Adaptation, Anticipation and Social Interactions in Happiness: An Integrated Error-Correction Approach Maarten Vendrik Maastricht University IZA
Research area Dynamics of happiness of individual people over time Underdeveloped area Fertile area for application of modern techniques from time series analysis of panel data
Three phenomena Three important phenomena: 1. Hedonic Adaptation E.g. lottery winners: first very happy, later fall back holds for changes in most determinants of happiness 2. Anticipation E.g. people who marry: already happier before wedding holds for changes in some determinants of happiness 3. Social reference effects E.g. happiness from consumption: how much in comparison to others
Contribution of my study Happiness literature: Phenomena mostly investigated separately from each other (but strongly correlated!) No control for similar dynamics in control variables Integrated investigation of implications of three phenomena for the dynamics of individual life satisfaction Focus on household-equivalent income, but control for similar dynamics wrt large set of control variables
Outline Dynamic life satisfaction equation Main results for income and social reference income Explanation of Easterlin Paradox for West Germany
1. Dynamic life satisfaction equation Estimations with individual fixed effects for German Socio- Economic Panel (SOEP), 1984-2007, West Germany Error-correction model (ECM) neat distinction between: short-run shock effects of explanatory variables on life satisfaction (future + current) long-run level effects of explanatory variables on life satisfaction Strength of adaptation to/reinforcement by change in X = = long-run effect of X - sum of short-run effects of X
Time path of life satisfaction as a result of income shock of 50% at t=0 Speed of adaptation ρ = 0.73
Income endogenous Income endogenous due to: Reverse causality and spurious correlations Measurement error and unobserved costs of income generation Instrumenting household-equivalent income by average household-equivalent labour earnings in occupations and industries of household members
2. Main results Anticipation effect of income insignificant (reverse causality) Strong current income effect and strong adaptation to income Long-run income effect insignificant complete adaptation cannot be rejected Social reference income has significant, negative and strong effect in the long run, but not in the short run Absolute income effect significant and positive in the short run, but insignificant in the long run Bias-corrected coefficient of lagged life satisfaction = 0.73 adaptation to income shock and reinforcement of reference income shock takes place for more than 90% in three years
6.9 7 7.1 7.2 7.3 7.4 Time path of life satisfaction as a result of absolute income shock of 50% at t=0-2 0 1 2 4 6-3 -1 3 5 Year
6 6.5 7 7.5 3. Easterlin paradox for West Germany: Income-life satisfaction relation across individuals in 1995 10000 20000 30000 40000 Mean Annual Real Net Income
6.8 7 7.2 7.4 7.6 Mean Net Annual Income 16000 18000 20000 22000 Average income-life satisfaction relationship over time 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 year mean life satisfaction mean net annual income
Possible explanation of Easterlin paradox In cross section richer people happier because of higher income relative to others income Over time everybody becomes richer incomes relative to others incomes remain the same if income distribution does not change People fully adapt to absolute income rises no significant effect in long run Both effects everybody remains at same happiness level in the long run
Publication More details in: Maarten Vendrik (2013), Adaptation, anticipation and social interaction in happiness: An integrated errorcorrection approach, Journal of Public Economics, 105: 131 149 Thank you!