LSE - Small Logo
LSE - Small Logo

Blog Admin

May 1st, 2012

Ten Commandments of good policy making: a retrospective by Sir Gus O’Donnell

5 comments | 2 shares

Estimated reading time: 5 minutes

Blog Admin

May 1st, 2012

Ten Commandments of good policy making: a retrospective by Sir Gus O’Donnell

5 comments | 2 shares

Estimated reading time: 5 minutes

Sir Gus O’Donnell recently delivered a lecture at the LSE on his time in the civil service. What follows is his overview of the triumphs and disasters, as well as his ten commandments of good policy making.  

In this article, I will review some of the lessons I have, sometimes painfully, absorbed over my time in the Civil Service. I will start with policy triumphs and failures and then consider the implications for how the public sector is run and how it needs to evolve. I find it useful to remember the motto that Wimbledon tennis finalists see at centre court

‘If you can meet with triumph and disaster and treat those two impostors just the same’

There is as much to be learned from success as from failure. Accountability, if it is to work well, needs to encourage well managed risk taking not to promote risk aversion. So let’s start with some triumphs, first in economic policy as this is the area I know best.

Triumph no 1: the Labour government’s decision not to join the Euro  

It is easy to look back at this decision and say it was obvious. But many senior members of the government, including prime minister Tony Blair and leading commentators, were strongly in favour of the UK joining from the start or as soon as possible. The decision not to join was a triumph of politics in that the politicians made the right decision and what I believe was a triumph for economic analysis.

The test carried out by HM Treasury was a comprehensive cost-benefit analysis of the pros and cons of joining. It has stood the test of time. If only it had been translated into Greek. And having given credit to the last Labour government for deciding not to join, I should also add that they would not have had that option if it had not been for John Major’s success in getting the UK an opt-out from monetary union.

The lesson I draw from this is the need to have an objective, thorough analysis of key economic decisions. Indeed I would go further and suggest that future decisions of similar importance should have their analysis crosschecked with the Office of Budget Responsibility (OBR). This could expand to cover major policy issues like the High Speed Rail Link. Where is the logic of having the OBR looking into the details of every budget decision, some quite trivial, and not looking at far more important policy issues where departments other than HM Treasury are in the lead?

Second triumph: the establishment of an independent monetary policy committee to set interest rates 

This was an interesting and rare example of politicians ceding power to technocrats. It has succeeded in de-politicising this subject and has enhanced accountability. The greater openness, with voting made public, regular press conferences and scrutiny from the Treasury Select Committee, has worked well. Of course, monetary policy has not been perfect. It never will be whatever the institutional arrangements, but I believe the new arrangements have produced better results than would have been the case under the former arrangements.

I stress that this is a belief and it is certainly an assertion that could be challenged. However the accountability gain is clearly very substantial. I should add that the coalition government’s decision to set up the OBR is another triumph that I expect in time will have just as positive an impact. But we should be careful not to measure the OBR’s success by its forecasting record. Its main achievement is to de-politicise the forecast and to ensure objective costings of policy decisions. I would hope that the OBR might reduce the bias in certain forecasts but I doubt that it will reduce the standard errors very much.

My third triumph: microeconomic examples 

There are so many that spring to mind it is difficult to choose one: the use of differential taxes to persuade everyone to convert to unleaded petrol; the contrast between lung cancer rates in France and the UK resulting from successive decisions by Chancellors to increase taxes on cigarettes; the abolition of exchange controls; the privatisation of companies like Cable and Wireless, BT, British Airways that led to massive increases in productivity; and, more recently, having started to exploit advances in behavioural science that are incredibly promising.

Some of these triumphs represent the application of well accepted economic principles. Others, like removing exchange restrictions and privatisation were hotly contested at the time. The lesson I draw from both examples is that there is an important place in applying standard economic solutions to many public policy issues, but there is also a case for well managed, but risky, initiatives. As Nigel Lawson made clear in ‘The View From No 11: ‘if he had waited for a consensus to emerge ‘we would still have exchange control today’. Sometimes you have to try something new because, unfortunately in the social sciences, we rarely have evidence from controlled experiments. Ideally we would use controlled experiments far more often. I am delighted that a lot of the work of the so called ‘nudge unit’ is being taken forward using carefully constructed experiments in different parts of the country.

Now the disasters

The first I would choose is the poll tax, or community charge to give it its official name. Economists have long argued that poll taxes are the least distortionary of all taxes. However, as good economists know, they went on to argue that it may well be necessary to offset any adverse distributional impact by other means. The poll tax would have resulted in ‘a pensioner couple in Inner London paying 22 per cent of their net income in poll tax whereas a better off couple in the suburbs would pay 1 per cent’.  A report by Department of Energy officials described the poll tax as ‘completely unworkable and politically catastrophic’. The Treasury were firmly opposed to its introduction. The poll tax did indeed prove catastrophic for the then Prime Minister Margaret Thatcher.

So what lessons should we draw from this episode? The first I call the Irish axiom: don’t start from the wrong point. Successive governments had allowed the financing of local government to become more and more centralised. This destroyed local accountability. The problem is still with us and I would be surprised if it is solved at all soon. The problem is that radical changes are now needed and that means large numbers of winners and losers. And the iron law of tax changes is that the losers scream and the winners remain silent. The second lesson is the need to have the Treasury on your side if you are making big changes. Money will be needed to compensate the screamers. If the Treasury aren’t supportive the money will not be forthcoming, the policy will fail, and the Treasury will be the first with the ‘I told you so’ response.

