SPECIALIZING INGENERALITY:

FIRM STRATEGIES WHEN FACTOR MARKETS WORK

Raffaele Conti

Catolica Lisbon School of Business and Economic

Alfonso Gambardella

Department of Management & Technology and CRIOS

Bocconi University, Milan

Elena Novelli

Cass Business School, City University of London

March 2016

PRELIMINARY – COMMENTS WELCOME

ABSTRACT

We study the interdependence between two strategic decisions of firms in vertical industries in which an asset is used to produce downstream products but can also be sold in intermediate markets to final producers. Our theory suggests that when intermediate factor markets are efficient, the strategic decision to specialize in selling upstream assets without entering downstream markets is complementary to the strategic decision to invest in making the assets fungible to many different downstream markets. We test our predictions using a sample of firms in the US laser industry between 1993 and 2001. A regulatory shock that affects the cost of downstream entry provides the setting for a quasi-natural experiment that corroborates our prediction. Apart from providing a direct test of the complementarity between two strategic choices (no-entry and investment in fungible assets), our study highlights the potential for exploiting economies of scope through markets rather than internal organizations.The traditional view by Penrose (1959) or Nelson (1959) is that fungible assets are the domain of larger firms that invest in fungibility because they can exploit the asset internally. Interestingly, their prediction would be opposite to what we find because in their case higher downstream entry costs ought to reduce fungibility of internal assets. Our result suggests that exploiting economies of scope inside an organization is bounded by the lack of well-functioning vertical markets.

INTRODUCTION

Valuable resources are at the core of firms’ competitive advantage (Barney 1991; Penrose, 1959). They constitute the essence of the “product opportunity set” of a firm, encompassing “all of the productive possibilities that its `entrepreneurs´ see and take advantage of” (Penrose 1959, p. 28). Drawing on this opportunity set, firms undertake strategic actions possibly resulting in greater profits. However, resources differ in their ability to be deployed in multiple uses, and so, in providing firms with strategic options – a characteristic referred to as the generality or general-purpose nature of the resource, or “fungibility” (e.g., Kim Bettis, 2014). Indeed, some resources are inherently specific to some applications while others are more fungible and so can be more easily reconfigured into alternative uses (e.g., Helfat Eisenhardt, 2004).

The leading argument underlying research in this area is that the more fungible a resource is, the higher the likelihood that it can be re-used in a new business, thus determining entry into new markets (e.g. Nelson, 1959; Penrose, 1959). More in detail, Penrose (1959) argues that firms grow by exploiting internal synergies, which arise from the possibility of redeploying fungible resources into new businesses at low or even zero cost. In a similar vein, Nelson (1959) suggests that diversified firms are most likely to internalize the externalities arising from the breadth of opportunities created by investments in basic research, which is conducted without a specific application context in mind and so might lead to inventive outcomes applicable across multiple settings. This view is still at the core of the RBV theory of entry and diversification (e.g., Nason Wilklund, 2015).

Yet, one of the core assumptions of the RBV theory for internal expansion is that “strategic factors markets” – where both resources and their deriving services can be traded (Dierickx Cool, 1989) – are not perfect or at least do not operate smoothly. In fact, while scholars in this area emphasize that “without market failure due to high transactions costs or imperfect mobility, the firm could simply sell the services of their redundant resources” (Peteraf, 1993, p. 183), they also tend to believe that market for resources are inherently imperfect (e.g., Teece, 1980) and, as such, firm growth is actually the only option firms have for exploiting their fungible resources. This paper contributes to the RBV theory in two ways: (a) it argues that in industries with well-functioning strategic factor markets, investment in fungible resource might be complementary to trading those resources in intermediate markets; (b) it explores the contingencies that determine the strength of this complementarity.

When strategic factor markets do not work, the only choice of firms to profit from their idle, fungible, resources is to expand into a new market. The incentive to do so is contingent on the level of economies of scope experienced by the firm. The lower the adaptation costs required for employing a resource in an additional downstream market, the higher is the incentive to enter, i.e. the stronger is the complementarity between investing in fungible resource and entering downstream. But the incentive to employ a fungible resource in a new market also depends on the costs of downstream operations in eachmarket after entry. It follows that if a firm faces an increase in these costs, it reduces the complementarity between resource fungibility and entry. Hence, in the Penrosian world, firms would be likely to respond to an increase in downstream costs by reducing the fungibility of their resources.

