Ec345 2024-25 Topic 6 - SWB and the Economy PDF

Ec345 Behavioural Economics:Theory and Applications Topic 6:The Economics of Happiness II Subjective Well-being and the Economy Alexander Dobson University of Warwick, Department of Economics November 2024...

Ec345 Behavioural Economics:Theory and Applications Topic 6:The Economics of Happiness II Subjective Well-being and the Economy Alexander Dobson University of Warwick, Department of Economics November 2024 1 Outline Previously - Economics of Happiness I: Subjective Well-being and Utility  Foundations  Measurement  Common Patterns  Easterlin Paradox? 2 Positional Economic Value as Explanation? (Clark, Frijters and Shields, 2008) 3 Outline Economics of Happiness II: Subjective Well-being and the Economy  Relative Economic Effects  Subjective Well-being and Income  The Relative Income Effect  Status concerns and Behaviour  Behavioural Agents  Macro and Subjective Well-being  Unemployment and Inflation 4 Subjective Well-being and Income Income as potentially important aspect in own right, but also crucially as proxy for economic consumption  Positive relationship but declining with income  Above a basic level of income, most of effect appears to be due to relative rather than absolute income Two separable (if related) components of reference-dependence: 1. Adaptation over time 2. Social Comparison 5 High Income Improves Evaluation of Life but not Emotional Well-being (Deaton and Kahneman, 2010) 6 Poverty Exacerbates the Effect of Adverse Circumstances (Deaton and Kahneman, 2010) 7 Top Income and Wellbeing in Context (Deaton and Kahneman, 2010) 8 Income and Adaptation Changes in income also exhibit the familiar pattern of subjective well-being adaptation seen for other well-being events German Socio-Economic Panel (Di Tella et al, 2008) 9 Consumption Of course, the well-being effect of income also presumably depends on what you spend it on  Economic theory assumes it is (solely) consumption that is the ultimate carrier of value One claim: spending money on others promotes happiness (Dunn et al, 2008)  Those who give proportionally more to charity are happier, regardless of income  People may also derive more satisfaction from experiential purchases than material purchases (Gilovich et al, 2014) 10 The Relative Income Effect  Substantial part of the Well-being subfield of the Behavioural Economics literature  Observed across many samples (despite empirical challenges of reducing social relativity concerns to a single, imposed aggregative reference)  Clark and Oswald (1996) UK BHPS, Blanchflower and Oswald (2004) US GSS, Ferrer-i-Carbonell (2005) German GSOEP, Graham and Felton (2006) 18 Latin American Countries, Knight and Song (2006) China … Layard et al (2018) 11 Social Comparison Affects Reward-Related Brain Activity (Fliessbach et al, 2007) 12 Relative Income Effect: Imposed Reference Group Structure If relative value – what reference/s?  As-if rational reference group choice (Falk and Knell, 2004): comparison with similar others (trading off self- improving upwards comparisons with self-enhancing downwards comparisons)  Comparison potentially automatic (and so dependent on exposure, and endogenous to action) Literature generally imposes a single aggregative comparison reference (entering as difference or ratio)  Ideal measure: weighted mean over reference group (e.g. by proximity)  Relative income effect literature: simple reference group mean (where reference group often restricted to similar age, gender, region etc.)  Potential confound: local cost of living 13 Neighbours as Negatives (Luttmer, 2005) 14 Neighbours as Negatives (Luttmer, 2005) 15 Relative Income Effect: Mechanism? (Luttmer, 2005) The array of alternative satisfaction measures provide suggestive evidence about the possible mechanism  Neighbours’ earnings significantly increase satisfaction with one’s city or town (suggesting local cost of living is not responsible)  Neighbours’ earnings significantly reduce overall subjective well-being, reduce leisure time, reduce satisfaction with leisure time and reduce satisfaction with friendships People appear to be giving up leisure, to allow their friendships to suffer, and to work more - perhaps in an attempt to mimic the material living standards of their neighbours  Consistent with the substitution of goods yielding extrinsic satisfaction (material possessions, status) for those yielding intrinsic satisfaction (time for family, friends, hobbies) because predictions of future utility from the extrinsic attributes of consumption are systematically biased upwards 16 Importance of Rank? Other aspects of the reference group distribution may be relevant to social comparison effects (beyond a reference group mean as a single aggregative reference) e.g. rank, distance from max/min etc.  