I have heard some argue that the poll tax was the triumph of economics over politics. I think you have to take a very narrow view of economics to make that case, but it certainly reinforces the point that, where distributional issues are important, it is vital to have clear political decisions which have cabinet support.

The second disaster must be our failure to spot the financial crisis of 2008 and the associated failures in the regulatory system. It is clear the regulatory system created incentives for excessive risk-taking in part because it was believed that the government would bail out the banks if the crisis was large enough. I take my share of the blame for this episode; I thought the banks knew what they were doing. That was wrong.

It took remarkable courage and persistence from the then prime minister Gordon Brown to bring the G20 together in London and help to prevent a retreat into protectionism and an even worse recession. And, as Alastair Darling explained in an earlier lecture at the LSE, he and Mervyn King had to work extremely hard to save the UK banks from complete collapse.

Some lessons are clear: we need to revise the regulatory system, ideally in a globally coordinated way; we need to have ways of handling banks deemed ‘too big to fail’. There is a lot to be learned from countries like Canada that weathered the storm remarkably well.

For my third failure I will use a microeconomic example; namely, the failure to agree on a planning system that achieves our desired goals. Every time there is a recession, or even of a reduction in growth below trend, there is a call for more ‘structural reforms’. Top of the list is always the planning system. It is blamed for holding back growth and development. The problem is, in fact, a classic example of not being clear about the outcome that is desired. If it is to boost GDP, then the answer is simple: concrete over the South East. But of course that’s not what we want and that’s because you would have to be an idiot to want to maximise GDP. It’s a highly flawed measure and I am pleased that we are at last starting to think more broadly about how as a society we measure success. I want to pay tribute to LSE’s Richard Layard for his tireless work on happiness and on the need to tackle mental health issues. The good news is that we are starting to apply the best economics to these issues. I am hopeful that the recent advances in methods of valuing environmental goods and bads will help decision makers make better choices once they are clear about what they want to achieve.

Good Policymaking

As I look back on all these events over my 30 years in the Civil Service some patterns emerge of what brings success. My ten commandments of good policy making are listed below:

1)    Thou shalt be clear about the outcomes that you want to achieve

Lack of strategic clarity, of knowing the problem you are trying to solve, is a cardinal sin.

2)    Thou shalt evaluate policy as objectively as possible

You need to be clear about how you are going to determine success. It is essential to relate the success measure to the desired outcome. This may sound obvious but far too often it is not done.

3)    Thou shall not bear false witness against thy neighbour’s policies

Most important policies operate across departmental boundaries so be sure to have collective buy in, and ownership of, the policy. We can’t tackle the problems of an ageing society or obesity or climate change simply by operating in one department.

4)    Thou shall not assume the government has to solve every problem

In a future lecture I will talk about the 3 kinds of problems that we face.  We have models in our heads, and sometimes written down, about how we think people, companies and governments make decisions. These models are probably embarrassingly naïve. As Keynes said many of us are slaves to some defunct economist. These three types of model failures must all be addressed.

5)    Thou shalt not rush to legislate

It is a fact that all our ministers sit in either the House of Commons or the Lords which are bodies that spend a very large amount of time legislating. My experience suggests legislation should be something of a last resort to be used when all the other possibilities have been tried and found wanting.

6)   Honour the evidence and use it to make decisions

There are now lots of excellent think-tanks. We also have the internet and masses of academic research. If the policy submission doesn’t cover, for example, what works in other countries then send it back. (There is a lot to be said for covetting thy neighbour’s policies). The only proviso is to be clear that most of the time the evidence will not be clear cut. So perhaps the right answer is to generate more evidence by the use of experiments. In every case judgement is needed. That is the most valuable skill needed by ministers and their senior officials.

7)    Thou shalt be clear who is accountable for what and line up the powers and the accountabilities

If ministers want someone else to be directly accountable to parliament for a specific area then they need to cede power over that area to the ‘someone else’, having obeyed commandments (1) and (2).

8)    Thou shalt not kill the messenger

Nobody’s perfect and every organisation needs a way of providing constructive feedback to the senior decision makers. If you don’t encourage internal debate then the first time you will learn about your mistakes will be from your enemies not your friends.

9)    Thou shalt not forget that it is a privilege to serve

In government, the taxpayers pay our salaries and they deserve good value for money and to know we will always strive to follow the Codes that encapsulate our values.

10)  Thou shalt keep a sense of proportion

Or in the words of a wise, now former, minister: ‘Thank God it’s only a  game’.  For some, every bad headline is a crisis. Keep a focus on what the real impact is on people’s lives.

Gus O’Donnell was speaking at the LSE as part of the British Government @ LSE events series. More information is available here.

Please read our comments policy before posting.

Note:  This article gives the views of the author, and not the position of the British Politics and Policy blog, nor of the London School of Economics. 

About the author

Sir Gus O’Donnell is a former British senior civil servant who, between 2005 and 2011, was the Cabinet Secretary (the highest rank in the British Civil Service), Head of the Home Civil Service and Permanent Secretary at the Cabinet Office. O’Donnell was the last man to hold these posts jointly before they were split.

Print Friendly, PDF & Email

About the author

Blog Admin

Posted In: Central government and core executive functions


Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported
This work by British Politics and Policy at LSE is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported.