Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change...

44
Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1

Transcript of Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change...

Page 1: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

1

Subjective Well-Being: Concepts and MeasurementsSiobhan McAndrewInstitute for Social ChangeUniversity of Manchester

Page 2: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

2

ObjectiveIn scope• Introduction to concepts,

measurement issues, main findings

• Discussion of what we know

• Work still to be done• Implications for public

policy and policy analysis

Out of scope• Presentation of own

research and findings

Page 3: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

3

What do I have to do with well-being?

•Economic historian by training•Avner Offer, Challenge of Affluence

▫1999-2001 M.Phil. seminar▫2001-2004 research assistance

•2005: the ‘happiness agenda’ takes off!•Whitehall Well-being Working Group, 2006-07•Policy position paper with Jonathan Lepper,

HMT•Religion and well-being; spatial models of

well-being

Page 4: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

4

Three strands of well-being

•Utility:▫pleasure (18c.); ▫preference satisfaction (20.c)

•Objective well-being: ▫welfare measures

•Subjective well-being: ▫self-reports of life satisfaction and

happiness

Page 5: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

5

Utility and ‘old’ welfare economics• Jeremy Bentham’s hedonic calculus• Utilitarianism: an action is right iff productive of

the greatest happiness of the greatest number – ‘the only right and proper end of government’

• Neoclassical economics:▫Edgeworth (1845-1926) and Pigou (1877-1959)▫Choices arise from preferences, constraints, and

expectations▫Utility can be cardinally measured (measurable and

comparable)▫To measure social welfare, add up utility functions.

Page 6: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

6

New welfare economics (Pareto, Kaldor, Hicks)

•Utility can’t be observed directly, but revealed in the prices people are willing to pay, and the choices they make (revealed preference theory)

•Society is allocatively efficient if goods are distributed to people who get the most utility from them

•Social welfare maximised if nobody can be made better off without someone else being made worse off (Paretian efficiency)

Page 7: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

7

Arrow’s impossibility theorem•To calculate a social welfare function via a voting

rule, we need to relax one of the following:▫Non-dictatorship▫Universality▫Independence of Irrelevant Alternatives▫Pareto efficiency

• No utilitarian social welfare function (fair voting rule) can convert individual choice into social optimality• i.e. Cannot add up individual happiness meaningfully.

Page 8: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

8

Measuring welfare, 1950-• Avoid direct interpersonal comparisons: ethical problem if

Government imputes a ‘universal preference ordering’ • Assume heuristically that underneath ‘we are all really

alike’ – our preference orderings are similar• Use proxies for utility:

▫ National income per capita (mean, median) UN System of National Accounts from 1953

▫ Adjust for unemployment, inflation, inequality▫ Leisure time, household production▫ Pollution and environmental costs

• But if income growth is not broadly shared, or particular groups lose, then can’t infer a Pareto gain.

• Must take care in interpreting real GDP increase as a good thing.

Page 9: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

9

Heuristics for public policy • Important to improve productive and allocative efficiency (e.g.

via technological progress, competition):▫ Higher income means budget constraints relaxed – so higher

levels of satisfaction possible▫ Allocative efficiency means that goods and services go to the

people who value them most• Pareto criterion may be impossible. So, assess interventions by Cost

Benefit Analysis: can losers be compensated by gainers? Can they be potentially compensated by gainers? (Kaldor-Hicks criterion)

• Examples:• Immigration – costs felt locally, benefits spread widely• Local economic development – infrastructure, construction, congestion• Reduction in tariffs – lower-priced imports but threat to domestic jobs• Opening up of professions

Page 10: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

10

• Commitment to value only outcomes is not neutral. • Is preference satisfaction an adequate conception of

well-being? ▫ Needs, individual dignity, opportunity, rights and fairness▫ Not necessarily reducible to utility

• Rawls: well-being should be measured by an index of ‘primary social goods’

• Sen: should not focus on the external means that permit individuals to attain various functionings. Should rather focus on ‘capabilities’: what individuals are free to do or to become. This takes into account all the relevant dimensions of life, rather than being purely concerned with either access to resources, or achieved utility.

