Better Statistics for Better Lives and the OECD Approach to Measuring Subjective Well-being.
Ahead of going to the polls in 1980, US presidential candidate Ronald Reagan invited voters to ask themselves: "Are you better off than you were four years ago?" Reagan's instinct - to appeal to how people are feeling about their lives - is probably one shared by many politicians. But an interest in how people feel has taken some time to filter through to the world of policy evidence and analysis. Nevertheless, times are changing. Thirty-five years later, research from the London School of Economics has suggested that the ups and downs of people's life satisfaction can help to predict European election results over and above standard macroeconomic variables (George Ward, "Is Happiness a Predictor of Election Results?", London School of Economics Centre for Economic Performance, Discussion Paper No. 1343, 2015).
The World Bank has highlighted the falling life satisfaction, even in the face of rising GDP per capita, that predated the Arab Spring (Elena Ianchovichina et al., "Inequality, uprisings, and conflict in the Arab World", World Bank Group Middle East and North Africa (MENA) Economic Monitor, 2015). And the New Economics Foundation, jointly with the United Kingdom's What Works Centre for Wellbeing, has shown that those areas where happiness inequalities were highest in 2016 were also those where the Brexit vote was strongest (Saamah Abdallah et al., "Measuring Wellbeing Inequality in Britain", What Works Centre for Wellbeing, 2017).
Understanding how people are feeling about their lives might, then, be rather important for politicians to grasp. But can we really put a number on people's happiness, and if so, are those numbers a viable tool for public policy making, beyond the politics of winning elections?
Measuring Progress "Beyond GDP"
To assess whether people are better off, economists have traditionally relied on numbers about economic growth, as measured by GDP. But GDP tells us what is happening in the economy at the macro level, not what impact that is having on how people live, the environment they live in, or the resources that will be needed to sustain today's living standards for many generations to come. GDP captures the quantity of economic growth, in bulk, but says nothing about how that growth is distributed and whether society at large is benefiting. Many of the things that people value - ranging from good health, to time spent with friends and family, to clean unpolluted air - are not traded in markets and have no prices, making them difficult to account for. So, what we need is a broader set of measures if we want to understand whether life is getting better - and for whom.
Ultimately, the goal of most governments is not just to grow economies, but to improve lives. Historically, there has been a close association between GDP growth and better living standards, but the two are not synonymous: if the benefits of growth are too highly concentrated, or if that growth comes with high social and environmental costs, for example, the relationship between growth and well-being can be put at risk. Increasingly, there is a recognition that growth is a means to an end, but not an end in itself, and the quality of economic growth matters, not just its quantity. This in turn has raised questions about whether our public institutions, including our national statistical offices, are measuring the right things to give policymakers the feedback they really need.
Better Statistics for Better Lives
Through its Better Life Initiative, launched in 2011, the OECD has been building the evidence base to help policymakers understand whether people are becoming better off - not just in terms of money, but also in terms of their health, happiness, safety, housing, and much besides. We have developed a framework for measuring well-being (Chart 1) that includes both the objective and subjective outcomes that matter for people's lives today ("current well-being"), how these outcomes are distributed across the population, as well as the stocks of resources that will help to sustain well-being over time ("resources for future well-being"). As part of understanding people's current well-being, it is important to address people's own experiences and reflections on how life is going - which in the framework is shown as the dimension called "subjective well-being", but can also be more loosely described as people's happiness.
Subjective Well-being: a Critical Gap in the Evidence Base
There are a number of important gaps in our ability to measure the dimensions of well-being shown in Chart 1, especially on an internationally comparable basis. A key pillar of the Better Life Initiative is to support new and improved statistics, to help fill these gaps. One of the first places we started with was subjective well-being, partly inspired by an influential publication in 2009 by Joseph Stiglitz, Amartya Sen and Jean-Paul Fitoussi, the Report of the Commission on the Measurement of Economic Performance and Social Progress. This report encouraged national statistical offices to add questions on subjective well-being in their standard surveys. At that time, there was considerable interest in the question of how and what to measure, with a variety of methods already in use, but a lack of international consensus on good practice, and a number of sceptics who needed further convincing about the validity and value of these data.
