Early Choices That Are Hard to Undo

The idea

Some choices are cheap to reverse. Others get harder to undo every day they stay in place, because other things grow around them. A file format chosen in a hurry becomes the format every tool depends on. A habit chosen once becomes the default. Economists call this path dependence. In W. Brian Arthur's models, small chance events early on can tip a whole market toward one technology, which then stays because everyone has built on it.

Why it attracts me

It joins two ideas I care about. Every choice has a cost in the options it closes off (What Every Choice Gives Up), and choices that cannot be undone make that cost permanent. And the future is a tree of branches (Mapping the Futures a Choice Opens), some of which close for good once I pass them.

An example

Imagine a small team laying out a new workshop. Putting the heavy press in the middle seems fine on day one. Five years later, power lines, walkways and storage racks all sit around it, and moving it would mean a week of downtime. The early choice is now a constraint, even though nobody remembers why it was made. A rule of thumb follows: move quickly on choices that are easy to undo, and slow down on the ones that are not.

Where it connects

Across generations, long chains of choices become legacy (What I Pass On to the Next Generation). What a family passes on is often the result of small decisions that hardened over time.

Questions I'm still exploring

• How can I tell early which choices will be hard to undo?
• Which of my current limits come from a reason that no longer exists?
• When is it worth paying the high cost of switching paths?

Sources and further reading

• W. Brian Arthur, "Competing Technologies, Increasing Returns, and Lock-In by Historical Events," The Economic Journal 99(394): 116–131 (1989)