However, when strategic factor markets work, the outcome may be just the opposite. Firms can trade their excess resources in intermediate markets. This implies that the net benefits arising from using a fungible resource to enter in a downstream market have to be weighed against the net benefits from selling that resource in the intermediate markets. To the extent that the latter benefits overcome the former, investing in more fungible upstream resources and trading those resources – rather than entering downstream – may become complementary strategic choices. Hence, when the option of selling resources in intermediate markets is available, we discuss the conditions under which reduced opportunities to operate downstream (or greater opportunities to trade in intermediate markets) encourage firms to specialize upstream and increase the fungibility of their resources. We also explore the conditions under which we are more likely to observe this phenomenon. In particular, we show that the complementarity between investing in fungible resources and trading them in intermediate markets is stronger for firms that (1) have not yet entered into any downstream market or (2) face a homogeneous – as opposed to skewed – distribution of potential buyers of their services in the downstream markets.

Our framework evokes some important strategic decisions of firms in recent years, which we briefly discuss in the next section as a motivation to our analysis. For example, it echoes one of the most significant strategic turnaround at the turn of the last century: the transformation of IBM from a mainframe producer to a service-based firm. As suggested Louis Gerstner (2002), the CEO who engineered this transformation between 1993 and 2002, the shock that triggered the change was the rise of UNIX and the PC, and the fact that IBM did not have a comparative advantage in these new businesses. However, over the years, it had developed a significant general capability of applying technologies to solve business problems. This turned IBM into a company that served “… customers [who] had a problem with a product from Digital, Compaq, or Amdahl” (p.128) – that is,IBM’s competitors in the PC business. Similarly, IBM realized that its considerable stock of technologies could be licensed making USD 1.5 billion in income in 2001 from $500 million in 1994,that the technological components that it used to make for itself could be used by many other firms, and that it could provide technological solutions to a wide range of problems of its business clients.

We test our predictions using a sample of firms in the US laser industry between 1993 and 2001. This is an ideal setting for our theoretical framework for several reasons. First, the crucial resource in this industry is the laser technology itself, and each laser technology might be more or less fungible: that is, it may have a lower or higher number of applications, each one linked to a specific submarket. In the particular time period taken into account, the applications of laser – and the corresponding submarkets – expanded considerably. Furthermore, the laser industry is vertically disintegrated: it is populated by laser producers – which produce the laser as a standalone technology – and laser system manufacturers – which embed the lasers into a “laser system” ready-to-use. This implies that intermediate strategic factor markets for the crucial resource (that is, laser technology as a standalone component) exist and work smoothly.

Finally, in the time period of our sample we can exploit an exogenous regulatory shock that increased the cost of downstream operations for firms operating in some US States, which allows for testing complementarity in investment in fungible resources (lasers) and in trading them. In so doing, we follow Brynjolfsson and Milgrom’s (2012, p. 58) suggestion that “legal and institutional changes are often ideal candidates to … [estimate complementarities in organizations] … because a change in a law or government policy can provide a precise date and specific geographic area or jurisdiction for the change to occur.” Our difference-in-difference analysis tests whether firms located in states enacting the regulation (treatment group) – so increasing their cost of entry into new downstream markets – are more likely both to invest in fungible technological resources and trade them (rather than enter in new downstream markets), compared to firms in the other states (control group). We employ both bivariate probit and linear probability models for the four combinations of investing/not investing in fungible resources and entry/trade, and find results consistent with our predictions.

In the following sections, after a discussion of the background of our research, we present an intuitive framework for our theory. We then present our data and methodology, discuss our results (including robustness checks), and conclude. The Appendix provides a full-fledged formal model capturing our main ideas.

THEORETICAL FRAMEWORK

Background

Anecdotal evidence is consistent with the increasing importance of the strategy to invest in fungible resources to be traded in intermediate markets when firms face high downstream production costs. The most valuable resource of IDEO, a leading design company and known for pioneering a new business model, is the overall creative process for designing new ideas. This resource is extremely fungible as it can be applied for designing new products in multiple market domains, including, for instance, electronics, robotics, and apparel. However, the production costs to operate in several of these potential markets are high. Hence, taking advantage of corporate downsizing in the 90’s and the creation of “markets for designing”, IDEO decided to trade the services deriving from its creative process to companies in several different markets – such as Apple Computer, AT&T, Samsung, Philips, Amtrak, Steelcase, Baxter International, and NEC Corp. Similarly, Echelon, an industrial automation company, developed a universal automated control system (Lonwork) with applications in sectors as different as elevators, manufacturing processes, cars, or utilities (Thoma, 2009). Because of the high entry costs in these markets, largely due to scale requirements, the company deliberately chose to not enter downstream, focusing instead on increasing the fungibility of its controller and expanding its span of applications (see also Gambardella McGahan, 2010).