The OECD Guidelines on Measuring Subjective Well-Being, released in 2013, thus aimed to provide support for national statistical offices and other data producers in designing, collecting, publishing and analyzing subjective well-being data.
Making Happiness Measurable
The first step in writing the OECD Guidelines on Measuring Subjective Well-Being was to provide a conceptual definition. Based on the literature, we adopted the working definition of: "good mental states, including all of the various evaluations, positive and negative, that people make of their lives and the affective reactions of people to their experiences" and identified three distinct elements that are core to people's subjective well-being:
- Life evaluation - a reflective assessment on a person's life or some specific aspect of it. This is usually measured through questions about life satisfaction, overall happiness, or instruments such as the Cantril Ladder (which asks people to position their lives on a scale from "worst possible" to "best possible"). It can also include domain-specific evaluations, such as satisfaction with health, or job satisfaction.
- Affect - a person's feelings, emotions and other states. These are typically measured with reference to a particular point in time (such as yesterday) and include both positive affect (e.g. happiness, calmness, enjoyment) and negative affect (e.g. sadness, pain, worry).
- Eudaimonia - a sense of meaning and purpose in life, or good psychological functioning. Measurement of this component has been approached in several different ways, ranging from asking people about the extent to which they feel the things they do in life are worthwhile, through to questions about autonomy, a sense of accomplishment, meaning and purpose, and other aspects of positive mental states or "flourishing".
Methodological Issues: How You Ask Questions Shapes the Answers You Get
An important part of the Guidelines walks readers through the challenges of getting the survey methodology right when measuring subjective well-being. Like almost all data collected through surveys, how questions are asked is key - indeed, the manner and words used affect how communication is received, understood, and acted upon. The Guidelines therefore summarize the available evidence on four topics of central concern to survey design:
- Question wording and response formats;
- Question order and context effects;
- Survey mode and timing;
- Response styles and the international comparability of the data.
We describe how each of these factors is known to affect people's responses to subjective well-being questions, and what can be done to either mitigate error, or (at a minimum) keep the sources of error constant across people, time, and different survey vehicles, so that meaningful comparisons can still be made.
Moving on to the practical concerns of going out and collecting data, the Guidelines also summarize what is known about good practice in terms of survey content and planning. This covers issues of wider survey design - including the key variables that should be collected alongside subjective well-being data in order for them to be meaningfully interpreted - as well as practical advice on survey implementation. Finally, the Guidelines are concerned with the output and analysis of subjective well-being measures. This describes the compilation and reporting of subjective well-being data, as well as some of the factors that affect its interpretation. It also discusses the use of subjective well-being data in a public policy context in particular - for example, as a complement to economic indicators, to better understand the drivers of well-being, and as an input to cost-benefit analysis.
Ready-made Question Modules for Statistical Offices
The Guidelines culminated in a series of prototype question modules, designed for inclusion in household surveys. The goal of these questions is to provide a common baseline to help improve the quality, consistency and international comparability of subjective well-being data. A "core" set of five short questions is intended to provide a minimal set of measures for widespread use - including a primary measure of life satisfaction, for when space constraints permit only a single question to be included (Chart 2). But the proposed modules also offered more in-depth batteries of questions, addressing affect as measured through time-use surveys, as well as scales for capturing domain-specific life evaluations, multiple aspects of eudaimonia, and a richer set of affective experiences.
Taking Stock of Subjective Well-being Measurement in OECD Countries
In the years since the Guidelines were published, subjective well-being measurement has gained momentum in OECD countries. In 2016, we conducted a survey of national statistics offices (NSOs), to gather information about the use of subjective well-being questions in official data collections (Exton, Siegerink and Smith, forthcoming). This found that 34 out of 35 OECD countries now collect life evaluation data and more than three-quarters of OECD NSOs have collected at least some data on affect and eudaimonia, which follow in some respects the recommendations included in the Guidelines. Japan is the only OECD country in which no subjective well-being data have been collected as part of the official statistical system (Chart 3).