Philosophical problems

Page 11: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

11

Conceptual problems, psychological evidence• People aren’t so rational

▫ framing effects▫habituation▫ loss aversion and risk aversion▫struggle to predict future tastes▫pleasure is related to novelty

• Smaller rewards given sooner preferred to larger, deferred rewards

• Building up picture of economy from individual choices and preferences flawed if utility functions are not ‘well behaved’

• Richer does not mean better off.

Herbert Simon: We are boundedly rational and use heuristics to make decisions

Kahneman and Tversky: Prospect Theory (losses weigh more than gains)

Page 12: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

12

Quality of life: objective measures•Index of Sustainable Economic Welfare

(Daly & Cobb, 1989)•UK Sustainable Development Indicators:

▫68 measures (though includes SWB and self-reported health) for Balanced Scorecard approach

•UN Human Development Index▫Life expectancy, education, income

•Results sensitive to variables included or adjustments made

Page 13: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

13

1975 1980 1985 1990 1995 2000 20050.8

0.820.840.860.880.9

0.920.940.960.98

1

UN Human Development Index, 1975-2005

UK Aus DenmarkGer US UK ranks 16/177

as of 2005

Page 14: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

14

Index of Sustainable Economic Welfare• Refined by Jackson and Marks for the New

Economics Foundation• GDP measured by consumer expenditure

▫minus spending on crime, private health and education, and pollution;

▫adjustment for longer-term environmental damage and depreciation of natural capital;

▫economic adjustments to give higher weighting to ‘prudent investment’ and trade balances;

▫adjustment for rising inequality; and▫adjustment for household labour

Page 15: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

15

‘Measure of Domestic Progress’

Page 16: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

16

Challenge of ‘happiness economics’

•1960s antimaterialism; 1970s environmentalism

•Fred Hirsch, Social Limits to Growth (1974)▫Economy = material economy + positional

economy ▫Socially scarce goods, or subject to congestion▫If no market pricing, then screening, queuing

or quality dilution▫e.g. leadership, suburban living, scenic land

•Brickman and Campbell (1971), hedonic treadmill

Page 17: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

17

Arrival of surveys!Eurobarometer

US General Social Survey

World Values Survey

European Social Survey

British Household Panel Study

GSOEP, HILDA

UK Sustainable Development Indicators

Gallup polls archive

Page 18: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

18

Subjective well-being indicators• Allows direct measurement of individuals’ well-being

Definition: ‘all of the various types of evaluations, both positive and negative, that people make of their lives...

reflective cognitive evaluations, such as life satisfaction and work satisfaction, interest and engagement

and affective reactions to life events, such as joy and sadness.

Thus, SWB is an umbrella term for the different valuations people make regarding their lives, the events happening to them, their bodies and minds, and the circumstances in which they live’.

Ed Diener, 2005

Page 19: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

19

Typical instruments• Single life satisfaction or happiness questions• Eurobarometer :

▫ ‘On the whole, are you very satisfied, fairly satisfied, not very satisfied, or not at all satisfied with the life you lead?’

• Data available 1973-present for EU members.• European Social Survey:

▫ ‘On a scale of 0 to 10, how happy are you?’ ▫ ‘On a scale of 0 to 10, how satisfied are you with life as a whole?’

• British Household Panel Study:▫ Using a 1-7 scale how dissatisfied or satisfied are you with your

life overall? • Eurobarometer periodically asks:

▫ ‘Taking all things together… would you say you are very happy, fairly happy or not too happy?’

Page 20: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

20

Life satisfaction ratings, 2007-2009

Page 21: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

21

Accuracy and reliability• Ceiling effects from finite scales• Cultural meanings, language effects• People remember peak or end experiences more clearly than moment-

by-moment life satisfaction• But self-reports correlate well with:

▫ likely attributes such as unemployment;▫ respondents’ recall of positive or negative life-events;▫ family and friends’ estimates; ▫ duration of authentic smiles; ▫ physiological measures (cortisol, ECG measures)responses to stress.

• Praag: ▫ ‘in reality we have never met a respondent who would refuse to answer

because his happiness level, being a ‘‘12’’ was not included... each respondent accepts and understands a finite scale’.

