Path Dependence

Path dependence emerged as a formal concept in the work of economists Brian Arthur and Paul David in the mid-1980s. David's celebrated 1985 paper on the QWERTY keyboard layout argued that an early technological choice, made for specific historical reasons, became locked in through increasing returns and network effects, persisting long after the original rationale had disappeared. Arthur, working in parallel, showed how small early events can tip markets and technologies toward one outcome among several equally viable possibilities, and how those outcomes become self-reinforcing over time.

The concept has profound implications for strategic management. Once a company commits to a particular technology platform, organizational structure, or market position, that choice generates investments, relationships, skills, and expectations that make it progressively more costly to switch to an alternative path. This is not simply about sunk costs; it is about the entire ecosystem of complementary assets, learned routines, and stakeholder expectations that build up around a strategic choice. Intel's commitment to the x86 processor architecture, for instance, was reinforced by decades of software development, manufacturing investment, and customer relationships that made switching architectures enormously costly, both for Intel and for the broader computing ecosystem.

Microsoft's dominance of the desktop operating system market provides perhaps the most studied case of strategic path dependence. The initial adoption of MS-DOS by IBM in 1981 was partly a historical accident; IBM's first choice, Digital Research's CP/M, fell through due to a failed negotiation. But once MS-DOS and later Windows gained market share, network effects and switching costs created powerful increasing returns. Software developers wrote for Windows because users were there; users stayed on Windows because the software was there. By the time technically superior alternatives existed, the installed base and ecosystem made switching costs prohibitive for most users and organizations.

Path dependence also explains why organizations struggle to adapt even when they recognize the need for change. Clayton Christensen's work on the innovator's dilemma shows how successful companies become locked into serving their existing customers with sustaining innovations, making it nearly impossible to pursue disruptive technologies that initially serve different markets. A Kodak engineer built the first digital camera in 1975, but the company's massive investments in chemical film processing, its dealer network, and its profit model created a path-dependent trajectory that made meaningful pivoting extraordinarily difficult.

For strategists, path dependence counsels two things. First, early-stage decisions deserve disproportionate attention because they shape the trajectory of all subsequent choices, often in ways that are difficult to reverse. Second, breaking free from an unfavorable path typically requires a crisis, a change in leadership, or a deliberate strategy of creating a separate organizational unit insulated from the legacy path. IBM's successful pivot to services under Lou Gerstner and Amazon's creation of AWS as an independent unit both illustrate how companies can overcome path dependence when they recognize its grip and take dramatic action.