Despite considerable recent progress on measurement, the range of different question formulations in use across different statistical offices continues to pose challenges for the international comparability of subjective well-being data. The greatest consensus exists around the use of a 0-10 life satisfaction measure, which is now available for 30 OECD countries. A particularly important breakthrough has been the 2013 EU Statistics on Income and Living Conditions (EU SILC) ad hoc module on subjective well-being, administered in all EU countries, plus Iceland, Norway, Switzerland and Turkey. This featured life satisfaction and eudaimonia questions that are consistent with the Guidelines, as well as affect summed over a period of four weeks. A decision has now been taken to include the central life evaluation question as a core part of the annual EU-SILC.
Meanwhile, data collections consistent with the Guidelines in Canada, New Zealand, Australia, Mexico, and South Korea are helping to gradually complete the OECD picture. This international perspective can be particularly important for benchmarking purposes in individual countries, providing a comparative picture on both average levels and the size of disparities between population groups.
The Guidelines represent a first attempt at international recommendations on measurement, but we are in the early stages of national statistics on subjective well-being, and several outstanding issues need to be better understood. For example, questions remain around the optimal number of items needed to capture people's affective experiences. Many different varieties of measures exist for capturing eudaimonia (or meaning and purpose in life), and more systematic comparison of these different methods is needed. Additionally, a number of OECD NSOs include a wide range of what the Guidelines referred to as "domain-specific evaluations" - i.e. measures designed to capture satisfaction with specific aspects of life, such as health, a person's job, personal safety, or relationships with others. Here again, measurement practice varies substantially, and would benefit from harmonization when it is clearer what good practice looks like.
Since their work is characterized by the large, representative samples needed to take good methodological work forward, NSOs have an important role to play in building up the evidence base. It is also critical that they continue to collect, on a routine basis, the life evaluation measures that are most well-established in the field. Developing a time series of high-quality data is important for better understanding, in the longer run, of how subjective well-being covaries with other aspects of people's well-being, and ultimately the role that might play in informing public policy.
Using Subjective Well-being Data to Broaden the Evidence Base
While it is clear that we are far from having solved all the riddles of measuring the complex notion of people's happiness, we do think that the measures that exist today can tell us something important about how life is going for people. In particular, there is growing evidence that subjective well-being measures are sensitive to aspects of people's lived experience that simple economic statistics tend to miss. For example, subjective well-being tends to show strong relationships with people's health status, social connectedness, and trust in others - things that are often hard to capture through other means. Across a wide array of studies, one of the strongest factors predicting subjective well-being tends to be people's employment status; in particular unemployment is a major drain on life satisfaction, worthwhileness, and people's affect (i.e. emotional states and other feelings).
But what we also learn is that the impact of unemployment on people's life satisfaction is much larger than we might predict based on the loss of income alone. This highlights the importance that work plays in our lives, beyond money - for example, as a source of identity, meaning and purpose, social connectedness and more, and illustrates why replacing lost work is about more than just replacing lost income. Finally, when we ask people what matters most to their well-being through our Better Life Index website (www.oecdbetterlifeindex.org), the main message we get back, from over 200,000 users around the world, is that people aspire to be healthy, happy and wise: the top three rated dimensions are consistently health, life satisfaction and education.
How's Life in 2017?
The OECD is using subjective well-being to broaden its own evidence base, and provide a fuller picture of how life is going, for example in our flagship report, How's Life? 2017: Measuring Well-Being (http://dx.doi.org/10.1787/how_life-2017-en). Here, subjective well-being is used alongside other objective measures to benchmark countries' performance, and perhaps more importantly to monitor whether life is getting better for people. Comparing life today with the conditions experienced in 2005, prior to the financial crisis, we see that most OECD countries now have higher average household incomes, higher annual earnings, and a longer life expectancy (Chart 4). In around half of all OECD countries, the employment rate has risen since 2005; the incidence of long working hours has fallen; more people say they feel safe when walking alone at night; and there are fewer homicides. Yet in some of these outcomes, progress has often been slow, unsteady or unevenly distributed.