As noted in the introduction, the transformation of IBM from a mainframe and computer producer to a general-purpose technology service company is another example of our story. The shock came from the rise of UNIX, the open system environment developed by Sun and HP, and the PC. Not only did this produced a tough competitive environment for IBM’s core business, but IBM was unable to react, or at least it did not have a comparative advantage with respect to its competitors. It is not uncommon that even successful companies like IBM, once under attack by new disruptive technologies, lose their war (Christensen, 1997). However, IBM sought a different strategy from the two natural reactions that we most often think of: resist to the competitive threats by improving the core business under attack, or trying to enter in the new business. As Gerstner (2002) puts it:

It would have been easy to follow HP and UNISYS and all the rest down this path. All of the pundits who followed the industry saw the dominance of this model inevitable. It would also have been easy simply to be stubborn and say that the changeover wasn’t going to happen, then fight a rear-guard action based on our historical view of a centralized computing model. What happened, however, is that we did neither. … we decided to stake the company’s future on a totally different view of the industry.” (p.123)

This different view was to become a supplier of “solutions” that integrate different technologies. Instead of serving customers, typically firms, with the PC or other specific products, IBM served them with general-purpose technologies and services to improve their processes. In so doing, IBM built upon its quintessential asset: “In services you don’t make a product and then sell it. You sell a capability. You sell knowledge.” (p.133) While the company applied these service capabilities only internally or to its own product, now it supplies them to many clients and for many products. Similarly, IBM has opened the “company store,” as Gerstner puts it, ranging from technology licensing to the sale of standard components or sophisticated services. With this strategy, IBM has reached the very same clients who use the products of companies in the software, networking, computing, or communication business in which IBM could have entered if it had bet on the idea that this was, inevitably, the way to go.In our framework firms that decide to specialize in generality serve the downstream firms that enter the new markets. More than serving these firms IBM serves the clients of these firms in the many downstream markets that it can reach with its fungible capability. But this is a detail. The thrust of our argument is that there is complementarity between the strategic decision not to enter downstream markets and the strategic decision to expand the fungibility of the firm’s resources to serve different markets – that is, the decision to increase the scale of operations by going broadly across markets than deeply with a large volume of activities in one market.

Why not all companies do so then? For one reason, you need a shock that triggers the decision to change business model – from entry and operations downstream to serving many markets with a fungible resource and focusing upstream. Moreover, the change in business model is not without costs. Gerstner notes that the change hit the traditional culture of IBM, and it was not easy to implement in practice, eliciting a lot of the diseconomies of scope, between different business cultures, highlighted for example by Bresnahan et al. (2011).

AnItalian firm specialized in producing military aircrafts provides a good example of a company that had the option to change, but did not. An engineer found that a technology for Ground Support Equipment (GSE, the support equipment used to service aircrafts between flights) could also be used, with proper modifications (e.g., modifying the fluids’ capacity or the hydraulic pumps) as a general technology for a larger variety of military aircrafts of similar size. The company was using the technology in-house and it was only selling the GSE as a bundle to its own product. It considered seriously the opportunity to make GSE more fungible to sell it to many aircraft producers, possibly pulling out or at least reducing its stakes in the downstream market. However, there was no reason to abandon its traditional business, there was no trigger or shock. As a result, the company decided to stay in the downstream market without investing in the fungibility of the technology. This helps us to emphasize that our paper makes no claim that specialization in generality is a better strategy, but only that, under some conditions, no-entry and generality are complementary strategic choices.

We would also like to note a few other ingredients of our framework. First, we focus on a shock to the downstream production costs of some firms as our trigger to switch between the two business models. This is the shock that we observe in our empirical analysis and we find it convenient to use it to illustrate our theoretical framework. However, we can think of other shocks, particularly factors that increase the efficiency of vertical market transactions. Second, our framework elucidates other aspects of the choice between transactions and integration. In particular, transaction cost economics suggests that stronger bargaining power of downstream buyers induce upstream suppliers to integrate downstream to reduce the hold-up problem. We show that staying upstream and investing in fungibility is another potential response of upstream firms. In so doing, our framework highlights another angle of the problem. Hold-up depends on asset-specificity, which transaction cost economics often takes as a given. We argue that asset-specificity can be a matter of choice, and firms, particularly upstream firms, can choose to make their assets less idiosyncratic to specific applications.