Why Good Intentions Produce Bad Systems: Why the structure, not the sincerity, decides the outcome

I built a smaller version of that trap with my own hands. In the workshop I once tied a bonus to units completed per day, meaning to reward the people who worked hard and finished. What I actually rewarded was speed, so speed is what I got, and a unit finished wrong still counted as a unit finished. The rework I paid for later erased the output I thought I had bought. My intention was fine. My structure was the problem, and the structure is what showed up in the numbers.

This book argues that outcomes follow incentives, not sincerity. Robert Merton named the modern study of this in 1936, calling its sharpest form the perverse result, an outcome running directly contrary to what was intended. Charles Goodhart gave it its most quoted line in 1975, later sharpened by Marilyn Strathern in 1997: once a measure becomes a target, it stops being a good measure. Donald Campbell said the same of social indicators, that the more weight you put on a number, the harder people push to corrupt it. The English NHS lived it out with an eighteen-week waiting target that rewarded hospitals for treating the patient who had not yet waited too long and leaving the longest waiters last, a target, in one official's words, that ended up incentivising precisely the wrong thing.

The old cobra and rat bounties make the same point, though I will admit a 2025 review found the famous Delhi cobra story historically thin. The lesson outlives the anecdote. A reward paid for a proxy buys the proxy, dead snakes and opened accounts and hit deadlines, never fewer snakes or served customers or treated patients. The discipline that prevents it is dull work. Before a rule ships, write down the laziest behaviour that still collects the reward, because that behaviour is what you have actually bought.

This is a short book about unintended consequences and perverse incentives, Goodhart's law and Campbell's law, the cobra effect and gaming a metric, incentive design as the real work of reform, and the pre-mortem that stress-tests a fix against the behaviour its own structure will pay for.

It is written from an Indonesian operator's context, though good intentions wired to bad incentives produce the same backfire in any other emerging-market or small-business setting.