Inequality aversion and risk attitudes Ada FerreriCarbonell ICREA

Inequality aversion and risk attitudes Ada Ferrer-i-Carbonell ICREA & IAE-CSIC & Xavier Ramos Universitat Autònoma de Barcelona & IZA

Inequality aversion § Empirical evidence § Subjective measures: GINI has a negative effect on self-reported well-being. Alesina, Di Tella, Mac. Culloch (2004) & Schwarze, Harpfer (2007) § To what is it related? § Risk aversion: only experimental evidence shows that risk aversion & inequality aversion are related concepts (Carlsson, Daruvala, Johansson. Stenman, 2005 ). § Mobility

Our contribution § Empirical test on whether inequality aversion found in Western Europe is related to risk aversion. § Uses direct measures of utility and risk § Large representative panel data set with about 20. 000 individuals.

Measuring subjective well-being In conclusion, we would like to ask you about your satisfaction with your life in general. Please, answer according to the following scale: 0 means ´completely dissatisfied´, 10 ´completely satisfied´. How satisfied are you with your life, all things considered? 0___1___2___3___4___5___6___7___8___9___10 completely dissatisfied completely satisfied

Assumptions: empirical evidence § Interpersonal comparability: § Individuals able to recognize and predict satisfaction level of others (pictures, videos). § Common understanding of how to translate internal feelings into a number scale. § Measurability: § Strong positive correlation between emotional expressions (smiling, frowning, brain activity) & self-reported happiness. § Individuals will not choose to continue activities which yield low satisfaction levels.

Our satisfaction equation § Individual risk attitudes (Ri) assumed constant § Inequality at the federal/time level (I) (GINI) § Other explanatory variables (Xit). § Individual effects (about 80%) ( i). E. g. intelligence, and capacity to adapt to adversities. § Time (T) and federal (F) dummy variables.

Empirical testing § Risk attitudes constant over time § “β 3 Rt” can only be included with individual random effects. § We present results with FE and RE § RE: Mundlack specification for those variables that may correlate with individual time persistent traits § Cardinal/ordinal assumption makes no difference in terms of trade-offs Ferrer-i-Carbonell & Frijters (2004)

Measuring risk attitudes How do you see yourself: Are you generally a person who is fully prepared to take risks or do you try to avoid taking risks? Please tick a box on the scale, where the value 0 means: 'risk averse' and the value 10 means: 'fully prepared to take risks'. You can use the values in between to make your estimate. 0___1___2___3___4___5___6___7___8___9___10 risk averse fully prepared to take risks

Willingness to take risks (WTR) § Validity of WTR question: § Correlation between reported WTR & individual behavior in an experimental setting § Correlation between reported WTR & reported behavior (e. g. holding stocks, smoking, & occupational choice) § Correlation between reported WTR & individual characteristics (e. g. women are more risk averse, and years of education and income correlate negatively with risk aversion).

Data § The German Socio-Economic Panel § 1997 -2004 = about 20000 individuals & 115000 observations. § Sample means § Mean risk attitude = 4. 42. Most individuals (22%) report 5 and 46% of them report a 4 or less. § Mean life satisfaction = 7 § Household income = 2535 (euros/month/net) § Equivalent household income =1374

Risk attit. & inequality aversion § Individuals are inequality averse § The more risk averse, the larger the effect of inequality on well-being § Similar effects when introducing Ri with RE

The effects explained § The FE results: (all results are very similar) § For the most risk averse individuals, the effect of inequality = -2. 07 § For the average person, the GINI coeff = -0. 58 § The effect of GINI is positive for 21% of the sample (7 or more WTR) § RE: inequality coefficient positive for WTR = 10

Is it risk attitudes? § Risk attitudes correlated with other characteristics: § may it be something else than risk? § E. g. lower educated or women are more risk averse & face larger income fluctuations. § To test for this § We introduce an interaction term between GINI and (gender, years of education, & household income). § Only significant interaction: § GINI & household income, which is negative § It does not change the relationship between GINI & risk attitudes

Conclusions § Inequality aversion has been often proxied with estimates of risk aversion. § This paper employs two direct measures of utility and risk § To identify & estimate inequality aversion and risk aversion separately. and § To explore the relationship between the two.

Summary of results § Risk attitudes & inequality aversion are related § Risk attitudes determine the effect that inequality has on life satisfaction. § Most risk averse individuals are also the ones who are more inequality averse. § Consistent with the results in experimental setups.