equal footing with other firms if it adopts the same practices, and for this it often needs agents of change (such as consultants) as transmitters. Observations of McKinsey interventions, for example, have given rise to one of the founding publications of neoinstitutional theory, the article by DiMaggio and Powell (1983), which was based on the two authors’ observation that McKinsey advice led to a number of isomorphic changes in public- and private-sector organizations.
Sociological neoinstitutionalism has been somewhat appropriated by the critical view on consultancy, as the theory seems to fit nicely into the critical camp's doubts about efficient outcomes from consulting assignments. However, the theory does not lend itself fully to the critical view. In fact, it has some elements of functionalism. For example, consultants as traders of legitimacy provide a service to a client even if their solution is similar to others, because it puts the consulted firm on a par with the others. Moreover, the sheer otherness of consultants in relation to client firms plays a central role in their ability to provide advice and gain legitimacy for it (Meyer 1996). And, as a central point, in their article on the institutional conditions for diffusion (of innovations, management practices, etc.), Strang and Meyer (1993) argue that any process of diffusion is accompanied or even preceded by a process of institutionalization. That is, before anything can disseminate as an idea or practice, it must be conceptualized and commodified as a term and concept, for only a communicatively transferable concept or explicit theory stands a chance of diffusing within or between professional groups. Consultants represent interpreters and theorists of individual cases and events. They often frame ambiguous information in new terms and theories, and thus develop and sharpen an interpretive consciousness within the client firm. Only this preceding theorization and term-building process enables an idea to diffuse. And, again, it is especially those consulting firms with a high public reputation that play a part in this process.
The application of sociological neoinstitutionalism to management consultancy has been a useful and important advance, as it has highlighted the role of consultants in legitimation processes and in the communicative framing that precedes the diffusion of management concepts. Nevertheless, relying solely on sociological neoinstitutionalism may narrow the focus on societal norms and divert researchers from looking at the deliberative processes of individuals. Although sociological neoinstitutionalism acknowledges the possibility of different degrees of deliberation in economic action (Meyer and Rowan 1977; DiMaggio and Powell 1983), the question of the conscious behavior of economic actors represents the Achilles heel of this theory. As DiMaggio (1988: 9) observes, “[s]elf interested behavior tend[s] to be smuggled into institutional arguments rather than theorized explicitly.” Sociological neoinstitutionalism has been developed to model the influence of norms on economic action, but it has difficulties with modeling autonomous action in the context of norms that economic actors are aware of. DiMaggio's (1988) distinction between institutionalization as a process and as an actual state then allows us to conceive of individual action at least in processes of institutionalization (see, in this context, Tolbert and Zucker 1996 and Barley and Tolbert 1997). If we take into account the possibility that clients are experienced and knowledgeable executives who can reflect on norms and act deliberately, then sociological neoinstitutionalism meets its limits and other theories suggest themselves.
In particular, economic signaling theory (Spence 1973, 1974, 1976) models deliberate signaling processes in the context of known norms. Signaling theory argues that, in markets of credence goods and quality uncertainty, providers invest in product or service features that signal status, quality, and reliability. Spence models graduate education (essentially, the reputation that different kinds of education involve) as a signal for graduates’ future productivities. At the center of attention are the costs of signaling (e.g. for graduates on the job market the costs of education such as loans and household credit, and the effort put into attaining the degree), the effects of signaling (type of job, salary, promotions of the hired employee), and the incentive structures to invest in signals. If a provider cannot prove the quality of the outcome prior to purchase, and not even for some period after purchase, then he resorts to proving input factors as an indicator for the quality of the outcome. Signals such as certificates concerning educational background reduce the information asymmetry between supply (graduate) and demand (employers) of labor. Spence's central point is that a good education works as an efficient mechanism to signal a graduate's future productivity because, for someone with lower future productivity, it would be much more costly (investments, efforts) to attain a renowned degree. A conceptually simple but methodologically unfeasible test of signaling theory would be, for example, to gather people of identical ability and randomly assign some of them a degree certificate. If those with the certificate later earn more, then signaling theory would be supported.
There is one fundamental difference between economic signaling theory and sociological neoinstitutionalism. The former assumes that the signaling mechanism works as an efficient device to connect supply and demand, the latter looks at deviations from economic efficiency that legitimacy-seeking behavior brings about. In other words, signaling theory assumes that the market clears efficiently and conceptualizes how this comes about by signaling mechanisms. This assumption of efficiency may appear absurd to sociological neoinstitutionalists, because they observe economic action oriented toward norms and anticipated expectations independent of or detrimental to economic efficiency. Indeed, the explanation of economic actions in cases where efficiency remains unclear is the main purpose of the theory.
Nevertheless, the two theories have two important aspects in common. First, both view the essence of economic behavior in aspects external to the immediate exchange relationship, such as the status of education at prestigious colleges/universities or the status of particular concepts of organizational structure. In other words, both focus on the orientation of economic behavior toward the norms within which exchange partners act, rather than toward the immediate features of the exchange partners. The second commonality is that both theories imply a decoupling of reputation from the actual quality of a service. For sociological neoinstitutionalists, the legitimacy effect is decoupled from the economic quality of a decision (e.g. regarding organizational structure). Alternatively, an economically positive effect arises as a result of the gained legitimacy rather than from any intrinsic economic quality of the decision. Firms make particular decisions not because they have proven economic effects but because the environment considers them useful. Signaling theory, too, relies on the assumed rather than the actual quality of education. That is, a graduate from a college of high reputation may have undergone a worse preparation for a job than someone from an unknown college. Nevertheless, the graduate from the high-reputation college is rightly assumed to have a higher future productivity. This is because those individuals with a high future productivity independent of the education have less costly access to colleges of high reputation. Thus the signaling mechanism works irrespective of the actual quality of the education.
