What is unintended consequences?

Unintended consequences are outcomes that nobody planned. They show up when a change in one part of a system sets off reactions elsewhere. They can be good, bad, or neutral, but the problem cases are those that cancel or reverse the intended effect.

How it works

Systems respond because people and processes adapt, feedback loops activate, and delays hide effects until later. A policy that rewards one measure leads people to improve the measure instead of the goal. A rule that restricts one channel moves activity to another. Models help because they make the loops visible, and they let you test an action across many scenarios before applying it. The habit to build is to ask who will respond, how, and after how long.

Not all surprises are harmful. Some side effects are helpful, and a model can reveal these too. The aim is not to avoid every surprise but to have looked for the likely ones before committing to a change.

Example

Say a city adds a parking fee downtown to cut traffic. Downtown traffic falls by 10 percent. But some shoppers move to a suburban mall, and downtown shops lose 8 percent of sales. Shops reduce staff, the area feels emptier, and fewer visitors come, so traffic drops further than planned but so does the city's tax revenue. The fee reached its target while hurting the economy it was meant to protect. This is an illustration with round numbers.

How to model it in Machinations

Draw the intended effect first, then add the reactions as extra Pools, Gates, and state connections: who changes behavior and how much. Use Levers for the policy settings. Play the diagram across several Lever settings and chart both the target measure and the side measures.