Make Code Changes Expensive Enough and Nobody Makes Them at All, Actually

The article introduces the 'Reverse Jevons Paradox': just as cheaper resources can increase total usage (classic Jevons), making a resource more expensive can drive total spend to zero. In software, piling red tape onto code changes—multi-level reviews, ticket webs, sign-off hordes—raises the cost of small refactors past the break-even point, so developers stop attempting them entirely. The result isn't slower incremental improvement; it's the complete disappearance of that improvement stream, which is disastrous for codebase health.