Jennifer Castle, David Hendry, 13 August 2014

A typical Oxford University econometrics exam question might take the form: “Data mining is bad, so mining with more candidate variables than observations must be pernicious. Discuss.” Similar questions may well be asked at other academic institutions, but there may be few outside Oxford University where the candidates are expected to refute both myths. This column explains why that is the right answer.


CEPR Policy Research