• Multi-item scales (e.g. BHPS GHQ-12; satisfaction with life domain questions) help solve measurement problems

Page 22: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

22

Data and path dependence• If multiple items available, can assess validity

of single items• Comparability of data across countries and

over time gives existing measures a comparative advantage

• Multi-item scales are more expensive• Continuing scepticism:

▫ ‘the concept sounds incredibly flaky and allencompassing based around subjective interpretations/expectations andexperiences’ (Political Science colleague)

Page 23: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

23

Dimensions of subjective well-being

•Newton/Defra: ▫hedonic well-being – includes happiness and life

satisfaction; and ▫eudaimonic or psychological well-being (PWB),

relating to meaning and self-actualisation.▫e.g. is it better to be the pig satisfied or Socrates

dissatisfied?▫Difficult to measure ‘self actualisation’ ▫Overlap with/reduction to hedonic well-being.

• Well-being and ill-being: distinct dimensions rather than mirror images

Page 24: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

24

Singapore

Kuwait

Cyprus

Spain

Italy

Britain

Mexico

Germany

NZ

Belgium

USA

Colombia

Malta

Ireland

Neths

Lux

Norway

Canada

Finland

Australia

Sweden

Austria

Iceland

Denmark

Switz

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00

Happy Life Expectancy, top 25 countries, 1995-2005

UK 20th/91

Page 25: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

25

Lux

Denmark

Australia

Greece

Canada

Ireland

UK

Italy

Cyprus

Austria

Switz

Malta

0.0 10.0 20.0 30.0 40.0 50.0 60.0

Happy Planet Index (West only), 2005

Adjusted for ‘ecological footprint’ (global hectares per capita). For full set, UK ranks 74 out of 173.

Page 26: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

26

Findings from survey data

•Positively related to SWB at individual level:▫income ▫marriage ▫job status ▫health▫religion▫being right-wing▫sex once a week (optimal number of

partners in past month= 1)

Page 27: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

27

Easterlin paradox

•Easterlin’s 1974 study suggested negative relationship between income and SWB across countries

•Income effect dependent on standards which were inherently changeable ▫expectation, habituation levels, and social

comparisons•Famous result that SWB flat in US since

1945, despite sustained economic growth

Page 28: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

28

Page 29: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

29

GDP, Wellbeing and MDP, 1973-2002

Page 30: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

30

Reasons for Easterlin paradox• UK GDP pc doubled since 1973, mean life

satisfaction has not:▫Basic needs largely met▫Relative income standards matter more for

well-being▫Hedonic adaptation

• Individual well-being increases with income, decreases with loss of position

• Positive income-wellbeing relationship at individual level, and across countries, but not national level over time

Page 31: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

31

Literature debate: relative effects

• Individual SWB responds positively to increases in income, negatively to increases in inequality and relative loss of position

• Income-SWB relationship stronger within a country at a point in time, than over time by country

• Status benefit of high income within a country – but zero-sum

• Clark & Oswald (1996): coefficients on income and relative income equal and opposite

• Helliwell & Huang (2005): life satisfaction completely relative in income.

• Alesina et al. (2004): individuals less happy if inequality is high. Effect stronger in Europe than the US.

Page 32: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

32

Dissenting opinions• For the very poor, absolute effect outweighs

relative effect• Diener et al. (1993):

▫‘To the degree that people compare themselves to virtually anybody, the theory becomes harder to rigorously test... if people’s standards adapt to virtually all conditions, why don’t they adapt to certain expectancies and social comparisons?’

• Some theoretical work suggesting people choose their reference groups

• Not only relative rank but distance (Hopkins 2008).

Page 33: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

33

Adaptation effects: pro-Easterlin•Bjørnskov et al. (2008): accelerated

growth is needed to influence trends in life satisfaction – because aspirations change over time

•Di Tella et al.; van Praag: effect of an income increase in long term only 40-42 per cent of effect after one year

•Gardner & Oswald study of moderate lottery wins: not complete return to original ‘set point’.

Page 34: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

34

Recent challenges• Economic security rather

than income; ill-being rather than well-being

• Veenhoven and Hagerty (2006):▫ data too fragile; ▫ Richer does mean happier;

inequality in SWB has fallen; happy life years increased.

