The Bank of England should get on with the job of setting interest rates and other monetary activities, and be held accountable for how well it performs. Instead, everyone wants to retain control over decisions while shedding responsibility for outcomes.
Public sector management is in deep confusion. Reform is in progress throughout the English speaking world – countries that speak French, or German have not seen much need for this process. The Americans, predictably degenerating into business clichés, have been reinventing their government. New Zealanders seem to have handed over most state functions to economists, private companies and, increasingly, to lawyers to argue about in the Courts.
Britain pioneered privatisation although, with appropriate civil service caution, change at the centre of government is limited, like the first efforts of a toddler to Next Steps.
The best text on this New Public Management is provided in Christopher Foster and Francis Plowdens “The State Under Stress” (Open University Press, 1996). Foster and Plowden are somewhat reluctant to reach firm conclusions. But some important ones emerge.
Almost every success of public sector management reform results from increases in decentralisation, pluralism and contestability – handing down responsibility, allowing new ways of doing things, opening up activities that were previously public sector monopolies. Privatisation has been successful – as in the transformation of British Airways or the extraordinary efficiency gains which have been made in power generation – when directly associated with the introduction of competition. Competitive tendering for services has produced extensive cost savings. The Citizen’s Charter works when it substitutes individual employee responsibility for the implementation of a bureaucratic rule book.
And most of the failures have to do with the substitution of formal contractual relationships for previous structures of informal co-operation. The breakdowns of morale in health and education. The growing complexity of regulatory rule books. The problems of the Child Support Agency and in prison service management. Rail privatisation illustrates the dichotomy well.
The gains are all from giving local managers more power – to try new services or fare structures. And from giving freedom to divide priorities. The nonsenses are all to do with the perceived need to write elaborate contracts between newly separated organisations.
The trick that’s required is to try to achieve decentralisation and pluralism without excessive contractualisation – and rail privatisation will only work if the former develops and the latter atrophies. It’s not easy, since the introduction of markets almost inevitably implies both. That’s why it’s all too common to confuse the two things. And why the introduction of markets is not always the best direction of reform.
Contractualisation is associated with a concept of separation. The old public sector rule was that there was a separation between policy and administration. Ministers decided what to do and their civil servants did it. Today we talk of other kinds of separation – between policy and operations, purchaser and provider, customer and contractor. Separation between policy and administration could never work very well because in any but the simplest of cases it was impossible to make sensible decisions about what to do without being involved in doing it and difficult to do it well without some knowledge of and sympathy for the reasons it needs to be done.
And the new forms of separation mostly don’t work for similar reasons. A split between purchaser and provider can only be satisfactory if the purchaser can decide what needs to be bought without the kind of detailed knowledge needed to provide it. This is fine if what you’re buying is baked beans, or possibly even issuing driving licences. It doesn’t work so well when what is being bought is health care, monetary policy or criminal justice.
And that leads to a distinction which much of the new public sector management reform fails to understand – the difference between decentralising responsibility for functions and decentralising responsibility for outcomes. You can’t successfully devolve responsibility for outcomes in situations that involve any substantial element of interdependence, and you cannot successfully devolve anything at all unless you devolve responsibility for functions. When you assign responsibility within a football team, you don’t make one player responsible for shooting the first goal and another for stopping the other side from scoring. You give people roles, like goalkeeper or centre forward, and assess their contributions in that role to the overall objective of the team.
Similarly, central bank autonomy does not mean – as it seems to mean in Britain today – making the Bank of England independently responsible for the outcome (the inflation rate) without giving it independent responsibility for the function (the operation of monetary policy). This leads to the pointless charade in which the Bank of England always says that inflation might be about to go up and therefore interest rates need to rise. It said it again this week. The Bank finds itself in exactly the same position as the footballer who is told he is responsible for the result but is not allowed on the pitch. All he can do is shout encouragement to score goals and that is what the Bank of England does. What independence and decentralisation really means is telling the Bank of get on with the job of setting interest rates and other monetary activities, and holding it accountable for how well it does the job. And not just in relation to the inflation rate, which is only partly under its control, but on how well it contributes to avoiding booms and busts in the economy.
The reason this doesn’t happen much is that what ministers, or for that matter chief executives or anyone else with authority, want to do is just the wrong way round. They want to shed responsibility for outcomes while retaining authority over what is done. This will go down in history as the “Michael Howard problem”. When John Major said that the creation of executive agencies did not affect the principle of ministerial responsibility, he was either talking nonsense or saying that what was being done was a meaningless charade. New public sector management still has many steps to take.