Boyce, Brown and Moore (2010) find income rank predicts life satisfaction in the BHPS, whilst absolute and reference income have no effect 17 Consumption, Status and Value in Economics  Adam Smith (Wealth of Nations,1776): “A linen shirt … is, strictly speaking, not a necessary of life. The Greeks and Romans lived, I suppose, very comfortably though they had no linen. But in the present times, through the greater part of Europe, a creditable day-labourer would be ashamed to appear in public without a linen shirt, the want of which would be supposed to denote that disgraceful degree of poverty which, it is presumed, nobody can well fall into without extreme bad conduct.” ▪ Thorstein Veblen, conspicuous consumption ▪ James Duesenberry, Relative Income Hypothesis ▪ Fred Hirsch, positional goods 18 Consumption, Status and Value in Welfare Economics The relevance of interdependent economic value, including social relativity and associated externalities, was also explicitly recognised in major post-war discussions and reviews of welfare economics ▪ For example, Little (1950), Graaff (1957), Mishan (1959) ▪ Scitovsky (1976) The Joyless Economy 19 Systematic, Layered Qualification I. Positional Externalties  Frank II. Optimal Response to a change in reference  Clark and Oswald (1998) III. Systematic Mistakes with endogenous social reference-dependence  Dalton and Ghosal, projection bias 20 Positional Externalities (Frank 1997, 2005) Does greater economic consumption lead to increased happiness?  Frank argues that for many goods, the evidence suggests not (beyond some point)  If alternative use of the same resources could enhance wellbeing, why don’t we do this? Possible explanations: i. Incomplete information about adaptation (discussed in Week 5: Projection Bias) ii. Positional externalities ensure that many forms of consumption are more attractive to individuals than to society as a whole (social dilemma/cooperation problem) 21 Positional Externalities Positional Competition Cooperative strategy Competitive strategy Cost ▪ Bull elk contest Moderate size antlers Large size antlers Potentially fatal ▪ Arms race Low defence costs High defence costs Misallocated resources ▪ Competitive consumption Constrained Unconstrained Misallocated resources ▪ Relative consumption or savings Adequate savings Inadequate savings Retirement well-being ▪ Labour supply, school district Moderate workweek Long workweek Family time 22 Positional Economic Value and the Easterlin Paradox (Clark, Frijters and Shields, 2008) 23 Positional concerns reducible to Envy? Generally: social reference-dependence in economic evaluation, behaviour and experience  Multifaceted issue, reflected in diverse literature and variety of terminology  (‘envy’ minor aspect)  Non-trivially related to SR and LR value/welfare 24 Positional concerns reducible to Envy? Roles of Social Comparison: i. Subjective perception - direct impact on quality of experience (relative value decreasing in the social reference and positive for downward comparison, negative for upward) ii. Position-dependent allocation and social functioning - reward via relative performance in contests, as well as more general social functioning (value decreasing in the social reference) iii. Information and motivation (higher references potentially positively related to value) where i. and ii. both imply negative positional externalities 25 Status Goods: Evidence from Platinum Credit Cards (Bursztyn et al, 2018) Directly testing for status concerns in consumption behaviour is challenging, as with observational data difficult to separate unobserved intrinsic consumption utility from a desire to signal high income and status Field-experimental evidence on platinum credit cards as status goods  Large bank in Indonesia  Urban, upper-middle-class sample  130m (of 330m) global luxury goods consumers located in emerging markets 26 Status Goods: Evidence from Platinum Credit Cards (Bursztyn et al, 2018) 27 Demand for Status Signalling, and in Social Contexts (Bursztyn et al, 2018) First experiment (n = 835): test demand for pure status component of the platinum card via contrast with a control product holding constant all the instrumental benefits  Demand increases to 21% take-up (at market price) vs. 14% for the control  7.3%pt. increase economically meaningful: a second call-back with a 25% discount only increased take-up 3.7%pts. Observational data for individual credit card transactions (n = 2,492) also showed platinum cards were more likely to be used in social situations (bars, restaurants)  Likely substitution away from cash and other cards as evidence actual restaurant visits don’t differ between platinum and standard card holders  At a cost – around half are forgoing discounts / cash-back rewards on other cards  Suggests platinum cards used to build social-image 28 Identifying Positional Externalities (Bursztyn et al, 2018) Their next experiment tested for the presence of positional externalities by offering a new, more expensive but functionally identical ‘diamond card’ to current platinum card holders A control group was told no more information, whereas a treatment group (n = 93) was informed that the income criterion for the platinum card they current hold has been reduced  The additional information nearly doubled take-up of the diamond card (41%, from 22%)  Making the status good accessible to lower income customers weakened its income signalling power, imposing a positional externality on higher-income consumers  Higher-income consumers then demanded a more exclusive status good 29 The Dampening Impact of Self-Esteem (Bursztyn et al, 2018) Bursztyn et al’s final experiments provide suggestive evidence that increased self-esteem may have a causal effect reducing the demand for status goods  Bank customers in a treatment group completed a self-affirmation task (describing an event that made them proud), whereas a control group undertook a placebo task (describing media habits)  Point estimates suggest large reductions in demand for platinum credit card use, but small sample size problems (n = 167) in identifying effect precisely  A parallel design online (n = 405) elicited preferences for luxury brand apparel (as a classic status good) vs. nonluxury apparel, and found that higher self-esteem resulted in substantially less demand for the status good  Self-image and social-image may be substitutes, rather than complements 30 A Theory of Rational Emulation (and Deviance) Clark and Oswald (1998) explore a model of rational behaviour with a comparison term in the utility function so that individuals care about (and respond to) the actions of others, developing the implications for conformity and deviance  Goal is to generate plausible reasons to imitate, without simply assuming a need to fit in  Concern for relative position is in itself insufficient to bring about emulation: the optimal reaction to the behaviour of others depends crucially on the curvature of the comparison term 31 Comparison-Concave Utility and Following Behaviour Additive comparisons model:  FOC for interior maximum:  Differentiating implicitly to give the individual’s response to others: Since denominator negative (by concavity requirement of maximisation problem), comparison-concave utility implies optimal following behaviour (comparison-convex, deviance) 32 Comparison-Concave Utility and Following Behaviour Comparison-concave utility Comparison-concave utility produces following behaviour: (1) Reference increases, 𝑎∗ ↑ ∗ (2) Position relative to reference worsens, 𝑎0 − 𝑎 ↓ (3) Marginal value of status increases, 𝑣 ′ (𝑎0 − 𝑎∗ ) ↑ (assuming concavity in status term) 𝑣 ″ (𝑎0 − 𝑎∗ ) < 0 (4) So optimal response is to increase action, 𝑎 ↑ 33 Simulated Response to Fall in Leaders’ Costs  Leaders: comparison-convex utility  Followers: comparison-concave utility  Time lags in information assumed 34 Expenditure Cascades (Frank, 2005) Sharply increased spending by top earners may exert indirect upward pressure on the median earner by shifting the frame of reference for those just below them, and so on “all the way down”  Frank claims that traditional theories of consumption predict that observed income growth in the US since 1979 should have resulted in spending increases by the median earner of only around 15 percent by 2005 (whereas the top 1% should be spending 3 times as much)  However, the median size of a newly constructed house has increased from 1,600 square feet in 1980 to 2,100 square feet by 2001, more than twice the predicted increase  US counties with higher earnings inequality have significantly higher median house prices, personal bankruptcy rates, divorce rates, and average commute times 35 Happiness and Comparison Intensity (Clark and Senik, 2010) 36 Alternative Wellbeing Measures (Clark and Senik, 2010) 37 Behavioural Agents: Feedback from Actions to Preference Parameters Social reference-dependence may be an important context in which to apply a general mechanism for systematic mistakes described by Dalton and Ghosal (2010, 2012)  In the model, frames or psychological states are considered as preference parameters that are endogenously related to the choice of action, rather than exogenously given  If comparison is to some degree automatic (e.g. Gilbert et al, 1995) then the reference group for an individual (and its weighting) are dependent on the actions of the individual via exposure  If individuals take this feedback from their actions to their social reference group into account, then the agent may be classed as (in this respect) sophisticated; if individuals fail to take this feedback (adequately) into account, they may make systematic mistakes 38 Failure to Recognise Reference Change - a simple illustration  Suppose the individual is initially in a moderate wage, moderate consumption, moderate reference environment, enjoying a payoff of 3  She faces a choice: if she sacrifices more of her leisure/family time, she can obtain high wages and consumption; alternatively, she can remain in her current situation  If her references remained moderate, she would enjoy an increased payoff of 4  However, in obtaining high wages and consumption she becomes exposed to high references, adapting to the benefits but still incurring the costs, resulting in a payoff of 2 39 Incorrect Projection of Current References This is an example of Projection Bias (see Week 5)  A sophisticated agent takes the feedback from her actions to her references into account, correctly anticipating her future preferences  A behavioural agent fails to recognise this feedback and so neglects adaptation to the action-dependent social reference context, incorrectly projecting her current references on to her perceived future preferences  In general, projection bias suggests we often simply overweight current preferences, rather than projecting them exclusively – we recognise some of the shift in preference, but neglect its magnitude 40 A General Channel for Projection Bias: Feedback from Actions to Frames Topic 5 illustrated Projection Bias using evidence from Conlin et al (2007) that catalog consumers had systematically failed to fully recognise the impact of variation in the weather on their cold-weather item preferences, variation that was essentially exogenous in nature However, as the mechanism characterised by Dalton and Ghosal (2012) emphasises, any endogeneity of context (including changes in social reference context) suggests a further potential channel for systematic mistakes  ‘Frames’ as preference parameters endogenously related to action, rather than exogenously given  If individuals fail to take this feedback (adequately) into account, they may make systematic mistakes Projection bias may have a wide variety of applications 41 Relative Economic Effects and the Economy - Some General Conclusions  Status and social comparison give rise to positional externalities and (in the absence of institutional solutions) distorted choices  The Relative Income Effect literature consistently finds a negative relationship between subjective well- being and the average income of an imposed reference group (possibly due to emulation, greater labour supply, and a neglect of friendships and leisure time)  Higher intensity of social comparison is also associated with lower subjective well-being  Comparison-concave utility may induce following behaviour, and so also potentially generate cascade effects  Neglecting the extent of adaptation to a dynamic, action-dependent social reference context may lead to systematic bias 42 Internalities and Future Adaptation Neglect Adaptation more generally also implies similar feedback from actions to own outcomes as ‘self-references’ for adjusted future preferences, a mechanism that may also be neglected  Potential for self-imposed ‘internality’, comparable to an externality Possible implications for life-cycle consumption (Loewenstein et al, 2003)  When consumption is habit-forming, rationality (generally) implies an increasing consumption profile  If projection bias leads to an underappreciation of the impact of current consumption on future utility (the negative ‘internality’ created by neglected adaptation), individuals may plan to consume too much early in life, and too little late in life  Projection bias may also lead to repeated readjusting of immediate consumption upwards relative to recent plans  Individuals may decide to increase consumption more than previously planned, reducing saving 43 Outline Economics of Happiness II: Subjective Well-being and the Economy  Relative Economic Effects  Subjective Well-being and Income  The Relative Income Effect  Status concerns and Behaviour  Behavioural Agents  Macro and Subjective Well-being  Unemployment and Inflation 44 Wellbeing and Unemployment Layard et al (2018) 45 Inflation and Unemployment: Trade-offs Modern macro textbooks embed the assumption of a social welfare function defined on inflation and unemployment. Although usually only implicit, SWB data can be used to explicitly estimate this function.  Di Tella et al (2001, 2003) show that life satisfaction is negatively associated with both inflation and unemployment  The relative size of the estimated life satisfaction coefficients attached to these variables allow the (reported life satisfaction) costs of inflation in terms of unemployment to be calculated  Recessions and high unemployment in the economy seem to impose substantial psychic costs above the loss in output and the impact on the unemployed – plausibly due to fear of unemployment  Higher unemployment benefits are associated with higher national well-being 46 Inflation and Unemployment: Trade-offs (Di Tella et al, 2001) 47 Inflation, Unemployment and Unemployment Benefits (Di Tella et al, 2003) 48