• Stevenson and Wolfers (2008):▫ Income-happiness gradient

robust across countries and over time; no evidence of a satiation point.

Life Satisfaction in 91 nations, 1995-2005, and log GDP per capita, 2005

Page 35: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

35

The final word?• Layard, Mayraz and Nickell (2009)• US GSS (data from 1972); W. Germany GSOEP (data from

1984); Eurobarometer from 1973• US GSS and GSOEP: comparator income has a negative effect

on happiness equal in magnitude to the positive effect of own income

• US data : perceived relative income has same effect as actual relative income – comparisons matter

• Europe from 1973: trend income effect small and weak – so long-run effect of higher income at country level is unimportant for indidual SWB

• Differ from Stevenson & Wolfers because their study is largely cross-sectional and includes poor as well as rich countries.

Page 36: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

36

Models (1)

•Difficulty establishing causality from cross-sectional surveys▫e.g. well-being and divorce, well-being and

unemployment▫longitudinal studies more rigorous▫Clark and Oswald (2002) from BHPS

analysis: ‘the biases in cross-section patterns may be less dramatic than has sometimes been supposed’

Page 37: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

37

Models (2)

•Self-selection:▫People can pay more to live in more

pleasant areas▫Choose to have more or fewer children▫Leave education or stay on▫‘these have a strong choice element to

them, so that at the margin we would expect the utility effect to be zero’ (Clark et al. 2007)

Page 38: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

38

Models (3)

•Make as few assumptions as possible:▫What causes what?▫Causal effect of health on well-being? –

mental health difficult to distinguish from SWB

▫Many health behaviours ‘chosen’ (self-medicating)

▫Examine health and SWB as joint dependent variables

▫Add alcohol consumption/exercise etc. as mediating variables in path analytic framework

Page 39: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

39

e.g. Tampubolon (2010)

Page 40: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

40

Models (4)•Multi-level effects

▫Individual level▫Household level▫Neighbourhood level▫National level

•Latent variables or manifest variables•Spatial and temporal effects:

▫‘everything is related to everything else, but near things are more related than distant things’

▫correct for spatial & temporal dependence

Page 41: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

41

Disciplinary wars•Economics and its discontents - the ‘dismal

science’; ‘autistic’•But:

▫‘Economics is a study of mankind in the ordinary business of life… [and] that part of individual and social action which is most closely connected with the attainment and with the use of material requisites of well-being’ (Alfred Marshall)

• ‘Well-being isn’t the same as utility’• ‘Anti-growth nuts’

Page 42: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

42

Policy future• Cameron interest in ‘General Well-Being’:

▫At height of asset bubble▫ ‘Red Tory’ angle▫Quality of life important to the retired; anti-development

attitudes in the South East• Previous Governments’ sustainable development strategy

document drawn up by DEFRA:▫ ‘the goal of sustainable development is to enable all

people throughout the world to satisfy their basic needs and to enjoy a better quality of life, without compromising the quality of life of future generations’

▫Well-being surveyed alongside battery of questions regarding consumption, energy use and the environment.

Page 43: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

43

But wicked issues...• ‘We must tell people to consume less!’ • Loss aversion• Resistance to congestion charging and petrol taxes• Alcohol duty increased inflation this month• Obesity: taxes on processed food?• Marriage tax break?• Avoidance of progressive taxes• Ameliorating climate change is costly – requires growth• Austerity reduces funds for long-term interventions• Fragility of IAPTs programme• Governments and institutions face race to bottom• Policy usually more political than technocratic.

Page 44: Subjective Well-Being: Concepts and Measurements Siobhan McAndrew Institute for Social Change University of Manchester 1.

44

Conclusions• Good and growing evidence of what correlates with well-being

at the individual level: health, relationships, employment, good governance.

• Weaker evidence on ‘place’ and causality – need further waves of panel data

• Relative income hypothesis appears confirmed but need to understand mechanisms

• Technocratic magic bullets, even if possible, would not be desirable: people care about quality and fairness of process, as well as outcome

• Current economic priority is macroeconomic stability and reduction of unemployment. Evidence also highlights importance of:▫ alleviation of chronic illness, including mental illness;▫ strengthening relationships and social capital.