Important for our purposes is the notion that whether the behavior of market participants leads to efficient or inefficient outcomes cannot be assumed a priori, hence there can be no prior nonnormative preference for either of the two theories. Rather, the essence is to compare the theories with regard to individual phenomena. Management consultancy can be perfectly modeled in terms of signaling theory. It represents a market of experience if not credence goods, and the quality of a consulting service is very difficult or impossible to prove in advance. As a result, management consulting firms signal output quality by input quality – i.e. by the quality of their human resources (see chapter 9 for details, and Pudack 2004). Although they accept applications from all universities, they hire actively only from the top business schools and universities. Independent of whether these institutions really deliver higher educational quality, these are the places that highly talented people covet and have less costly access to than less talented people, and therefore these are the places where the best graduates can be hired.
In fact, the large and renowned management consulting firms play a crucial role in what could be theorized as a signaling economy. The most talented students are drawn to the most renowned universities, irrespective of whether the educational quality is proven to be better there. Renowned management consulting firms hire from the most renowned universities and actually obtain better graduates than from other universities, again irrespective of proven educational quality. By hiring from these universities, the renowned consulting firms signal high output quality, can charge higher fees to their clients, and thus can offer higher salaries to their graduates. For consulting firms of lower reputation which cannot charge such high fees, it would be more costly to hire the same graduates, as they cannot carry over the higher personnel costs to clients in the same way. The renowned business schools, in turn, can signal quality by referring to the coveted jobs their graduates obtain, which, again, renders it more costly for less talented people to secure placement there.
This signaling circle can be extended to consulting clients. Large corporations hire the most renowned consultancies because the latter signal better consulting quality through their top business school graduates. Signaling high-quality advice means gaining legitimacy for management decisions and thus signaling management quality, which leads to advantages in the capital market. If the capital market rewards better talented students with less expensive loans and thus lower costs of obtaining access to coveted universities, then the signaling circle would be closed. But this might drive the signaling argument to the extreme and be too far-fetched to apply to future consultants. This point is taken up in the conclusion (chapter 10).
The orientation toward norms external to the immediate business relationship is an important factor influencing the consulting market, but only one factor. If we take into account the possibility that clients can make rational decisions about whether to hire consultants or not, then we need to look more closely at the deliberations that characterize the immediate exchange relationship. Transaction cost economics (Coase 1937; Williamson 1975, 1985, 1986, 1988) helps theorizing when or for which business problems internal solutions or market provisions are beneficial. From this viewpoint, cost considerations are the crux of economic action, but more closely related to the immediate features of the transaction than signaling theory suggests. To transaction cost theorists, rationality (bounded by the available information and processing capabilities), calculativeness, and opportunistic behavior in business relationships represent an imperfect yet still the best possible set of assumptions for modeling economic behavior. The point of departure is the assumption that a company's costs can be classified in two categories: production costs and transaction costs. Production costs are those directly attributable to the productive capacities, such as manufacturing or logistics. Transaction costs, by contrast, are those associated with organizing economic activity. The latter comprise costs that occur prior to or that lead to a transaction, such as costs for gathering information, for negotiation, and for finalizing a contract, and costs that emerge after a transaction has been agreed upon, such as costs for interpreting contract clauses, enforcing contractual conditions, monitoring, conflict solving, or adjusting the contract. The decision of whether a task or service is to be conducted in-house (“hierarchy solution”) or purchased in the market (“market solution”) is based on a comparison of the sum of production and transaction costs.
Due to the detailed consideration of make-or-buy decisions, transaction cost economics is useful to outline clients’ decision if and when to hire external consultants. Williamson conceptualizes make-or-buy decisions on the basis of three factors: the uncertainty, frequency, and asset specificity of a transaction. The argument is that, the higher the uncertainty, frequency, and asset specificity of a transaction, the more efficient an in-house solution is, because the transaction costs of elaborating and enforcing a reliable contract with an external provider would be higher than monitoring internal personnel. Vice versa, transactions with a low degree of uncertainty, frequency, and asset specificity can more sensibly be outsourced.
The question of which management functions are better conducted in-house and which should be outsourced to a consulting firm can be approached with the transaction cost tool by a trivial example. Suppose a chemical engineering company faces continuous engineering challenges in order to maintain and improve its products and production processes. These core engineering activities demand a highly specialized workforce and engineering equipment (high asset specificity), the challenges occur on a regular basis in the context of process maintenance (high frequency), and uncertainty is relatively high because the challenges are often accompanied by research and development (R&D) issues. As a result, outsourcing those activities would be inefficient, and the company will retain an internal workforce of chemical engineers to take care of them.
The same chemical engineering firm may face a new challenge, for example an opportunity to acquire or set up a plant in a new region with a much lower cost structure. Such a situation requires an analysis that represents a one-off activity (low frequency), it does not involve specialized machinery (low asset specificity), and the process of analyzing the situation is less burdened by long-term uncertainty than research and development processes. As a result, the chemical engineering firm may reasonably outsource this analytical service to an external provider, for example a consulting firm. This may be a trivial example of applying transaction cost economics to the consulting business, but it conveys the consideration that explicitly or implicitly underlies a decision to hire consultants.
Information economics (Stigler 1961; Alchian and Demsetz 1972) belongs to the same family of theories as transaction cost economics. It compares the usefulness of information with the costs of obtaining it. With regard to a comparison between an internal and an external solution to gaining the necessary information, its existence is attributable to the advantages of monitoring joint inputs. That is, in comparison to a market solution, it is easier for a firm to monitor internally those inputs that are not attributable to individual providers, such as employees. Put differently, “The ability to detect shirking among owners of jointly used inputs in team production is enhanced (detection costs are reduced) by this arrangement and the discipline (by revision of contracts) of input owners is made more economic” (Alchian and Demsetz 1972: 794). The limit to a firm's size is reached when the “specialized knowledge about inputs becomes as expensive to transmit across divisions of the firms as it does across markets to other firms” (794). To translate the argument to the consulting market, in simplified terms the limit to a client firm size is reached when the costs of transferring information within the firm, for example from the bottom of the hierarchy to the chief executive officer (CEO) or between divisions, are higher than transferring this information through an external provider. In a similar vein, the limits to a firm's size are reached when labor law renders it difficult to dismiss employees. Hiring consultants may often mean purchasing short-term enhancement of analytical capacities that are in principle available for in-house employment but that would constitute overcapacity after the task has been finished.
