Paul Allin was director of the national wellbeing programme at ONS when the project was first launched. Once the programme was established, Paul retired from a career in official statistics and joined David Hand as a visiting professor in the department of mathematics at Imperial College, London.
David Hand is the senior research investigator and professor of mathematics at Imperial College and a former President of the RSS. He has also advised various public sector bodies down the years and currently sits on the ONS National Wellbeing Technical Advisory Group.
Both David and Paul are well placed to comment on where the idea of measuring national well-being now stands and how it can be used in future to improve society. We asked them about the questions their book sets out to answer.
The measurement of well-being has now gained traction at the upper levels of many governments. At this stage of the concept’s development, what did you aim to demonstrate with the book and who did you aim it at?
We hope that we have demonstrated that the measurement of well-being is practicable and is now being taken up by many governments around the world, at national, city and local levels. However, we reckon it is still early days in the use of such measures - well-being measures will only have an impact when policy-makers explicitly adopt them in evaluating the impact and effect of policy decisions. So, we aim the book not only at statisticians and economists in government, but also at policy and decision makers in government and business. We also trust that the discussion on the use of wider measures of well-being and progress will strike a chord with the media, non-governmental organisations and the public.
How do you find the right balance between survey-based data on subjective well-being and data from existing economic indicators of income, consumption and wealth?
The simple answer is that it depends on what matters to the people of the country, city or community for which new measures of well-being and progress are being produced. We realise that this makes comparability between places difficult so, down the line, we anticipate that some standard coverage may also emerge. A good starting point for that would be the range of measures proposed by the Commission on the Measurement of Economic Performance and Social Progress (the Stiglitz, Sen and Fitoussi Commission).
Also, we are not immediately proposing a single, overall index of national well-being, so the question of weighting together the diferent types of measures is not paramount. Again though, we think that will need to be tackled in time (and the Stiglitz Commission tasked national statistical offices to provide the information needed to aggregate across dimensions). This is because headline measures of well-being and progress are likely to be needed to stand alongside the headline measures of economic progress, such as GDP and GNP, before there is widespread use of the wider measures. Incidentally, subjective well-being and economic aspects are only two of the relevant dimensions - issues such as sustainability are also critical to the longer-term well-being of a nation.
What specific areas of policy do you think can be better informed by a well-being measurement?
Our book is not about the policy implications of national well-being measures, but rather the broader question of why and how to go beyond the established way of assessing progress only in terms of economic growth. So, that means we are concerned with broad questions about the kind of society we want and how we see human development and sustainable development. That said, we have pointed in the book to a number of issues that appear to impact on individual well-being and which have public policy implications, such as income and how it is distributed, how we spend our time and the state of the natural environment. Or perhaps another answer to your question is to bounce it back - what areas of policy would not be better informed by an understanding of how decisions in those areas impact the well-being of a nation and its people?
How detailed and comprehensive should a well-being survey design be? Should the survey be conducted at different points of the year and should the questions vary accordingly?
It’s certainly the case that the survey should be conducted at multiple points of the year. The impact of phenomena such as seasonal affective disorder are well known. In the survey work we know best, where the ONS measures subjective well-being, the same set of questions is asked throughout the year. This means that the analysis of the results can focus on the personal characteristics of the respondents at the time, and external factors, rather than variations in the questions. We think this is a good way forward, though of course how it is implemented elsewhere will depend on the resources available to conduct a survey. However, we do think that using the same questions over time is important. As to your more general question, about how detailed and comprehensive a well-being survey should be, the obvious answer is the more the better. But again, do not forget that well-being surveys tap into only some aspects of the well-being of a nation.
What kind of life events (such as having children) affect the measurement of well-being and how can a survey take account of this?
Should the question really be about what sort of life events affect well-being, rather than its measurement? Because it is well-established that life events do have dramatic impact on well-being. In attempting to measure national well-being, we are attempting to produce an aggregate summary - which is exactly what a survey seeks to do. And so that it is as valid and accurate as possible, such a survey should have as broad a coverage as possible. The subjective nature of subjective well-being surveys is both a strength and a limitation. It is a strength because the research into how people answer these type of questions suggests that they are answered spontaneously, without anticipating what the questioner wants. However, the limitation is that we do not know exactly what has caused the respondent to give the answer that they have. This is why the Stiglitz, Sen and Fitoussi Commission included a recommendation that subjective well-being measures should be included along with objective measures of quality of life when building new measures of well-being and progress.
What will new data, technologies and changes in society mean for the future of measuring well-being?
We are sure that ‘big data’, open data and new electronic data capture technologies will all provide new sources and new ways of assessing how a country is doing. We can imagine more use of indirect methods of assessing individual well-being, for example, rather than overt surveys. However, the more urgent need is to draw on and use existing statistics that will paint a broader picture of social progress. It is the need to use statistical measures that we want to keep emphasising, not just the development of new statistics.
How long do you think it will take for the concept of well-being to become as established as other statistical indicators in society? Can you take a guess as to when well-being results might become a major talking point in a UK general election campaign?
Interesting question! We do see a lot of aspiration out there for the ‘elaboration of additional measures that better capture the importance of the pursuit of happiness and well-being’, as a recent United Nations resolution put it. But we also see economic statistics continuing to feature prominently in the news and in politics. This is presumably largely because of inertia since, as anyone who knows anything about how economic measures are constructed knows, they have substantial arbitrariness. For example in the definition of what counts as work, as well as uncertainties in the data at their foundations. Any notion that they are more solid or valid than well-being measures is an illusion. We can but dream that our book might help take public and political debate to a broader assessment of how society is doing and what progress we are making, and that this is based on well-being measures including GDP, if not going ‘beyond GDP’.