DISCUSSION

81

manufacturer, in a directional distribution setting, requires high investment in tangible assets (the Y axis) and they contribute to the product value but only little to procedural activities (the X axis), the slope will be steep but the length of activities short. Hence, complex distribution will typically cross a steeper slope from the use of non-codified entrepreneurial activities and use of idiosyncratic resources, but possibly also a longer stretch of activities before the product is sold.

The flatter the slope, i.e., the lower the value creation, and the shorter the stretch, i.e., the fewer activities to integrate, the less problematic the governance challenges. The steeper the slope, particularly if the value creation derives from distributed idiosyncratic plastic resources, and the longer the stretch of activities, the higher is the difference in the resources and capabilities contributed by two sequential firms.

82

incentives are effectively internalized through residual ownership to value-creating assets without changes in organizational structures.

In long-linked value chains (Thompson, 1967) where distribution serves as a necessary mediating instrument to bridge the manufactured products with downstream product markets, various interdependencies between adjacent value chain activities emerge. The deployed organizational technology must satisfy the demands for both instrumental and economic efficiency (ibid) in upstream manufacturing as well as downstream distribution. However, the coordinating responsibility and authority as well as the incentives to conduct activities are very different depending on the role of distribution as a conduit to the final product market.

Under directional distribution, product complexity originates from resources, capabilities, and entrepreneurial efforts that reside with the upstream manufacturing processes. The adjacent distribution activities are linked to manufacturing that defines the asset specific tangible investments needed to secure that the products are sold and distributed with only a limited need to focus on intangible distribution investments. The economic base for this governance approach are the rents created from the competitiveness of the manufacturer’s products and ability to codify two things: the activities and investments. Both are needed to accommodate the downstream distribution. The coordination of the long-linked mediating interdependencies between manufacturing and distribution is managed through planning and standardization of processes that are instigated through the manufacturer’s authority. In this way the technological core of the manufacturing firm is sealed off from potential influences of fluctuations in the product market, thereby ensuring operating efficiency in manufacturing.

To attract the necessary specific investments and efforts from the market bridging distribution entities enforcing the coordinating authority of manufacturing there must be sufficient incentives and rewards to motivate the distribution agents. This is done by manufacturing sharing the rents in amounts that, at a minimum, are equal to the distributors’

external opportunity costs (Klein, 1995; Lafontaine and Raynaud, 2000). In short, the manufacturer has sourced authority and coordination rights from the distributors to decide on resource allocation and specific asset investments by sharing rents and that align their interests (Simon, 1951; Williamson, 1975).

83

However, as observed from the empirical research on forward integration, directional distribution is not immune to transaction costs. These costs relate to incentive misalignment from sequential monopolies. If the manufacturer holds some kind of brand recognition in the final product market, the distributors can free-ride on the brand value to internalize profits and externalize costs to the manufacturer and other distributors (Carves and Murphy, 1975). While the required investments in specific tangible distribution assets (Klein et al., 1978; Klein, 1995;

Williamson, 1979, 1985) are relative easy to police, they can be subject to moral hazard if they are plastic (Alchian and Woodward, 1988). However, the implied self-enforcing contracts (Klein, 1995; Lafontaine and Raynaud, 2000) will lose their correcting effect if the distributors’

efforts cannot be related (directly) to output, or the relationship is difficult to verify. Therefore the manufacturer might share quasi rents by paying a provision without necessarily receiving a return from the distributors’ effort.

This situation has been a main driver of forward integration, because it improves the ability to measure the effort provided by the distribution (e.g., Anderson and Schmittlein, 1984;

Brickley and Dark, 1987; Kosová et al., 2013; Woodruff, 2002). The double marginalization problem (Eccles, 1985; Riordan, 2008) can be exacerbated by information asymmetry that may cause the manufacturer to share more quasi rents with distribution than is necessary thereby diluting the profits of the upstream manufacturing. In this context it is worth noting that empirical studies find that the need to incentivize upstream (manufacturing) effort leads to more forward integration (e.g., Lafontaine, 1993; Lafontaine and Shaw, 2005; Nickerson and Silverman, 2003; Scott, 1995).

The complex distribution setting presents diverse challenges. The competitiveness of this long-linked, mediating, and intensive technology (Thompson, 1967) depends on both manufacturing and distribution contributing to the product complexity in final product markets.

The role of distribution bridging the manufactured outputs into the final goods and services required by the end-users is more complex. The distribution must engage own value-adding resources and capabilities that are specific to the distribution activities. This includes specific market knowledge that often is tacit in nature and embedded in idiosyncratic customer relationships that can ensure end-user satisfaction and loyalty for future business engagements (Brickley and Dark, 1987; Kosová et al., 2013; Lafontaine and Slade, 2007). This makes the interdependencies and incentives between manufacturing and distribution very different

84

compared to the conditions under directional distribution. This more symmetric location of needed resources and capabilities with reciprocal interdependencies requires a governance approach, where responsibilities and authority are distributed more evenly between the value chain activities (Aghion and Tirole, 1997; Eccles; 1985; Galbraith, 1983; Thompson, 1967).