With regard to management consulting, one of the points is that clients economize on the acquisition of knowledge and information when hiring an external consultant. They do not necessarily buy the direct supply of information but, rather, information-gathering or knowledge acquisition skills. If internal employees had (inexpensive) methodological means to collect and distribute them internally, or if collecting this information represented an interplay of joint inputs rather than a service attributable to a particular party, then the use of external consultants would be less economical. In short, methodological skills for tasks which occur rarely or aperiodically in any one firm, which involve high costs of internal coordination and distribution, which can be attributed to a particular provider, and which can be more economically acquired and applied across firms are the crux of the consulting business. It is indeed surprising that few scholars of management consultancy have drawn on transaction or information cost economics (Canbäck 1998a, 1998b, 1999 and Kehrer and Schade 1995 represent notable exceptions), although these arguments explicitly address the alternatives from the client perspective. Chapter 2 will provide more detailed analyses and examples.
In conceptualizing mechanisms internal to the immediate exchange relationship, transaction cost economics has been challenged over the past twenty years by embeddedness theorists (Granovetter 1985; Powell 1990; Granovetter and Swedberg 1992; Uzzi 1996, 1997; Dacin et al. 1999). To them, transaction cost considerations represent an “undersocialized” (Granovetter 1985) conception of economic action. Embeddedness theory argues that organizational economists are not necessarily wrong in their cost comparisons and assumption of calculativeness, but that they ignore or underestimate the point that most economic action takes place in established lanes of social ties and networks. Calculativeness is not absent, but it is often bounded by social ties, and the efficiency of a transaction is only one consideration or possibility.
Embeddedness theorists distance themselves from transaction cost economics by arguing that management decisions, for example between subcontractors or make-or-buy alternatives, are primarily based on the structure and quality of the relations between decision-making executives of different companies, which can only imperfectly be captured in terms of transaction costs. At the heart of the embeddedness paradigm is the structural aspect of social relations, especially the significance of personal and business networks for economic transactions. Business relationships, for example between consultants and their clients, are rarely characterized by arm's-length relations and opportunistic behavior by the two parties, but typically by long-term relationships of trust and/or a social embeddedness in networks of business partners. As a result, transactions may be inefficient without the participants either noticing or calculating it as such. A transaction cost analysis of such processes may then represent an ex post rationalization of an otherwise inefficient solution.
From this perspective, the question emerges as to how consulting assignments come about and under which circumstances consulting firms compete with each other. For example, from the embeddedness perspective the make-or-buy decision is not based primarily on a calculative comparison of costs between hierarchy and market solutions, but the very question of whether to outsource or not emerges only from social ties. For example, a client may learn about a particular kind of consulting service only through business contacts and social ties to clients, suppliers, or competitors. Calculative cost considerations may then be a complementary feature of the make-or-buy decision, but the primary mechanism is a social relation one. More often than not, the decision-making process does not follow the sequence of, first, deciding about making or buying, and, second (in the case of buying), checking who would be a good provider. Rather, the social interaction with business partners is often the trigger for a buy decision, and possibilities of in-house solutions or other external providers are not considered.
Indeed, empirical findings suggest that competition between providers in the consulting sector is not based on price or costs (Dawes et al. 1992; File et al. 1994; Clark 1995; Page 1998). The degree and significance of the uncertainty that clients face when choosing and interacting with a consulting firm is high. Objective quality measurement of the mostly immaterial consulting services is difficult to achieve, and management consulting (as well as most other knowledge-intensive business services) is performed subsequent to the contract, which shifts the risk as to quality or partner adequacy toward the client. In such situations, informal social institutions such as trust, reputation, and word-of-mouth effects take center stage. These may also save costs, such as for gaining information or screening quality, but they may also prevent price or cost considerations on the buyer's side. As a result, the quality of a network tie to a client decision-maker is the main competitive advantage of a consulting firm. Chapter 3 outlines these circumstances in greater detail.
Embeddedness theory offers an additional perspective: applying the focus on social ties and networks to firm-internal matters. In the literature on knowledge management, the “communities of practice” approach (Brown and Duguid 1996, 1998) has obtained broad attention. While this approach tries to develop its own theory from the knowledge-based theory of the firm (Spender 1996; Grant 1996a, 1996b), it places the effects of strong and weak ties on knowledge creation within the firm at the center of attention. For example, Brown and Duguid (1998: 97) write,
[M]ost formal organizations are not single communities of practice, but, rather, hybrid groups of overlapping and interdependent communities. Such hybrid collectives represent another level in the complex process of knowledge creation. Intercommunal relationships allow the organization to develop collective, coherent, synergistic organizational knowledge out of the potentially separate, independent contributions of the individual communities.
While Brown and Duguid, and others who pursue the communities of practice approach, rarely refer to embeddedness theory as a point of orientation and source of information, the above quote shows that their arguments are practically the same. Not only strong ties (within a community of practice) but also weak ties (between communities of practice) lie at the center of their attention. Hence, their approach can duly be seen as an application of embeddedness notions to firm-internal matters – with one difference: authors such as Brown and Duguid praise communities of practice as being functional for firms but do not look at the inefficiencies that embeddedness effects may involve. In fact, one of the key insights that embeddedness research has brought about is that economic exchange does not always follow the lines of efficiency, precisely because it is bound by social ties. In a similar vein to the communities of practice approach, Högl and Gemünden (2001) and Högl et al. (2004) show that not just teamwork quality but inter-team coordination as well are crucial factors for the success of innovative projects. Thus, it is not only the quality of strong ties but also the management of weak ties within a firm that matter for innovativeness and corporate performance. But, unlike embeddedness research, Högl and Gemünden (2001) and Högl et al. (2004) do not look at the limits to efficiency that network ties often involve. This discussion will be taken up in chapter 8 on organizational design, governance, and knowledge management in consulting firms.
