On 15 June 2016, the paper 'New statistics for old? Measuring the wellbeing of the UK’ was presented by Paul Allin, visiting professor in statistics at Imperial College London and David J Hand, emeritus professor of mathematics, also at Imperial College London. The meeting was chaired by Michael Baxter, for the Official Statistics section, which hosted the event.
The authors summarised four key messages from the paper:
- The paper is not just about the measurement of wellbeing, but also about how and why it is used – it does not automatically translate into policy.
- The history of wellbeing measurement needs to be considered – the ideas covered are not new.
- It is important to acknowledge that the measurement of wellbeing is a 'work in progress' – there are challenges still faced in this area.
- The Office for National Statistics (ONS) is an exemplar in wellbeing measurement and to a large extent it is the work of the ONS in this area that enabled the paper to be written.
The question posed is simple – 'How is the country doing?' – not just economically, but individually; however, the answer is not as straightforward as the question. Some of the statistical challenges of measuring wellbeing were outlined, including challenges in international comparison of wellbeing measures, definitions of wellbeing concepts and producing an optimal measure, or set of measures, that will be representative, as well as easy to use and simple to interpret.
The authors encouraged the ONS to raise the profile of wellbeing statistics and to promote wider use of the full set of available measures, emphasising to users the importance of the full picture, as opposed to resorting to a single measure of economic progress such as gross domestic product (GDP). The importance of wellbeing measures as a starting point for a new performance criterion was highlighted, and the open question of whether there is a need for a new single measure of wellbeing to effectively compete with GDP was posed, forming the main basis for the ensuing discussion at the Ordinary Meeting.
A vote of thanks was proposed by Paul Smith (University of Southampton). He outlined some of the different possible approaches being used to develop wellbeing measures and highlighted the potential influence of a new single measure of wellbeing on an unbounded scale, similar to GDP, as a driver for behavioural change. However, would this be like throwing away the 'old genie' of GDP for a new measure that is less well-understood? And what are the “three wishes” for this potential new single measure?
The vote of thanks was seconded by Abbie Self (wellbeing director at ONS), who highlighted that the question is about national, as well as personal wellbeing. For the moment the “wellbeing” focus appears to be more on measurement than take-up, but the seventeen global goals for Sustainable Development (on health, education, environment and economy) provide leverage for further take-up across Government departments, to guide their policy progress. While a single index of wellbeing would be easier to communicate with impact, the main challenge with this is to identify which measures are the most important and which can be excluded. In the national debate on the issue, it appears that in fact there is a desire for more information on wellbeing, not less – each of the seventeen global goals carries equal weight and their number cannot be reduced further. In the discussion of how to measure wellbeing it is also important not to lose sight of the ultimate goal to improve wellbeing.
The vote of thanks was passed by acclamation. The chair then invited contributions from those who had indicated in advance that they wished to speak, and further contributions from the floor.
Summary of contributions from other speakers
Juliet Michaelson, associate director, New Economics Foundation
Juliet commented on three key aspects of the paper: methodological challenges, national wellbeing as a concept and a lack of progress in wellbeing policy. She drew upon other social statistics to highlight that wellbeing does change nationally, also highlighting the work of the New Economics Foundation in this area, which explored five key factors of wellbeing. She highlighted the need for increased accountability of policy-makers, driven by the media and the public.
Alan Spence, chief statistician and chief analyst, Health & Safety Executive
Alan focused on the 'competition' between wellbeing measures and GDP, bringing out the deficiencies of GDP as a measure, which are mentioned in the paper. It is preeminent in the field of economics, but it is a relatively recent measure and is an abstract concept, which could be said to 'measure everything except what matters'. There is a need for a measure of wellbeing, to complement the continuing use of GDP.
Jonathan Haskel, Imperial College London
Jonathan highlighted the increasing survey fatigue of the public, which may result in a loss of respondents to wellbeing questionnaires. He also suggested that there may be a degree if organisational fatigue and proposed that expertise from the private sector may help government to improve the current work on wellbeing measurement.
Hazel Wardrop, Equality & Human Rights Commission
Hazel focused on the effect inequality may have on wellbeing, suggesting that more factors to account for inequality should be considered. She also mentioned the challenges in making international comparisons of wellbeing measures, for which a common framework across nations ought to be balanced by differences in the issues that matter most to different nations and cultures.
Fionn Murtagh (with slides), Royal Holloway, University of London
Fionn focused on the methodological challenges of collecting data on wellbeing, highlighting that data based on self-selection and social networks is inherently biased, which suggests that 'big data' would not provide a robust basis for measuring wellbeing (citing the paper 'Geometry & topology of data and analytics'). However, 'big data' would be a valuable resource for calibration and evaluation of more robust survey-based measurement.
Guy Nason, University of Bristol
Guy also spoke about the methodological challenges of measuring wellbeing, including the definition of the population (which can become outdated very quickly) and the measurement of uncertainty (suggesting that the underlying variability of the ONS figures should be much higher). The many wellbeing statistics tell us many things, but currently there is no clear overall picture, while GDP, despite its shortcomings, is a much simpler thing to define and measure.
Woody Caan, fellow of the Royal Society for Public Health
Woody discussed the importance of a person’s life course in their overall wellbeing, drawing on evidence from the field of psychology to illustrate that adverse childhood experiences can help to predict (and prevent) negative outcomes.
Harvey Goldstein, University of Bristol
Harvey spoke about the contribution of the Organisation for Economic Co-operation and Development (OECD) to the measurement of wellbeing in the UK and linked this to the previous discussion on challenges in international comparisons. Wellbeing is likely to be differently understood in different countries, and currently the focus seems to be on highly industrialised Anglophone countries. He suggested that fundamental questions of meaning such as this are in fact more important than the detail of how measurement should be carried out. He also asked for the authors’ views on ways to guard against certain wellbeing statistics being 'cherry-picked' by policy-makers and also on the scope for making causal statements on the basis of wellbeing statistics.
Tom King, University of Newcastle
Tom discussed the importance of wellbeing measures for children; GDP does not cover them, and neither does the annual population survey. He suggested that wellbeing should not be an age-invariant measure, as people are likely to consider their wellbeing over a long time range. Small children live in a small world, with less sense of coherence than adults, but with horizons that expand over time. The development of children is a complex and difficult aspect to incorporate into the measurement of wellbeing (not least due to the practicalities of data collection), but is worth being considered.
Written contributions to the discussion had also been made, but there was insufficient time for these to be read out.
The authors thanked the discussants for the wide variety of interesting topics and valuable points raised which they would need to consider in detail. Written contributions and fuller considered responses from the authors will follow in the journal.
The event was filmed and can be viewed on our video page.