Anomalies in Economics

Systematic patterns in data or behavior that do not fit a commonly used benchmark model.

An anomaly is a recurring pattern in economic data or decision-making that conflicts with the prediction of a benchmark model.

Why anomalies matter

Economists use models to organize evidence. When a regular pattern keeps showing up but the model says it should not, researchers have to decide whether the problem comes from behavior, frictions, omitted variables, or measurement.

Examples

Anomalies are always relative to a specific benchmark. Common examples include:

  • the Allais paradox against expected utility,
  • present-biased saving behavior against exponential discounting,
  • persistent return patterns against simple market-efficiency benchmarks,
  • slow price adjustment against fully frictionless models.

Some anomalies lead to better theory. Others fade once the data, risk adjustment, or sampling procedure improves.

A caution about evidence

Not every surprising result is a true anomaly. Data mining, publication bias, and weak identification can generate patterns that do not survive replication. That is why economists treat anomalies as prompts for investigation, not instant proof that a whole theory is wrong.

Knowledge Check

### An anomaly in economics is: - [x] a systematic pattern that conflicts with a benchmark model - [ ] any one-time surprise in the data - [ ] a guaranteed market failure - [ ] a legal violation by a firm > **Explanation:** The key idea is repeated conflict with a specific theory or benchmark, not merely something unusual. ### Why are anomalies useful to economists? - [x] They force researchers to improve models, measurement, or identification - [ ] They make theory unnecessary - [ ] They prove every existing model is useless - [ ] They eliminate the need for replication > **Explanation:** Anomalies help show where a model is too narrow or where the evidence needs to be interpreted more carefully. ### A pattern that disappears after better data cleaning is: - [x] a weak candidate for a true anomaly - [ ] stronger proof that the benchmark was wrong - [ ] evidence that theory does not matter - [ ] a reason to ignore data quality > **Explanation:** Some anomalies are artifacts of measurement or sample selection rather than robust economic behavior.