IBM and Linux (A)

Carliss Y. Baldwin; Siobhan  O'Mahony; James  Quinn

Case | HBS Case Collection | June 2003

IBM and Linux (A)

by Carliss Y. Baldwin, Siobhan O'Mahony and James Quinn

Abstract

In the fall of 1998, Dan Frye, member of IBM's emerging technologies and business team, is trying to decide whether to forge a strategic alliance with the Linux Development Community (LDC). Just two years earlier, IBM had its first exposure to an "open source" software program when it selected Apache as the web server program for the web site of the Atlanta Summer Olympic Games. Based on its success with Apache, and Frye's intuition that Linux could be a critical, strategic step in the new "network-centric" computing environment, Frye and his colleagues are trying to decide whether an alliance between IBM and LDC would serve their interests--and, if so, how IBM could manage the alliance with a loosely affiliated group of software developers.

Keywords: Mergers and Acquisitions; Open Source Distribution; Problems and Challenges; Alliances; Cooperation; Computer Industry; Information Technology Industry;

Format: Print

19 pages

Educators Purchase

Citation:

Baldwin, Carliss Y., Siobhan O'Mahony, and James Quinn. "IBM and Linux (A)." Harvard Business School Case 903-083, June 2003.

In March 2013, Apple Computer has a very large cash balance, and is under pressure to return cash to shareholders. Hedge fund manager David Einhorn thinks Apple can "unlock value" by issuing perpetual preferred stock, dubbed iPrefs. Henry Blodget, CEO of Business Insider, disagrees, saying "you can't just wave your magic wand and make something of nothing." This short case is designed to support a discussion of "perfect" markets and the Modigliani-Miller capital structure irrelevance propositions. The case focuses on two questions: (1) From a shareholder's perspective, how is Apple's cash different from cash in a bank or money market account? (2) Can Apple create significant value for shareholders by splitting each common share into an iPref plus a common share?

The purpose of this chapter is to investigate the means by which firms capture value in open platform systems. I begin by arguing that the surplus value created by complementarities within a technical system will be split among the owners of the unique and essential components—the strategic bottlenecks in the system. However, most platforms must also execute a series of steps that are subject to “flow production” bottlenecks. Finding and fixing these flow bottlenecks is another way to capture value. In addition, two types of platform improvements provide further opportunities for value capture. “Accelerators” speed up the processing of options, while “subsidiary” platforms increase the range of options available to users. Finally, members of a platform system or new entrants may seek to supplant the owner of a strategic bottleneck by “disintermediating” platform components. I describe four generic methods of disintermediation: substitution, reverse engineering, platform independent complements, and platform envelopment.

The purpose of this chapter is to use the theory of bottlenecks laid out in previous chapters to better understand the dynamics of an open standards-based platform. I describe how the Wintel platform evolved from 1990 through 2000 under joint sponsorship of Intel and Microsoft. I first describe a series of technical bottlenecks that arose in the early 1990s concerning the “bus architecture” of IBM-compatible PCs. Intel’s management of buses demonstrates how, under conditions of distributed supermodular complementarity, a platform sponsor can reconfigure the modular structure of a technical system, property rights within the system, and its own zone of authority to increase system-wide throughput, while protecting its own strategic bottleneck from disintermediation. I go on to describe how Microsoft used platform envelopment to establish a second strategic bottleneck in productivity software and later to respond to the threat of disintermediation from platform-independent internet browsers. I end the chapter by discussing the conditions under which shared platform sponsorship can be a long-term dynamic equilibrium.

Faculty & Research

IBM and Linux (A)

Abstract

Citation:

About the Author

More from the Author