In a piece published here in the FT on February 21st and available at http://www.ft.com/cms/s/0/5a996db2-5c93-11e1-8f1f-00144feabdc0.html#axzz1nCFZ4BWd, but behind a paywall you have to at least register to access) Le Grand and Cooper suggests that the argument over the future of competition is a case of, following Ayers, ‘empiricists’ versus ‘intuitivists’ with the latter becoming ‘crowded out’ and their evidence of competition improving the NHS being ignored.
They go on to say that patient choice was introduced in 2006, and four studies based on hundreds of thousands of patient observations have clearly and consistently shown that ‘public hospitals facing more competition from other public hospitals improved their outcomes and became better managed’ and that ‘the introduction of competition in the NHS could be credited with saving hundreds of lives’.
They write that many of the principles in the NHS bill were sound, but the debate around the bill has empowered provider interests and the intuitivists who have made unsupported claims that have led to dead-end debates, and the ‘robust evidence base’ for competition being likely left on the cutting room floor.
There are a large number of problems with their argument.
First, there is more than one way to do empirical social research. Le Grand and Cooper seem to be saying that the methods they like, predominantly large-scale regression-based models, are the right way. The imply that theirs are only methods which are robust and so the only ones upon which evidence should be based. They seem to be suggesting that they have announced their findings on competition in the NHS, and that they are definitive.
The empirical parts of the findings from the research by Cooper, Propper and others are important – they suggest the NHS has seen changing referral patterns and increases in quality in particular locations. But they do not make any empirical link between these findings and competition, rather they assert rather than demonstrate that NHS hospitals faced competition after 2006, and so their results must be due to that change.
But their work takes no account of empirical research from other disciplines that finds those working in hospitals being largely disinterested in patient choice or competition because they already have waiting lists and see little benefit in competing for more patients. Equally, it makes no attempt to explain exactly how their empirical findings came about.
A more open-minded approach to considering the health reorganisations of the 2000s tends to regard the introduction patient choice and payment by results as one of many changes, so it is near-impossible to attribute the outcomes found in their research to a single factor. It is not enough to simply assert that competition appeared in 2006. Surely an empirically-driven research project would seek to explain not only what happened after that date, but how. We do not dispute the empirical parts of their work, but we do think that the claim that those changes occurred due to competition are at least contentious, and probably mistaken.
For competition to have caused the changes found in the research Le Grand and Cooper favour, even in the most stripped down and unrealistic form, two things must have happened. First, that that the introduction of patient choice plus payment by results led to competition, and second that competition is able to drive up clinical quality. They don’t show either, but rather assume both.
It is hard to see how choice plus payment by results equals competition. A general point is that economic theory itself tell us that market structures with only a providers, as is the case in the NHS, are more likely to be collusive than competitive with little incentive to improve quality. Empirically, it is hard to say how the introduction of patient choice led to hospitals competing. Only around half of the patients surveyed after the introduction of choice even remember being offered it to them and only a tiny proportion made use of the NHS Choices website specifically launched to help them make them in an informed manner. Even where patient choice was both in place and informed – so that patients chose shorter waiting time and better performing hospitals – the hospitals they did not choose faced little in the way of penalty provided they were able to keep their demand levels stable from other patients waiting for treatment. A shortage of demand does not seem to be a problem in the NHS (or why do we spend so much time worrying about waiting lists?), and unless patient choices lead to real financial problems for hospitals not chosen, the causal link is broken.
Can competition drive up clinical quality? To assert this assumes that those working in hospitals have both the motivation and the means to respond to falls in demand especially by improving care. Provided local areas have sufficient demand for a hospital’s services, there seems to be little motivation to react to patients choosing to go elsewhere. Even if hospitals do face financial pressures, it doesn’t seem to be the case there is a threat of closure – a mere two weeks ago the government gave a £1.5bn bailout to 7 hospitals in financial trouble over their PFI contracts.
Different groups of workers in hospitals also respond differently to the threat of losing patients through competition, even where we assume it to be present. Managers, who may be on performance-related contracts, will regard them as a more serious threat, and yet raising clinical standards requires manager to get clinicians to improve their practices. Empirical research over decades suggests that it is pretty difficult for managers to get clinicians to do things differently in the best of times. Simply assuming away this problem as an inevitable consequence of competition seems rather extraordinary – and completely ignores detailed work carried out by researchers who believe there are considerable problems with the underlying research upon which these claims were based, and which we published in the Lancet.
What this leaves us with is the view that Le Grand and Cooper are dismissing those that don’t use their methods, or believe in their particular kind of economics, as ‘intuitivists’. This is rather closed-minded, and a little odd as the data they use wasn’t collected through their own personal observations, and seems to be have been interpreted according to their own theories rather than the more inductive approach which is at the heart of Ian Ayers’ book (which I included myself in my recent book on research methods (http://www.amazon.co.uk/Designing-Social-Research-Guide-Bewildered/dp/1849201900/ref=sr_1_1?ie=UTF8&qid=1330159375&sr=8-1) as making a range of interesting contributions
There is more than one way to do empirical social science. There is more than one way to explain research findings. I’d like to invite Le Grand and Cooper to be rather more open about their assumptions and to regard their findings as the beginning of a discussion about evidence for NHS reorganisation, rather than, as they seem to be believe, the final word.