David Aaronovitch: In the NHS, the law of perverse incentives rules

'Some prefer a past when the professional was seen as an artiste, whose judgement was too fine to be scrutinised'

I love the idea of "perverse incentives". The phrase doesn't refer to a system aimed at rewarding stocking fetishists or creating a new generation of spankers, but to the unlooked-for negative consequences of positive actions. Centralised economies are full of perverse incentives, partially because they are also full of positive actions. The executive committee sets a five-year plan for increased garden gnome production, say, and it is clear that the job prospects of apparatchiks further down the line will be enhanced if the gnome plan is met or even exceeded.

Under the slogan of "Gnomes for All!" gnome producers are set targets, for which they (naturally) request the maximum amount of resources. Since people are imaginative, some ornaments (herons, for instance) are reclassified as gnomes, while production of other things, notably garden furniture, almost ceases. Meanwhile, defective gnomes, with broken hats are passed fit for service and sit, unwanted on the shelves, while customers search in vain for deck-chairs. The plan is fulfilled, but the executive committee targets furniture as its next big priority area.

It happens here too. Frank Field always argued that we had built hundreds of perverse incentives into the social security system. "What would some claimants do, but lie and cheat and be idle?" he asked. "We have created a system that rewards dishonesty." And some interpretations of this week's National Audit Office report suggest that a similar process has been at work, as the NHS has struggled to implement the Government's targets for reducing waiting lists. Figures have been fiddled or massaged, patients have been reclassified and – worst of all – clinical priorities have been distorted.

According to the report, more than 50 per cent of consultants said that they had treated non-urgent cases before serious ones in order to meet waiting-list targets. One fifth said that this happened "frequently". Examples cited were of patients operated on to reverse vasectomies, while patients with bladder tumours were forced to wait. This particular set of perverse incentives has dominated the discussion of the NAO report, as the rhetorical fur has flown.

Let's note this, but with two reservations. The distinction between "serious" and "urgent" is nothing like as clear-cut as the bandying around of the magical phrase "clinical priorities" seems to suggest. Very long waits for "non-urgent" treatment can also cost lives. Patients can have several conditions at once (say, hip and heart) and one may well affect the other. The second reservation is to wonder at grown-up clinicians and managers who, for whatever reason, will consciously invert their priorities. Isn't there an Oath about this sort of thing?

Even so you can see what happens further out when a blunt target is set at the centre. And though the focus has shifted from waiting lists to waiting times, it can only be a year or so before we start seeing the perverse incentives in the new targets. There will be other discrepancies and distortions. Some of these may not matter, others will.

What bothers me just as much, however, is the way that the problems of target-setting will be used by those who dislike measurement. Targets and measurement are not the same thing, but many professionals dislike both. They don't like collecting the information (bureaucratic and time-wasting) and they don't like what is done with the information once collected (misleading and punitive). Some prefer a past in which the professional was seen as an artiste, whose judgement was too fine and intuitive to be scrutinised. I certainly like to be regarded in this way.

The NAO report, however, stresses the importance of measurement. How else could we ask why it is that Dorset Health Authority manages to be in the top five for low outpatient, inpatient and trauma and orthopaedic waiting times? Indeed, some of the differences that the report reveals are truly staggering. Judged by the raw statistics, different parts of the NHS (and not the ones you'd expect) could be on different continents. Whatever Dorset has got, we want a bit of it spread around.

That's why I was alarmed by the announcement, a few days ago, at the Welsh Assembly that school league tables are to be abolished in that country as from next year. The Welsh Education Minister Jane Davidson argued, apparently, that performance tables were "divisive" and placed an "unnecessary burden" on schools. Schools will publish results, but only in their prospectus and governors' annual reports. So only existing parents will see them.

Well, hold on a moment. How is this going to be any less of a burden, if the figures still have to be produced? That argument, at any rate, is just a diversion. The real object is clearly to prevent prospective parents from making school-to-school comparisons. And not just parents. Ms Davidson also said that she wanted to stop local newspapers publishing league tables. This is so nakedly producerist, that the teaching unions have outdone each other in congratulating the minister.

The tables were not targets. No money rides on them. No one dies if you slip a place or two. Above all, it is always open to schools to explain their results, or indeed to learn from the results of others. But if the information is withheld, then there is little chance of everyone learning from success.

What complex public services need is high information and low central targeting. In the wake of the NAO report, the director of the Patients Association, Mike Stone, told BBC News Online: "We're meant to live in a country with a national health service. Whether you live in Newcastle or Truro, you should be receiving the same service." This is almost exactly wrong. It leads to the fixation with one or two national statistics (something that can fit neatly on an election mug or pledge card) and to the incapacity of the health debate to deal with anything more complex than what can be explained to an amoeba.

The NAO puts it much better in a line of one of its press release. "The need for wider implementation of good practice to manage and reduce waiting lists and times in the NHS in England is highlighted today in a National Audit Office report" is how it began. It continued, "wider implementation of innovative practices would make a significant contribution to reducing waiting lists and times." It then identified "five key areas" of good practice.

You cannot have best practice without measurement and without continual innovation. If there isn't measurement and publication, then the public is excluded from the discussion and there is no real accountability. If I were a Welsh parent I'd be pretty alarmed at what was now being done in my name. As for the innovation, that can happen under private or public management, and it is just ideological eyewash to suggest that it can't. But you certainly can't get it if local initiative and autonomy are being rolled over by five-year centrally determined plans.

Open publication plus innovation plus best practice plus accountability equals no more misshapen gnomes.

David.Aaronovitch@btinternet.com