In addition, several other elements of people's well-being have been left behind: voter turnout, long-term unemployment and housing affordability have each worsened in around half of all OECD countries since 2005, while labor-market insecurity is higher in four-fifths. Feelings of life satisfaction and social support have also fallen in at least one-quarter of OECD countries. So as economies begin to regain their momentum after the crisis, there are many people who are not yet feeling the benefits, in several aspects of their lives.
How's Life in Japan?
The situation in Japan mirrors the OECD average in some respects (http://www.oecd.org/statistics/Better-Life-Initiative-country-note-Japan.pdf). Household net adjusted disposable income has increased, cumulatively, by 7% since 2005 - in line with the OECD average. Life expectancy, perceived safety, satisfaction with water quality, and exposure to PM2.5 air pollution have all improved in the last decade. Some aspects of jobs are moving in a positive direction: at 74%, employment levels in Japan are now 5% higher than in 2005, and the long-term unemployment rate has fallen below 2005 levels (having peaked in 2010). The share of employees experiencing job strain also fell by 5 percentage points between 2005 and 2015. Nevertheless, average annual earnings have shown little sustained progress since 2005, at odds with the 8% cumulative rise for the OECD on average.
Housing has become less affordable - with the average share of household disposable income spent on housing costs rising by almost 1 percentage point since 2005. And in line with the OECD as a whole, both life satisfaction and the share of people reporting that they have friends and family to count on in times of need have fallen since 2005. Meanwhile, voter turnout in the 2014 elections was almost 15 percentage points lower than in 2005 - again, consistent with wider OECD trends.
Countries' Experiences in Using Well-being and Happiness Metrics in Policy
So how can countries use these data to inform policy? In a recent contribution to the Global Happiness Policy Report 2018 (www.happinesscouncil.org) the OECD documented countries' recent experiences in developing well-being metrics, and the structured mechanisms that they are using to introduce these metrics into policy-making. Our chapter of the report features 13 countries' well-being measurement frameworks, as well as 10 countries' methods for bringing well-being into policy decisions, and more detailed case studies of seven of these countries (Ecuador, France, Italy, New Zealand, Scotland, Sweden, and the UK). Five broad conclusions can be drawn from this work:
1. Countries' measurement frameworks for well-being are all multi-dimensional (i.e. they consider several different aspects of people's lives).
2. Most include subjective well-being measures, but not always as a headline indicator for analysis.
3. Well-being metrics are used at different stages of the policy cycle, from strategic analysis, to policy design, implementation, and the evaluation of policy interventions.
4. Efforts have been initiated in different institutions in different countries (e.g. the prime minister, cabinet, parliament, NSO) with varying commitment and results.
5. These are still early days: progress has been significant in the last decade, but most initiatives are still evolving. Continued monitoring over time will be crucial in order to learn deeper lessons from their experiences.
Measuring What Matters Most to People
To understand whether people are better off, it is important to consider the views of people themselves - since sometimes these diverge from the picture we gain from the standard suite of economic statistics most commonly used to inform policy. In the future, the Better Life Initiative will continue to advance the statistical agenda on measuring well-being, including subjective well-being, to fill the remaining measurement gaps. At the same time, we will continue research on how these measures can be used to support the OECD vision of putting people's well-being at the center of its analysis, and helping member countries to deliver better policies for better lives.
|Printer friendly Cite/link Email Feedback|
|Publication:||Japan Economic Foundation (Tokyo, Japan)|
|Date:||Jun 30, 2018|
|Previous Article:||Interview with Dr. Francis Fukuyama, Olivier Nomellini Senior Fellow at the Freeman Spogli Institute for International Studies at Stanford University.|
|Next Article:||The Reversion of Cyberspace to the World of Classical Realism.|