In summary, all four theories have important things to say about management consulting, although they are based on different assumptions and look at different issues. Signaling theory and transaction cost economics are both driven by the assumption that observations of calculativeness and cost considerations teach us most about economic behavior, while sociological neoinstitutionalism and embeddedness theory argue that observations of social mechanisms and the limits to calculativeness are more informative. Signaling theory and sociological neoinstitutionalism, by contrast, have in common their orientation toward norms and thus on mechanisms external to the immediate transaction partners, while transaction cost theory and embeddedness theory focus on the immediate features of the transaction and its participants. Table 1.1 summarizes these juxtapositions.
Table 1.1. Four theories and their focus
| Focus on cost considerations (“economics”) | Focus on social mechanisms (“sociology”) | |
| Mechanisms external to the immediate exchange relationship | Signaling theory | Sociological neoinstitutionalism |
| Mechanisms internal to the immediate exchange relationship | Transaction cost economics | Embeddedness theory |
At this point, the differences and applications of the three theories need to be outlined more thoroughly. The scientific community is often split in its assumptions concerning the nature of business relations and human behavior, and this split is reflected in the use of theories. Typically, the divide lies between the disciplines of economics and sociology (Swedberg 1990; Zukin and DiMaggio 1990; Friedland and Robertson 1990; Lie 1997). The debate between transaction cost economics and embeddedness theory represents this divide and, accordingly, is outlined first. The difference between transaction cost economics and sociological neoinstitutionalism follows different lines but is equally representative of the two disciplines. However, there are differences even between theories of the same academic discipline. As mentioned above, neither embeddedness theory and sociological neoinstitutionalism nor transaction cost economics and signaling theory form coherent approaches, as they differ in their assumption as to whether mechanisms outside or inside the immediate transaction relationship are more important. These theories cannot be applied to management consulting without discussing their differences or without discussing the debate on paradigm incommensurability, which will be carried out on the basis of critical rationalism.
Transaction cost economists tend to conceive of economic action in terms of the calculativeness (bounded rationality) and opportunism of market participants. This does not mean that market participants have to have egoistic motives; Homo economicus may have egoistic or perfectly altruistic motives. What is important for the economic paradigm is only that he pursues these (possibly altruistic) motives in a calculative and opportunistic manner. Through this lens, modeling business decisions in terms of cost considerations is a logical consequence, for costs represent a useful unit for invested effort.
Embeddedness theorists, by contrast, do not consider this approach particularly useful. As outlined above, they hold that the degree of calculativeness in economic transactions is often limited, because actors are embedded in social relations. These may be direct trust relations between suppliers and clients, or webs of social relations in which weak ties enable transactions between participants who are not connected through direct ties. From the embeddedness perspective, calculativeness and opportunism, in the sense of cheating when the cost of reputation damage is lower than the gains of cheating, be it modeled by transaction cost or game theorists (von Neumann and Morgenstern 1944; Axelrod 1984; Fudenberg and Tirole 1991; Kreps 1991), represent an “undersocialized” (Granovetter 1985) image of economic behavior. Bound by social ties, economic transactions may be inefficient, and transaction cost economics may at best be able to rationalize individual behavior ex post rather than sketch the reality of economic transactions at the time of decision-making. The arm's-length relationships between economic actors that economists often assume by default might be an accurate assumption for a consumer transaction such as buying a shirt or a pair of shoes, but not for purchasing business services in markets of credence goods and quality uncertainty (DiMaggio and Louch 1998).
Such theories are often considered either or paradigms for scholars. Either one models business relations as arm's-length ones with opportunistic behavior (essentially, as relationships in which cooperation is based on calculation, as game theory does), or as embedded relations with limited calculativeness. However, this either or relationship must be put in perspective. Embeddedness theorists emphasize that buyers often choose transaction partners within preexisting noncommercial ties. If these ties do not entail an appropriate provider, they seek recommendations from noncommercial and trust-based commercial ties to identify and assess transaction partners with whom they did not previously have relations. This is a process of actively pursuing word-of-mouth effects, which DiMaggio and Louch (1998) call “search embeddedness.” It does not preclude economizing behavior but emphasizes its limits. As Powell (1990: 323) argues in his article on network forms of organization,
Economizing is obviously a relevant concern in many instances, especially in infant industries where competitive preserves are strong. But it alone is not a particularly robust story, it is but one among a number of theoretically possible motives for action – all of which are consonant with a broad view of self-interest. Clearly many of the arrangements discussed above [network forms of organization] actually increase transaction costs, but in return they provide concrete benefits or intangible assets that are far more valuable.
This argument is supported by the analyses of Burt (1992, 2004), who finds that good ideas are connected with brokerage positions between network holes, which feed back on individuals in terms of performance evaluation, compensation, and promotions.
Since the embeddedness approach has become prominent, economists have come to be aware that networks play a critical role in economic transactions. Williamson (1991: 291), for example, outlines a network as “a nonhierarchical contracting relation in which reputation effects are quickly and accurately communicated.” He has become aware of the role of social ties in economic transactions, but insists that acting within strong and weak social ties still entails a lot of economizing:
It's my feeling that a rather huge fraction of what is going on in these network enterprises can be interpreted usefully in transaction cost terms. Actually, one of the things that is probably frustrating to noneconomists is that economics is so incredibly elastic. Once the economic content of a concept is understood, economics finds a way to embrace it. So I anticipate that networks can probably be incorporated – at least to some degree – into an extended version of transaction cost economic. (interview with Williamson in Swedberg 1990: 122)
And, indeed, game theory (von Neumann and Morgenstern 1944; Fudenberg and Tirole 1991) is based on exactly this notion: that an offeror gives trust as a specific investment if he reckons with rents, if reciprocated, or with comparatively lower costs, if not. The other part, the decision-maker, reciprocates trust if it involves advantages (Axelrod 1984; Raub and Weesie 1990; Kreps 1991).
