Other Mental Models · OM-29
The specific sequence of decisions and events that got a system to its current state can lock in a particular outcome, even a genuinely suboptimal one, simply because reversing course has become more costly than continuing — the past shapes the range of what's realistically possible now, not just what happened before.
A concept describing how decisions and events available at one point in time can be constrained or determined by the specific sequence of decisions and events that preceded them, such that a system can become locked into a particular trajectory — even a suboptimal one relative to some hypothetical alternative — because the accumulated costs, investments, and structural commitments of the path already taken make switching to an alternative prohibitively expensive or otherwise difficult, regardless of the alternative's underlying merit.
Developed extensively in economics by Paul David and W. Brian Arthur beginning in the 1980s, with David's widely cited 1985 paper on the QWERTY keyboard layout serving as a canonical illustrative case of a technology standard persisting due to accumulated switching costs, rather than continued technical superiority.
The Mechanism
Once enough has been built on top of a decision, undoing it costs more than living with it
The QWERTY keyboard layout, originally designed in the 1870s partly to reduce mechanical typewriter jams by spacing out commonly paired letters, has persisted for well over a century — even though subsequent keyboard layouts have been proposed with claimed typing-efficiency advantages, because the accumulated cost of retraining an entire population of typists and manufacturers away from the entrenched standard has, in practice, exceeded the benefit any alternative layout could plausibly offer.
01 · SWITCHING COSTS, NOT CONTINUED SUPERIORITY, ARE WHAT SUSTAIN THE LOCKED-IN PATH
This is the specific mechanism distinguishing path dependence from ordinary continued good performance
A path-dependent outcome persists specifically because the accumulated cost of switching to an alternative exceeds that alternative's benefit — not because the original path remains objectively the best option available; distinguishing genuine path-dependent lock-in from a path that simply remains the best option on its own merits requires actually estimating what a switch would cost versus what it would gain, not assuming persistence itself proves continued superiority.
02 · EARLY, SEEMINGLY MINOR DECISIONS CAN HAVE DISPROPORTIONATE LONG-RUN CONSEQUENCES
The timing of a decision within a path's development matters as much as the decision's own content
Because each additional layer built atop an initial decision increases the cost of later reversing it, decisions made early in a system's development — before much has accumulated on top of them — can have outsized long-run influence on the system's eventual trajectory, even when those early decisions were made under significant uncertainty or for reasons that later prove to have been comparatively minor or arbitrary.
03 · RECOGNIZING PATH DEPENDENCE EARLY CREATES A DELIBERATE OPPORTUNITY TO CHOOSE A BETTER PATH BEFORE LOCK-IN OCCURS
The practical value of the concept is largely in anticipating, not undoing, lock-in
Because switching costs accumulate progressively over time, the window for changing course at comparatively low cost is largest early in a system's development and narrows considerably as more gets built on top of the initial choice — recognizing a potentially consequential early decision as such, and investing extra deliberation in it before path dependence sets in, is a far more tractable intervention than attempting to reverse a deeply entrenched path later.
Where It Fails / Inversion
Where it fails / inversion
Not every persistent pattern reflects genuine path-dependent lock-in — some outcomes persist simply because they remain the objectively best available option on their own current merits, entirely independent of any switching-cost lock-in effect, and attributing persistence to path dependence without actually checking whether a genuine switching-cost barrier exists can lead to unwarranted assumptions that change is harder than it actually is.
How To Use It
Worked example · evaluating whether to migrate away from an entrenched legacy system
A company evaluating whether to migrate away from an aging but deeply entrenched internal software system should distinguish between two possibilities: the system remains genuinely the best available option on its current technical merits, or the system persists mainly due to path-dependent switching costs (integrations, trained staff, accumulated custom workarounds) that now exceed the benefit a superior alternative would provide — the correct decision depends entirely on which of these is actually true, which requires a real cost-benefit estimate of the switch rather than assuming persistence itself is proof of either continued merit or pure inertia.
How to use it
When a system, process, or standard has persisted for a long time, ask whether it remains genuinely the best available option, or whether it persists mainly because the accumulated cost of switching away from it now exceeds any plausible alternative's benefit — the two situations call for very different responses, and only an honest cost-benefit estimate of an actual switch reveals which one you're facing.
See Also