The use of laboratory experiments in behavioural economics has come under sustained criticism in recent years. Here are some of the criticisms made:

1. Skewed psychological testing using a WEIRD cohort - i.e. many of the people involved in the tests were drawn from Western, Educated, Industrialised, Rich and Democratic Countries and therefore not a representative sample
2. People’s observed behaviour in the laboratory is not realistic - usually small stakes involved in the dilemmas with few real and lasting financial consequences for participants
3. Self-selecting nature of volunteers - often graduate students who are taking psychology
4. Many experiments are rough and ready - they are sensitive to how they are constructed/delivered by psychologists
5. Discoveries about the past from behavioural experiments do not easily generalise to the future - the social context for one generation is often different from another.

Evaluating these criticisms

• Increasingly, behavioural interventions are being tested outside of the laboratory. For example the growing use of randomised control trials in the field. Consider for example the work of Duflo and Banerjee in Poor Economics
• Developments in neuroscience are giving us a more complex understanding of brain behaviour and how we behave/choose in different situations.
• Behavioural economics is evolving quickly and is now less reliant on crude, simplistic experiments in labs
• Behavioural interventions (nudges) are probably best seen as a complement to rather than a substitute for standard policies when dealing with market failure e.g. most economists would favour the use of a carbon tax and/or stronger regulations in helping to cut carbon emissions in the long term
• Attention is also switching to market design / engineering as a way of addressing fundamental / chronic market failures. Consider for example the ground-breaking work of Nobel winner Al Roth in developing matching algorithms for kidney exchange.