From this perspective, nothing stands in the way of modeling trust in cost terms. Network ties simply enable actors to save the costs of searching and assessing product or service quality, and pursuing transactions in webs of strong or weak ties saves the monitoring and contract enforcement costs. Even reputation then emerges as a result of iterated games of calculative refraining from monitoring (Raub and Weesie 1990). Hence, economists are perfectly able to model cooperating and gaining a reputation as a fair player, which could then be called “trust,” as a result of calculative behavior (Ripperger 1998; Axelrod 1984). Williamson (1993) reserves the term “trust” for purely noncalculative relationships and argues that calculative trust is a contradiction in terms. On this basis, he argues that economic relationships simply do not entail trust, but only calculative cooperation. By contrast, Ripperger (1998) argues, in line with game theory, that trust and calculativeness cannot sensibly be divided. Reserving trust for purely noneconomic relationships assumes a schizophrenic concept of human beings, whose deliberations and behavior precisely distinguish between different spheres of life. Ripperger (1998: 247–8) further argues that calculativeness forms a basis for trust rather than a contradiction of it, because moral behavior such as trustfulness cannot be learned without reason, deliberations, and thus calculation.
As a result, trust in economic relationships can be conceptualized as the degree to which an actor refrains from screening, monitoring, or demanding explicit contractual security or incentives. In games of iterated prisoners’ dilemmas, cooperating, in terms of refraining from short-term opportunism, may pay off in the long run, and an economics of trust emerges as the calculation of when this makes sense (Fudenberg and Tirole 1991; Kreps 1991; Ripperger 1998). Such deliberations can be perfectly virtuous if the ends and motives are altruistic. Vice versa, monitoring may lead to a spiral of distrust and thus to an inefficient economic transaction (Ripperger 1998: 70). The assumption of opportunistic behavior is no contradiction of this notion: only the possibility and danger of opportunistic behavior establishes those risks without which trust would not exist as a social institution (Ripperger 1998: 60). Tacit collusion, defined as two organizations using cues but without direct communication in order to achieve mutual gain, may then result in the formation of implicit cartels (for an overview, see Ivaldi et al. 2003).
A client–consultant relationship can be taken as an example. Transaction cost economists can model clients' cost-based decisions between alternative providers. The relevance of trust relations between consultants and clients can, in principle, be acknowledged and integrated in terms of search, information, and anticipated monitoring costs. From the embeddedness viewpoint, by contrast, such cost considerations either fail to be precise enough to guide a client's decision, or they do not even emerge between trusted business partners. According to this view, it is not that transaction cost considerations are absent but that they are often overruled by qualitative decisions based on social tie quality.
What emerges here is a central problem of monotheoretical perspectives. Adopting only one perspective may result in losing sight of phenomena that the other perspective would discover. For example, for the duration of one or two projects, a client may be perfectly satisfied with a provider's services, and a trust relation emerges between the executive and the senior consultant. This trust may go so far as to prevent the manager from seriously considering alternative providers (so-called “overembeddedness”; Uzzi 1997: 59). Even if a competing consultant enters the scene and offers a potentially better or more economical service, the manager cannot or does not want to examine the new provider's quality and thus may not seriously consider him. In particular in such a market for credence goods, a client may not engage in serious cost-benefit analyses, and he may not even think of disappointing his present partner. Thus, cost considerations interwoven with social tie considerations cannot be expressed purely in cost terms; real events rarely follow the clear lines of a single theory. This discussion will be taken up in chapter 10.
The distinction between transaction cost economics and sociological neoinstitutionalism follows a different line: the emergence and nature of social institutions. For a transaction cost economist, social institutions such as contracts and laws, and even trust and reputation, are arrangements that emerge in the context of economizing on negotiations, monitoring, and enforcement. As DiMaggio and Powell (1991: 3–4) put it in their demarcation of sociological neoinstitutionalism from other approaches, to economists “[i]nstitutions arise and persist when they confer benefits greater than the transaction costs … incurred in creating and sustaining them.” Information is costly, people behave opportunistically, and rationality is bounded; thus the features of transactions such as asset specificity, uncertainty, and frequency give rise to economic institutions. DiMaggio and Powell (8–11) usefully outline the difference between new institutionalism in economics and new institutionalism in organizational sociology along three lines: (a) the degree of consciousness and deliberation in creating institutions; (b) whether institutions are considered results of individual actors’ preferences and interests, or outcomes of collective social constructs; and (c) the degree to which institutions are malleable according to the imperatives of efficiency.
Transaction cost economics conceives of institutions as results of calculative action in order to save search, information, monitoring, and enforcement costs in economic transactions. They may be collective in the sense that groups of individuals have collectively designed them, but economists do not conceive of them as outcomes of collective social constructions that cannot be reduced to individual or groups’ deliberations (DiMaggio and Powell 1991: 9–10). Moreover, economists assume that institutions are also subject to change according to the deliberations of individuals or groups. Sociological neoinstitutionalists, in contrast, take the opposite stance on these matters. They conceive of institutions as social constructs without preceding processes of calculativeness; they argue that institutions cannot be reduced to the interests or preferences of individuals or groups; and they hold that changing or adapting them is highly complex, if not impossible.
Again, these differences can be applied to the consulting business, for example to the choice of organizational structures or outsourcing decisions that consultants may recommend. To an organizational economist, institutionalizations such as the multidivisional organizational structure (M-form) in the decades after the Second World War (in North America it began before the war) or the establishment of in-house consultancies in the 1980s and 1990s (see chapter 5) represent deliberate solutions in order to save transaction costs. In the case of the M-form, the divisionalization of the organizational structure was considered the best means of coordinating multinational corporations to respond to the increasing internationalization. It was assumed to reduce the information overload on top management and to reduce communication problems via standardization. Moreover, the divisions of an M-form have no excuse for poor performance and thus use resources more efficiently, which accounts for the cost advantages triggered by motivation (Chandler 1962; Williamson 1975).
