In pursuit of wellbeing

Government decisions based on quality of life and behavioural insights would provide the efficiency savings we badly need

By Gus O’Donnell

Follow Gus on Twitter @Gus_ODonnell


When David Cameron became prime minister in 2010 he announced two potentially revolutionary changes. First, he defined government success as raising the quality of life, or wellbeing, of the people. Second, he wanted government to engage with people in ways that helped them change their behaviour in order to raise their own wellbeing. I always saw these two initiatives as intimately linked. The role of the Behavioural Insights Team, inevitably dubbed the ‘Nudge Unit’, was to develop ways of raising wellbeing based on a better understanding of how real people, with all their faults and foibles, made decisions. Both changes were necessary, particularly as he inherited a deficit of over 10% of GDP and needed efficient ways of maintaining public services while cutting spending.

But despite the efforts of David Halpern, the head of the unit, this connection between the two never permeated government as a whole. Nudge policies had their triumphs: they got people back to work faster, encouraged them to pay tax on time, and improved student grades. But departments were hesitant to apply these tools — or any other tools, for that matter — to boosting wellbeing per se.

Yet the need for these linked revolutions is as great as ever. It is clear that the UK’s public finances will not be in surplus any time soon. We will still be in deficit by the time of the next election, now pencilled in for 2022.

This is not, to be fair, for lack of trying. But the struggle to repair our fiscal fortunes has been extended by disappointing economic growth, which forecasts suggest is likely to slow still further. So we face the continuing, worsening challenge of meeting ever-rising demand out of much smaller increases in tax revenues. How can we create a benign square out of this vicious circle?

One option, of course, is to increase taxation. But taxable capacity is limited, with a historically high proportion of total revenue coming from the highest paid already. Another option is to increase borrowing. But while there are good arguments for increasing infrastructure investment to support growth, letting go of the deficit to finance current public services risks descent into a financing crisis at a time when the UK can least afford one.

A third option is to increase productivity in public services. But some ambitious efficiency targets are already built into budgets, on a scale that (in the health service, for example) exceed anything that has been achieved historically, coming as they do on top of the savings that have already been made.

Efficiency is measured by relating inputs to outcomes. Inputs are relatively easy to count: they are the financial costs of public programmes. But measuring outcomes is much harder. How do we best assess whether a public service is ‘better’ or ‘worse’?

This is where the wellbeing lens makes an enormous difference. It could provide a much clearer, more coherent view of the trade-offs that have to be made in allotting taxpayers’ money to public programmes. Let’s take a few examples. Start with the biggest spending pressure cooker: health. To maximise the impact of expenditure on wellbeing, the budget should be rebalanced to give more to mental health services and less to building (and keeping open) general hospitals. In education, it would mean measuring the wellbeing of our children with the same rigour that we assess academic achievement. For older people, it would mean giving priority to programmes designed to keep them out of hospital. And it should place a barrier before the kind of practice rightly condemned by the most senior judge in the Family Division of the High Court, of driving elderly couples out of their own homes on ‘health and safety’ grounds, only to separate them when they are placed in ‘care’.

These few examples illustrate the difficulties as well as the opportunities, since a shortage of hospital beds, a decline in academic attainment or an uptick in accidents in the homes of the elderly would inevitably cause an outcry. But it would throw the spotlight on some particularly poorly targeted areas of spending, such as the pensioners’ winter fuel allowance, which costs over £2bn each year.

Reality check

A focus on wellbeing should lead to better outcomes, but only if we can also improve the way programmes are delivered. In the past, we have tended to ignore the fact that policies are made for humans, and should allow for their humanity.

Real people make mistakes, regret earlier decisions and at times need help, in all sorts of different ways. This is where policymakers’ growing understanding of behavioural economics should kick in.

Its basic message is very simple: focus on what people actually do, rather than theorise about how you might expect them to behave. This means listening to them, but remembering that actions speak louder than words. Do juicy tax reliefs persuade everyone to save? No. So enrol them automatically in a pensions programme. If you do not want to be heavy-handed, let them opt out, but make it the default option, and see if that works.

Allowing people to learn from their mistakes is good, for both human freedom and the public purse: it reduces dependency and helps people take better decisions for themselves. But some errors (like failing to save anything until you’re too old to earn) cannot be reversed. Then an early ‘nudge’ is justified, its sharpness depending on the necessity of action. And it has proved successful in spreading the habit of saving for retirement into groups uninspired by tax reliefs alone.

