The empirical dimensions of the scale effects problem

Empirical work has refuted the prediction of new or endogenous growth theory that growth exhibits a scale effect: the rate of growth has not accelerated and larger countries do not grow faster than smaller countries do. The theoretical dimensions of this ‘scale effects problem’ have been explored at length, but this has not yet led to a satisfactory solution. Concerning its empirical dimensions, it has been suggested that problems of measurement, in particular the ‘quality-improvement problem’, causing price increases to be overstated and real growth to be understated can account for the absence of the predicted growth scale effects. We offer the first test of this possible explanation and find that these biases do not offer a solution to the scale effects problem. The upward bias in measured prices has not worsened and our empirical analysis dismisses the increasing importance of ‘unmeasurable’ sectors as a possible explanation. It shows that the bias due to problems of measurement needs to be implausibly high for growth to exhibit a scale effect. Since neither theoretical nor empirical work seems able to come up with a satisfactory solution to the problem of scale effects, it might be much more fundamental than often realised