To a sociological neoinstitutionalist, this may be true, but the amazing dissemination of the M-form cannot be fully explained by proven efficiency. The M-form may or may not represent an efficient solution for some corporations; much more important has been the collective belief in its efficiency, which has accounted for its widespread, international adoption. To an individual firm or a decision-making executive, it was probably impossible to ascertain the economic benefits or drawbacks of adopting the M-form. Nevertheless, the institutionalization of this solution as a good organizational structure constituted a strong normative pressure to adopt it. Whether an executive actually believed in the efficiency of a solution, whether he was forced by legal standards, or whether he just reproduced it because comparable firms did the same, the crucial criterion of implementation was the firm's increasing legitimacy in the environment if it adopted the M-form. Moreover, as Bartlett and Ghoshal (1993) and Hoskisson et al. (1993) have pointed out, the term “M-form” covers a large variety of solutions. It is very questionable whether these different forms subsumed under the term represent more efficient responses to corporate internationalization than others. From a sociological neoinstitutional perspective, the diffusion of the M-form may have been a classical case of an institutionalized rationality myth.
A similar argument applies to the growth of internal management consulting in the 1980s (North America) and 1990s (Europe). A transaction cost consideration suggests that, if certain features apply to an analytical task, then in-house consulting may be a more economical solution than external consultancy (see chapters 2 and 5 for details). Again, from a neoinstitutional perspective this may be the case, but a transaction cost view does not suffice to explain the sudden wave of internal consultancies being founded in the 1980s. From a sociological neoinstitutional perspective, these processes of organizational isomorphism were a matter of normative or mimetic adaptation to perceived expectations – that is, to institutionalizations of what was considered efficient rather than as economic responses to what actually was efficient. Nevertheless, to economist and historian North (1990), for example, even broadly shared ideologies represent means to curb opportunistic behavior and to substitute formal rules. As an empirical proof for either theory cannot finally be achieved, it is again the case that more than one theory may inform the explanation of the phenomenon.
Signaling theory and transaction cost economics have several commonalities. Not only do they both take root in the economic paradigm and view cost considerations as the crux of economic action, but they also consider institutions such as contracts or trust as efficient outcomes of deliberate and calculative behavior. Moreover, both are concerned with mechanisms to reduce adverse-selection problems and information asymmetry between transaction partners. Nevertheless, the two theories have dividing lines. First, signaling theory looks at cost alternatives (different kinds of education) that emerge in anticipation of future productivity ascribed by a third party, such as an employer for graduates in the labor market. In transaction cost economics, at least in the basic model there is no such third party, and the alternatives are oneself (make) and a second party (buy). Thus, signaling theory is oriented toward mechanisms external to the immediate alternatives, while transaction cost economics is oriented toward the attributes of the immediate transaction alternatives. Like sociological neoinstitutionalism, therefore, signaling theory involves a macro component of societal expectations or norms, while transaction cost economics operates at the meso level of immediate transaction alternatives.
The second difference is that signaling theory models the behavior of the better-informed party to reduce information asymmetry, while transaction cost economics models the behavior of the worse-informed party. Signaling costs emerge for those who know about their future productivity and have to convey it to the other. Transaction costs of screening, selecting, monitoring, and enforcing emerge for those who have to learn about their transaction partner rather than convey information about themselves. The question of which party takes the initiative to reduce information asymmetry gives rise to the question of market power, which will be addressed in chapter 4.
Sociological neoinstitutionalism conceives of institutions as social and cultural constructs in the sense of unreflective routine behavior and taken-for-granted norms and views. From this perspective, signaling theory seems just the opposite – intentional participation in signaling games based on a full awareness of norms such as good education. Yet sociological neoinstitutionalism and economic signaling theory address two sides of the same coin. The former tries to explain the emergence and effects of norms and institutions, the latter models the way how individuals act in the context of these taken-for-granted norms. Sociological neoinstitutionalists, for example, observe the decoupling of an official organizational structure from the actual functioning of an organization (Meyer and Rowan 1977). Signaling theory, by contrast, observes how individuals such as job-seeking graduates behave in a market that is characterized by taken-for-granted institutions and thus provides a cost consideration of decoupling effects.
Nevertheless, important differences remain. The most important one has been mentioned above: sociological neoinstitutionalism does not suppose that economic decisions, for example selection principles in job markets, are particularly rational or efficient (as signaling theory holds). Rather, sociological neoinstitutionalism is interested in systematic distortions from economic efficiency, for example in hiring graduates from renowned universities although they may be less productive. These doubts about economic efficiency are particularly relevant in markets of credence goods. Consulting services comprise a variety of procedures on a continuum between experience and credence goods. To sociological neoinstitutionalists, it is unclear how approaches such as signaling theory can sensibly model such markets on efficiency assumptions. Hence, sociological neoinstitutionalism looks at symbolical and cultural processes within which institutions emerge and inefficiencies persist, while signaling theory is concerned with deliberate investments in sending imperfect signals, which result in efficient matching processes between supply and demand.
Because of its focus on the influence of institutions, sociological neoinstitutionalism may lead researchers to underestimate the intentionality with which market participants act. Calculative behavior by individuals who are aware of social institutions is a possibility, but the thrust of sociological neoinstitutionalism is the unconscious effect of institutions on individual behavior. Self-interested behavior tends to be “smuggled into institutional theory” (DiMaggio 1988: 9) after the original setting has abstained from it. Indeed, if empirical research increasingly identifies actors’ behavior as calculative and aware of institutions, sociological neoinstitutionalism gradually loses its explanatory power. By contrast, signaling theory may lead researchers to erroneously assume efficient outcomes of economic action. Job market signaling, for example, is considered an effective way that leads the market to clear (Spence 1973, 1974). The theory is interested in explanations of how and why a market clears, rather than in market failure. For credence goods, such assumptions are shaky. If empirical research finds that a job market does not clear, or that individuals with lower productivity do not have higher signaling costs than high potentials, then signaling theory meets its limits. If the degree of intentionality of market participants varies, or if market clearance is but one out of several observed results, then the juxtaposition of signaling theory and sociological neoinstitutionalism may lead to learning effects.