“Policies are made for humans, and should allow for their humanity”

The Nudge Unit has chalked up numerous other successes. Some followed better (behaviourally sensitive) language in official communications. A minor change in the wording of a letter to people who owed tax demonstrated effectively how emphasising social norms could speed up payment. Similarly, the unit found that jobseekers were nearly twice as likely to turn up for a job fair if the text from the job centre used their names, and nearly three times as likely if the person sending the text added ‘good luck’.

The point is not that nudges are small tweaks, delivering incremental change, but that they work from the big understanding that officialdom is dealing with humans, not robots; an understanding the commercial world was quicker to reach than the public sector. This, in turn, explains why behavioural economics and a focus on wellbeing are so closely connected. The second revolution, however, has proved much harder to bring about.

Scepticism runs deep, fed by the inbred Treasury suspicion that such ideas are always an excuse to spend more, never a reason to spend less. And this combined within government with a reluctance to move away from the discipline of a monetary calculus. If, in short, quality of life — or wellbeing — were to be the policy success measure, we would need to find a way of measuring it.

The alarmingly simple solution was to ask people about their own wellbeing, and so the Cabinet Office asked the Office for National Statistics to start collecting subjective data. While we still have not done so for long enough to smooth out cyclical effects, this was a huge step forward.

The measure of it

Subjective data need careful analysis, but without them we are blind, or at best only partially sighted as to the impact of policy on wellbeing. Just looking at life expectancy, for example, leaves us in the dark about the quality of life. None of our ‘objective’ data sources picks up the effect on wellbeing of factors such as loneliness or unemployment.

As the new, subjective sources of information matured, we began to experiment with their application to policy. The chief problem we encountered, and which I underestimated, was not that nothing of the kind was being done already — it was being done quite exhaustively — but that often the wrong measures were being used.

All major investment projects, for example, are (or should be) launched after a cost-benefit study, using the Treasury’s Green Book as the methodological bible. All other policies are supposed to be accompanied by similar ‘impact analyses’. The traditional coinage for these calculations was exactly that: money. Rarely do calculations of the benefits side take full account of the distributional effects, which may change the wellbeing calculus a lot. To take the example of the winter fuel allowance again, an informed wellbeing calculus might encourage us to concentrate much of the £2bn on the poorest pensioners, using the rest to reduce isolation and loneliness among all of the elderly. Indeed, the Conservative manifesto contains a commitment to means test the allowance in some way and use the proceeds to fund social care.

More broadly, the pursuit of wellbeing using behavioural techniques would lead to a massive switch in spending towards prevention and away from cure. Starting young, if we measured and targeted the wellbeing and resilience of our children at schools, our chances of having less remedial work to do with them in later life — through the courts, mental hospitals and social services — would be greatly enhanced. But the Treasury, case-hardened towards spend now, save later pleas, will take some convincing.

A NICE approach

There is, however, one example that should encourage Whitehall’s purse-minders. Perhaps the only wholehearted adoption of a wellbeing approach has sprung from a desire to take the politics out of very difficult decisions. For years NICE has been deciding which drugs can be afforded on the basis of their addition to quality-adjusted life-years (QALYs). Within certain parameters, this practice is widely accepted, and has acted as a brake on spending as well as a consistent way of making tough trade-offs. Using QALYs would also have avoided the debacle of the Cancer Drug Fund, whereby a popular election pledge resulted in £1.27bn in spending on treatments outside the purview of NICE that later proved to be largely unbeneficial.

We should go further with the QALY process, applying it to the politically difficult business of deciding which hospitals should be closed. Perhaps the prime minister may see the opportunity to deploy QALY analysis. It would be worthwhile to study precisely how NICE managed to change the way things are done. And a wellbeing approach would help inform a much better debate on assisted dying.

An even more radical approach would be to take a helicopter look at the impact on wellbeing of everything government does, and how it is organised departmentally. Some landmarks will stand out, not all of them what the sceptics might expect. For example, the subjective data show that security is fundamental to wellbeing, so defence would remain a priority. The connections between health and social care would become even more apparent, as would the nonsense of thinking separately about tax and benefits.

Taking stock today, it is fair to say we have the tools of a revolution, but need the impetus. Data on wellbeing are building up, to the point where they can meet the need for rigorous analysis. Indeed, it has now been incorporated into the Treasury’s Green Book, validating its use in cost-benefit analysis. Its advocates must make it clear that it is not a soft option: departments will still have to make policy choices within an overall spending constraint. But politicians, who instinctively prefer their own calculus of trade-offs to the discipline of such analysis, are the most difficult to convert of all. It would be a great pity if the start made by David Cameron proves to be a wasted legacy.


Gus O’Donnell is non-executive chairman of Frontier Economics. He served as Cabinet Secretary between 2005 and 2011.

This article first appeared in the RSA Journal Issue 1 2017

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.