Work and Well-Being: Insights from Happiness Research

Transcription

1 Work and Well-Being: Insights from Happiness Research Is happiness measurable and what do those measures mean for policy? Rome, 2-3 April 2007, University of Rome Tor Vergata Alois Stutzer, University of Basel Outline Received wisdom about work and well-being in economics (text books) Happiness research in economics enriches the picture research questions selective choice of findings Focus: Why does unemployment hurt the employed? Concluding remarks Alois Stutzer 2

2 Received Wisdom about Work and Well-Being in Economics (Text Books) People offer their work in exchange for a salary that allows buying consumer goods finally providing utility Work is only means to an end and is not itself contributing to people s well-being Potential insights from happiness research: What makes for intrinsically more rewarding jobs? Alois Stutzer 3 Received Wisdom Work place amenities and disamenities are reflected in compensating wage differentials On a competitive labor market (and homogenous preferences), people enjoy the same utility level given their income generating potential Potential insights from happiness research: To what extent does the idea of compensating wage differentials hold? Alois Stutzer 4

3 Received Wisdom Unemployment is a waste of productive resources In new classical economics, people are voluntary unemployed as their reservation wage is above the salaries offered or transfers received Potential insights from happiness research: What are the psychic costs of unemployment? Alois Stutzer 5 Happiness from Work Three applications Volunteering as intrinsically rewarding work Gains from sorting and matching in the labor market Rents in the public bureaucracy Alois Stutzer 6

4 Volunteering as Intrinsically Rewarding Work Two views (Menchik and Weisbrod 1987) Intrinsic work enjoyment / people enjoy helping others per se People volunteer instrumentally in order to receive a by-product of volunteer work Empirical anlysis (Meier and Stutzer 2007) German Socio-Economic Panel, Reported life satisfaction Frequency of volunteering 4 weekly, 3 monthly, 2 less frequently and 1 never Alois Stutzer 7 Volunteering as Intrinsically Rewarding Work Meier & Stutzer (2007). Forthcoming in Economica. Alois Stutzer 8

5 Collapse of the infrastructure of volunteering after reunification Report of the Enquete-Commission of the German Bundestag on the Future of Civic Activities More than 37 percent of all the volunteers in the GDR stopped their volunteer work between 1989 and percent reported that they stopped volunteering due to the termination of groups and organizations which previously provided opportunities for civic engagement, i.e. societal mass organizations or publicly owned firms. Alois Stutzer 9 Volunteering as Intrinsically Rewarding Work Meier & Stutzer (2007). Forthcoming in Economica. Alois Stutzer 10

6 Volunteering as Intrinsically Rewarding Work Meier & Stutzer (2007). Forthcoming in Economica. Alois Stutzer 11 Gains from Sorting and Matching in the Labor Market Two emphases Market forces bring about that skills are employed in their most productive use so that goods and services are supplied at lowest costs Marginal worker determines the compensation for labor in its many specificities Other workers individually get a utility premium if they experience an optimal match between job characteristics and their preferences Private benefits from sorting and matching are the larger the more heterogeneous the preferences are in a society ceteris paribus Alois Stutzer 12

7 Gains from Sorting and Matching in the Labor Market Empirical analysis (Luechinger, Stutzer and Winkelmann 2007, Mimeo) European Social Survey, 2002/2004 Reported life satisfaction Sorting between private and public sector Estimation procedure ordered response model with endogenous switching Alois Stutzer 13 Alois Stutzer 14

8 Alois Stutzer 15 Compensation of Amenities Direct test for differentials in SWB Clark (2003), Clark and Senik (2006) Empirical analysis (Luechinger, Meier and Stutzer 2008) Private sector vs public sector employment Rents from corruption (e.g. bribes)? European Social Survey & Latino Barometer Life satisfaction differentials in 42 countries Correlation with perceived corruption Alois Stutzer 16

