Despite the difficulties, one should be accountable more for outcomes than for the decision process.
Gordon Brown has wisely insisted on conditionality in many of his new public spending plans – the money depends on meeting targets. But it is as yet barely clear whether poor performance leads to more money or to less. There is a lot of hard thinking about accountability for public spending still to be done.
Accountability is necessary because delegation is necessary. Mostly, delegation is necessary because those at the centre do not have enough time, information, or expertise to do things themselves. It is for all of these reasons that Mr Blunkett has to delegate responsibility for Amanda Smith’s GCSE grades. And there is an additional reason – the difficulty politicians have in making truly binding commitments – which prompts Mr Brown to delegate responsibility for setting interest rates to the Monetary Policy Committee.
But accountability can take many forms. You can be made accountable for the process of taking a decision, for the decision itself, or for the ultimate outcome. And you can be held accountable at the time of making the decision, or when the outcome is known. That gives rise to at least six possible mechanisms of accountability. They are tabulated below.
Some of these mechanisms are counter-productive. Accountability that relates to process, and is before the event, for example. Here the key question is whether the decision is being made in the proper way and this matters much more than what the decision is, or what results from it. It is exemplified in the civil service memo which is circulated to a list of recipients as tall as Nelson’s column. By the worried university administrator who asks “but has it been to the procrastination and circumlocution committee?” And in the battles that take place within every bureaucracy, corporate or public, to ensure that when any discussion takes place, our department is there.
In some parts of the public sector, this emphasis on procedure in accountability has a rationale. Because the state is so powerful, the legitimacy of its actions matters. That is why it is necessary to insist on due process. But we should never think that because a decision is properly made it is right
The same error creeps into another debilitating form of accountability – accountability for process after the event. This is what judicial review, which began as a check on arbitrariness in public decision-making, has increasingly become. The best protection is to establish an appropriate paper trail, rather than to make wise judgements. Quality standards such as ISO 7000 have some similar effects.
We do better with mechanisms of accountability that are aimed at decisions themselves and their outcomes. But not much better with accountability for decisions before they are made: the type of accountability that looks over people’s shoulders and undermines, but does not transfer, responsibility for decision and outcome. Accountability for outcome before the decision ought to be impossible. But in fact it is all too common. It means second guessing not only the decision, but its consequences. People who do this have not understood the rationale of delegation.
So we turn to the two mechanisms of accountability at the bottom right – accountability, after the event, for the decision and its outcome. At first sight, it seems better to focus on the outcome. What GCSE grades did Amanda get? What actually happened to inflation?
But, as both these examples illustrate, things are not so easy. There are two groups of problem. First, it is difficult to disentangle the effects of the decision from the environment. How do we know what grades Amanda should have got? And if she was lazy, how much of the responsibility for that is hers and how much should be laid at the door of her teachers? We can try to hold the Monetary Policy Committee responsible for the rate of inflation. But how do we decide what is a reasonable target? And how can we make them responsible when there are other important factors – the government’s fiscal policies, what happens to oil prices – which are critical to inflation but entirely out of their control?
Second, by stressing outcomes we invite emphasis on those things that can easily be measured, and on the selection of easily measured things that we have chosen. Good GCSE grades are important, but no-one thinks they are the only purpose of education. Low inflation is not the sole objective of economic policy and interest rates affect many other things than inflation. A teacher who is told to focus on GCSE scores, or a Monetary Policy committee assessed by reference to the inflation rate, will inevitably run classes excessively oriented to examination success, and set interest rates higher than are appropriate in a world of more complex objectives.
In these situations, performance appraisal must focus not just on the outcomes but on the decisions. It is both not enough, and too much, simply to look at Amanda’s scores and recorded inflation. We have to ask direct questions about the teaching itself, we need to assess whether the interest rate moves which were made were reasonable, given the imperfect information and understanding at the time. But to do so in a way that does not in any way blur responsibility for decisions at the time they were made. Movement away from focus on process, and to accountability after the event rather than before it, is clearly movement in the right direction.