This has important implication in several ways. With the distribution using its own idiosyncratic resources and capabilities to accomplish entrepreneurial value-adding activities towards final customers, importance increases due to the reciprocal interdependencies (Thompson, 1967). The knowledge generated by downstream distributors serves as important coordination input to the upstream manufacturer as the means to improve the intermediary product, based on mutual adaptation between manufacturer and distributors (ibid). Furthermore, high plasticity (Alchian and Woodward, 1988) of distributor specific assets and resources makes the codification of the distributors’ efforts to the final product complexity virtually impossible.

The plasticity of specific assets combined with asymmetric knowledge makes measures of the distributors’ input very imprecise and costly to generate (Alchian and Demsetz, 1972). Hence, it is clear that possible transaction costs related to the distributors’ behavior are higher compared to the situation under directional distribution, where improved metering of effort is the answer.

Therefore segregated ownership of assets where each party remains the residual claimant to their own unobservable inputs like behavior and entrepreneurial effort is the proposed answer to deal with this situation consistent with findings in empirical studies (Brickley and Dark, 1987;

Kalnins and Lafontaine, 2013; Kosová et al., 2013; Lafontaine and Slade, 2007; Woodruff, 2002).

This discussion relates mostly to moral hazard issues, where the prioritization of entrepreneurial effort and behavior are necessary precursors to future profits. The inability to observe entrepreneurial effort and multitasking in downstream distribution is not necessarily a cover for potential moral hazard costs (Shepard, 1993; Slade, 1996), but may reflect a necessary prioritization of effort to generate future cash flows from specific assets ownership (Baker and Hubbard, 2004; Brickley and Dark, 1987; Woodruff, 2002). The ability to profit as a residual claimant to assets ownership incentivizes and rewards these otherwise non-contractible and unobservable efforts (Grossman and Hart, 1986; Holmström and Tirole, 1991). Firms that choose to integrate forward in a complex distribution setting may incentivize unobservable effort using aggregated indicators related to asset value from business unit performance

85

(Chandler, 1977; Holmström and Tirole, 1991; Lazear and Gibbs, 2014; Oxley and Pandher, 2016; Silverman and Ingram, 2017) when no direct observable indicators are available. This presents a governance dilemma. While incentivizing unobservable entrepreneurial effort related to the performance of distribution will enforce the prioritization of own asset value, it also removes the coordination advantage from integrated ownership.

Looking at complex distribution from Holmström and Milgrom’s (1994) view of the firm as an incentive instrument reveals some interesting insights related to the boundary conflicts discussed above. By providing the integrated distributors with job security to attain control over the integrated distribution (Coase, 1937; Simon, 1951; Williamson, 1975), the responsibility and authority to direct distribution effort is placed squarely with the manufacturer. The challenge here is that manufacturing does not have the necessary knowledge about the product complexity or the capabilities to codify this. So, while common asset ownership can remove opportunistic appropriation of quasi rents (Klein et al., 1978; Williamson, 1979, 1985), it also mutes the incentives to integrate the distributors’ unobservable personal effort when they are not residual claimants (Grossman and Hart, 1986; Holmström and Tirole, 1991; Kosová et al., 2013;

Silverman and Ingram, 2017; Woodruff, 2002). This is particular relevant under complex distribution where idiosyncratic resources and competences needs to be incentivized. A similar issue arises when the direct indicators of multitasking efforts are imprecise (Alchian and Demsetz, 1972; Holmström and Milgrom 1991). Under directional distribution, the specific asset investments are defined by the manufacturer. Under complex distribution, the distributors use their own idiosyncratic knowledge to decide on investments in specialized assets and resources. For a manufacturer that chooses to integrate forward, this makes job design more problematic because the downstream competitiveness is embedded in idiosyncratic resources that are unknown to the manufacturer. A job design that removes the distributors’ prioritization of idiosyncratic resources requires that it overall would be more profitable to prioritize the manufacturer’s assets as under directional distribution; or, the common incentives from specific assets along the long-linked value chain can be somehow aligned. These issues illustrate the conflicts in practice between different theories and their prescriptions about the governance of forward integration.

Another aspect of directional and complex distribution is whether these governance approaches are static or evolve as a function of a changing market context. Some manufacturing

86

firms have a product where the complexity in the final product markets are so valuable that it is insensitive to minor developments in product complexity in the distribution industry. Yet, other manufacturing firms may be faced with increasingly complex and dynamic market contexts for their final products that they are gradually forced to consider a new governance approach more that is appropriate to deal with complex distribution. Major changes in the market of final users may increase product complexity over time, and disruptive competitive innovations in distribution may cause dramatic changes on demand conditions that require a shift from directional to complex distribution. For a path dependent manufacturing firm operating a directional distribution the requirement to innovate in downstream advanced value, adding activities exploring for new resources and capabilities can pose severe challenges (Galbraith, 1983, Nooteboom, 2004; Teece, 2010; Teece, Pisano and Schuen, 1997). Accordingly, empirical studies (Benedettini, Neely and Swink, 2015; Gebauer, Fleisch and Friedli, 2005; Neely, 2008;

Visnjic, Weingarten and Neely, 2016) find that manufacturing firms often struggle with the transformation from being driven by the manufactured product into profiting from downstream complex value adding activities.