Signaling theory and embeddedness theory may appear intellectually too far away from each other to be sensibly compared. Interestingly, however, both are rooted in observations of the same topic: job markets. Granovetter's (1974) study of job market participants introduces the now common distinction between strong ties as direct trust relations and weak ties as rooted in rare encounters, mutual acquaintances, or mechanisms of recommendation. It gives center stage to the notion that weak ties can make the difference in accomplishing a transaction such as getting a desired job. From the embeddedness perspective, the essence of successful job market behavior is the mobilization of weak ties, rather than applications without previous contacts or reliance on strong ties. This represents a big difference from Spence's (1974) study of job market signaling, where both strong and weak ties are largely left out of the picture. Job market signaling is based on the assumption of arm's-length relationships, or even complete anonymity, between seekers and providers of jobs. A commonality is that both theories portray the job seeker as the one who takes the initiative in reducing the other party's informational disadvantage. In real life, the two ways of reducing information asymmetry may complement each other. For example, the mobilization of weak ties may be strongly facilitated when diplomas from renowned educational institutes can be presented. And women or ethnic minorities may not succeed in the job market as much as white males do, even though they have the same contacts, because their signaling costs are higher, which may render it more difficult for them to build on their contacts. Alternatively, they may not even obtain these contacts because of their higher costs of signaling future productivity. In fact, if job market participants obtain contacts by way of attending a renowned educational institute, then the two effects reinforce each other and are only analytically distinct.
Nevertheless, these analytical differences are important. First, embeddedness-based analyses of job markets tend to play down the cost aspect of contacts. Obtaining or maintaining strong and weak ties does not come for free but requires investments. And signaling theory, rather than embeddedness theory, looks at these investments in reducing information asymmetry. Second, embeddedness theory looks at the immediate features of the relationship through which a transaction comes about. In the case of job markets, this is the quality of the relationship to a potential employer. Signaling theory, by contrast, neglects these immediate features and focuses on mechanisms external to the immediate exchange relationship; in the case of job markets, on the reputation and therefore the value of education and on the investments in obtaining it. Like the difference between transaction cost economics and signaling theory, the levels of observation are meso versus macro. I will return to this discussion in the conclusion in Part III.
Both embeddedness theory and sociological neoinstitutionalism represent critiques of transaction cost economics, but important differences between them remain. Embeddedness theorists see the essence of economic transactions in the social relations between individual actors and in webs of mutual obligations. They focus on the quality of social relations in terms of strong or weak ties. Sociological neoinstitutionalists, in contrast, consider economic action to be driven by systems of rules or assumptions about efficiency that are shared not only by members of a business network but by larger entities, such as entire business sectors, organizational fields, nation states and their cultures, or even “world society.”
In the field of management consulting, the diffusion of innovations and the role of consultants in this process constitute good cases for outlining the difference between the two theories. As mentioned above, embeddedness theorists try to reconstruct or explain the diffusion of innovations by looking at the social ties within which innovative products are brought about or disseminated. Extensive studies of social networks account for detailed analyses of the relationship between network position and innovation success (Burt 1992, 2004; Powell 1996, 1998; Powell et al. 1996; Tsai 2001; Beckman and Haunschild 2002; Reagans and McEvily 2003; Ritter and Gemünden 2003). Management consultants are considered in the context of technology transfer and bridges of innovation between formerly unconnected actors or firms. The analysis of innovations and their diffusion is thus located at a meso level of social relations and their effects.
Apart from this reinterpretation of consultants as carriers of economic functions, neoinstitutionalism views them in yet another role. Those consulting firms with a high public reputation may bestow legitimacy upon an innovation or business idea and thus validate a management concept. Consultants contribute to the institutionalization of an innovation not only through interpretation and term-building processes but also by being an icon of management knowledge. The top consulting firms in particular represent rationality and quantitative-analytical competence. Embeddedness theorists may reply that being an icon of management knowledge at a macro level of institutionalized myths may be useful but barely suffices to get a consulting contract. Only social tie quality at a meso level enables this, and this is true for large as much as for small providers. Thus, embeddedness operates at a meso level while sociological neoinstitutionalism represents a macro level of sociological inquiry.
| Image not available in HTML version |
In summary, the differences between the four theories can be sketched in figure 1.1.
Looking at these differences between theories, the immediate concern of incommensurability may emerge. In organization theory, the assumption that different paradigms are incommensurable was first proposed by Burrell and Morgan (1979). They argue that the four paradigms of organizational analysis that they identify (radical humanist; radical structuralist; interpretive; functionalist) are based on entirely different assumptions about human nature and about the nature of social science (ontology, epistemology, methodology). Hence they say that researchers should adopt one perspective and contribute to research and knowledge production within it. Since all four theories outlined above represent different approaches to economic phenomena and are based on different assumptions about the nature of business relations and human behavior, how can they all be applied to management consulting in one book? Signaling theory and transaction cost economics assume calculative behavior and model it in cost terms. Sociological neoinstitutionalism and embeddedness theory look at the limits of calculative behavior and point to social phenomena that guide or at least influence the behavior of market participants. The gulf between, for example, embeddedness theory and transaction cost economics appears to many scholars to be too large to be bridged. The argument is that one can only either assume that human behavior is fully calculative or that it is bound by social ties, or one can only either assume that individuals act opportunistically or that they do not. The potential for integration seems low.
However, the above debate between embeddedness theory and transaction cost economics has indicated that actual business relations – for example between clients and consultants – are characterized by a mixture of personal trust, calculative cost considerations, reliance on weak ties, and arm's-length search behavior. These elements of the business relationship and their individual weight vary from case to case, or even overlap within a single decision-making process. Ignoring either embeddedness theory or organizational economics may be appropriate in one situation but misleading in another. Screening out one category by defining strict assumptions of human behavior may lead to a clean model but distract from a comprehensive image of what is really going on.
