David Rose of Essex University, who helped devise the scheme, says that it turns out to be a good predictor both of income (even though it is not based on income) and health (except intriguingly for the self-employed in class four, who are just as healthy as those in classes one and two).