Economic theory predicts that, ceteris paribus, increasing patient choice will increase productivity. Choice enables a better use of existing capacity. Better capacity utilisation leads to higher productivity. Further, if choice is coupled with competition, theory will again suggest that there will be positive incentive effects, driving efficiency and productivity improvements. Both of these factors would lead to the prediction that productivity will increase as choice and competition are rolled out. And as Andrew Street and Padraic Ward’s work has shown, productivity has indeed increased in the English National Health Service (NHS), following the widespread introduction of patient choice and provider competition.

Now, of course, post hoc does not necessarily mean propter hoc, and, moreover, ceteris was not completely paribus. There were other things going on during this period, including extra resources and the continuation of the targets and performance management regime. However, I find it hard to think of other comparable developments in that particular period that we would predict would lead to increases in productivity (as distinct from increases in output or production). Two things have to be explained: first, why there was an increase in productivity (not just production); and second, why the productivity increase started at the time that it did. Although I am certain that resources, targets and performance management led to increases in production, I cannot think of a strong theoretical or empirical reason for arguing that they led to increases in productivity – indeed, rather the reverse, given the inefficiencies involved in meeting many of the targets. But, even if they did increase productivity, then why only after 2004–2005, when resources had been increasing and targets had been in place since around 2000? Payment-by-results (PBR; which, along with choice and competition, was being rolled out more widely after 2004), and the deficit problem may have played a role; but one of these (PBR) is an essential part of the quasi-market competitive ‘package’ and the other was arguably a consequence of the package’s implementation.

There are two additional points. First, Street and Ward mention three possible drivers of their results: (i) increases in the number of patients treated; (ii) improvements in the quality of care that patients received; and (iii) a slowdown in staff recruitment and use of agency staff. It is worth noting that all of these are what competition theory would predict in a world of fixed prices, such as the English NHS, with agents looking to compete on volume, quality and cost-cutting. Second, it is worth noting that an (much less sophisticated) analysis of productivity changes during the 1990s internal market that I did found the same result as that of Street and Ward (Le Grand, 2002).

Overall, it would be difficult to claim, on this evidence at least, that the predictions from choice and competition theory were incorrect and had no effect (even if they were not the whole story). If anyone does want to claim that, then it seems to me that the ball is in their court; it is now their turn to produce evidence to prove their assertion.