causal

Interaction Effects Defeat One-at-a-Time Optimization

Testing 200 types of oregano sequentially while holding everything else constant ignores the economics of interaction effects: because ingredients combine in vast numbers of combinations, sequential one-variable-at-a-time optimization cannot find the true optimum, and the universe lacks enough time to try all combinations of pizza ingredients.

causalpending

Speaker

Russ Roberts

Evidence Quote

“he's probably a little weak on what we would call an economics interaction effect... there's a lot of combinations. The universe probably doesn't have enough life left in it or ever to try all the combinations”

Source

A.J. Jacobs on Solving Life's Puzzles— EconTalk
Created: 6/13/2026, 3:37:29 AM

Your notes

Loading notes...