Likewise, transaction cost economics and sociological neoinstitutionalism are analytically different too. Social institutions emerge through calculative behavior (economics) or through norm-based conduct (sociological neoinstitutionalism). Nonetheless, scholars such as Ruef (2002) point out that some elements are compatible. The process of interpreting and framing ambiguous information in new terms – which, for example, consultants perform – is a central element of social institutionalization processes. At the same time, it represents a service to clients who would incur higher transaction costs if it was carried out internally. As mentioned above, before anything can disseminate as an idea or practice, it must be conceptualized and commodified as a communicatively transferable concept. As interpreters and theorists of individual cases and events, consultants form a part of institutionalization processes, and at the same time save clients transaction costs by sharpening an interpretive consciousness at lower costs than in-house personnel could.
The person who rejected the thesis of paradigm incommensurability early and most forcefully was Karl Popper. In his book The Myth of the Framework (1994; based on articles he wrote in the 1960s and 1970s) he explicitly deals with the argument that paradigm incommensurability – in his words “the doctrine that truth is relative to our intellectual background and changes from one framework to another” – builds an appropriate basis for scientific progress. In fact, he considers this doctrine an intellectual fashion, and points out that its supporters are either ideologists who seek to render their theory immune to critique, or relativists who assume that truth is relative to the applied framework.
This doctrine [paradigm incommensurability] is, logically, an outcome of the mistaken view that all rational discussion must start from some principle or, as they are often called, axioms, which in their turn must be accepted dogmatically if we wish to avoid infinite regress. […] Usually those who have seen this situation either insist dogmatically upon the truth of a framework of principles or axioms, or they become relativists: they say that there are different frameworks and that there is no rational discussion between them, and thus no rational choice.
But all this is mistaken. For behind it is the tacit assumption that a rational discussion must have the character of a justification, or of a proof, or of a demonstration, or of a logical derivation from admitted premises. But the kind of discussion which is going on in the natural sciences might have taught our philosophers that there is also another kind of rational discussion: a critical discussion, which does not seek to prove or to justify or to establish a theory, least of all by deriving it from some higher premises, but which tries to test the theory under discussion… (Popper 1994: 60; emphasis in original)
One of the components of modern irrationalism is relativism (the doctrine that truth is relative to our intellectual background, which is supposed to determine somehow the framework within which we are able to think: that truth may change from one framework to another), and, in particular, the doctrine of the impossibility of mutual understanding between different cultures, generations, or historical periods – even within science, even within physics. […]
The proponents of relativism put before us standards of mutual understanding which are unrealistically high. And when we fail to meet those standards, they claim that understanding is impossible. Against this, I argue that if common goodwill and a lot of effort are put into it, then very far-reaching understanding is possible. Furthermore, the effort is amply rewarded by what we learn in the process about our own views, as well as about those we are setting out to understand. (33–4)
Popper thus makes two points. First, the doctrine of paradigm incommensurability errs in its assumption that mutual understanding or communication between paradigms is impossible. Rather, as he points out later in the book (48–53), communicating between paradigms is like learning a different language: it is difficult but not impossible. The proponents of paradigm incommensurability confuse difficulty and impossibility. In other words, they do not make an appropriate effort and are content simply with working within one paradigm, and they justify their lack of effort by saying that another paradigm is incommensurable with their own. The doctrine thus represents a comfortable excuse for not making oneself familiar with a different theory or method.
Second, and more importantly, Popper argues that the doctrine of paradigm incommensurability errs with regard to scientific progress. Certainly, scientific progress can take place within the framework of a single theory. However, this is only one possibility, and probably not the best one. Rather, a mutual critique represents the central basis for scientific progress, and in particular a mutual critique between different theories and their assumptions. Rather than a declaration of intellectual independence, the comparison and mutual critique of different theories is the hallmark of scientific progress, in both the natural and the social sciences.
There is a third shortcoming of paradigm incommensurability, and Popper considers this one not only futile but even dangerous: the encouragement and justification of intolerance between theories.
The myth of the framework can be stated in one sentence, as follows. A rational and fruitful discussion is impossible unless the participants share a common framework of basic assumptions or, at least, unless they have agreed on such a framework for the purpose of the discussion. […] As I have formulated it here, the myth sounds like a sober statement, or like a sensible warning to which we ought to pay attention in order to further rational discussion. Some people even think that what I describe as a myth is a logical principle, or based on a logical principle. I think, on the contrary, that it is not only a false statement, but also a vicious statement which, if widely believed, must undermine the unity of mankind, and so must greatly increase the likelihood of violence and of war. (34–5)
Thus, in addition to the concern that the assumption of paradigm incommensurability limits scientific progress, we have an even more forceful statement here. Thinking in terms of incommensurability immunizes one's viewpoint against critique, because one can always say: “Your critique of my viewpoint is not valid because your paradigm is a different one,” or “… because your ontological assumptions are different from mine.” This self-immunization against critique fosters a belief in the infallibility of one's standpoint and nurtures intolerance of others – other theories, other ways of thinking, and, ultimately, other ways of life. As a result, thinking in terms of paradigm incommensurability increases the likelihood of using force, communicatively and – ultimately – physically, against people who think differently.The remainder of this book is based on these principles and draws on the four theories outlined above. These theories do not need to agree and the degree to which they complement or contradict each other will vary. But in their complementarities or mutual critique they will illuminate the aspects of management consultancy in a way that individual theories could not achieve. From now on, the phenomena of management consultancy shall take center stage, and the theories will be taken as searchlights to shed light on them.
1For a summary of this debate, see Scherer (1998). For significant contributions, see Gioia and Pitre (1990), the edited volume by Hassard and Pym (1990), Jackson and Carter (1991), the debate in Organization Studies, 1993, volume 14, no. 5, and the “Comments” in the special issue of Organization, 1998, volume 5, no. 2.