Why bad science persists | The Economist