A triple test for behavioral economics models and public health policy

We propose a triple test to evaluate the usefulness of behavioral economics models for public health policy. Test 1 is whether the model provides reasonably new insights. Test 2 is on whether these have been properly applied to policy settings. Test 3 is whether they are corroborated by evidence. Where a test is not passed, this may point to directions for needed further research. We exemplify by considering the cases of social interactions models, self-control models and, in relation to health message framing, prospect theory; out of these, only a correctly applied prospect theory fully passes the tests at present.