9 Figure 2. Corruption and rents in the public bureaucracy GR CZ PY DE AT LU CL UY EE PT PE BR SV SK CO MX PA PL NI BO AR HN EC VE GT UA IS DK FI GB NL NO CH IE FR BE ES SI IT CR SE IL Irregular payments to bureaucrats Notes: (1) GLS estimation; (2) shaded area is 95% confidence interval; (3) life satisfaction differentials are corrected for log(gdp per capita) and separate constants for Europe and Latin America, respectively. Sources: ESS 2002/3 and 20 04/5, Graham and Felton (2005) based on LB 1997, 2000, 2001 and 2003, Heston et al. (2002) and Gwartney and Lawson (2004). Luechinger, Meier & Stutzer (2008). Forthcoming in the Journal of Law, Economics and Organizations. Alois Stutzer 17 Unhappiness from Unemployment Becoming and being unemployed is for many people a most negative experience Unemployment decreases mental health leading to higher death rates, suicide, drug usage, etc. Argyle (1989), Clark and Oswald (1994) Scarring effects Clark et al. (2001) Social sanctions due to social work norms Clark (2003) Stutzer and Lalive (2004) Alois Stutzer 18

10 Why Does Unemployment Hurt the Employed? Preliminary results from joint work with Simon Lüchinger (University of Zurich) and Stephan Meier (Boston Fed) Life Satisfaction and Unemployment, West Germany Unemployment rate Life satisfaction Source: GSOEP and Statistisches Bundesamt Deutschland. Alois Stutzer 20

11 Why Do Employed People Care About Unemployment? Job security decreases in recessions Job insecurity decreases well-being and increases psychosomatic problems and physical strains (e.g. De Witte 1999) Harsher working conditions Pressure on working time and wages (see Blanchflower and Oswald 1994 on the wage curve) People feel empathy with unemployed people Worries about general conditions in society, e.g. public finances, crime,... Alois Stutzer 21 Testing the Value of Job Security Comparison of workers in two sectors which differ sharply in job security - over the business cycle Public and private sector in Germany, US, and Europe Hypothesis Public servants suffer less from high general unemployment Proxy for individual welfare: reported life satisfaction Alois Stutzer 22

12 Testing the Value of Job Security Empirical setup Private/public sector life satisfaction gap over the business cycle Data for West Germany, US, and Europe Results Life satisfaction gap and unemployment Job security and unemployment Alois Stutzer 23 Differences between Public and Private Sector Employment Significant difference in job security, due to... difference in protection from dismissal In Germany, public servants can only be dismissed if (i) convicted of at least one year in prison for criminal charges or (ii) six month in prison for homeland security charges (paragraph 48 of laws on civil servants) difference in sensitivity to business cycle Private firms are much more likely to declare bankruptcy Employment in private sector is more cyclical than in public sector (e.g. Blank 1993 for US and UK) Alois Stutzer 24

13 New Entrants into Unemployment, West Germany New Entrants into Unemployment 6% 5% 4% 3% 2% 1% 0% Year Unemployment Rate Previously public servant Previously public sector Previously private sector Unemployment Rate Source: GSOEP and Statistisches Bundesamt Deutschland. Alois Stutzer 25 Data Set: Germany German Socio-Economic Panel, ,582 observations (working population in West Germany) Subjective well-being How satisfied are you with your life, all things considered? Scale from 0 completely dissatisfied to 10 completely satisfied Information on public employment. Two definitions: 5,598 (1,706) working in public sector (public servants) 15,182 (17,866) working in private sector (non public servants) Unemployment across Länder (year average) Data source: German Federal Statistical Office Alois Stutzer 26

14 Life Satisfaction Gap in Germany Source: GSOEP and Statistisches Bundesamt Deutschland. Alois Stutzer 27 Public Servants Suffer Less from High Unemployment Dependent variable: Satisfaction with life Ordered Probit Ordered Probit Coef. Coef. Unemployment Rate (mean adjusted) (0.007)** (0.007)** Public servant (Beamter) (=1) (0.025)** Public servant (Beamter) * UR (0.009)** Public sector (=1) 0.072(0.015)** Public sector * UR 0.018(0.005)** State dummies Yes Yes Year dummies Yes Yes Individual fixed-effects No No Clustered Individual Individual No. of observations 105, ,582 No. of individuals 19,225 18,926 Chi ** ** Notes: Dependent variable: life satisfaction on a 11-point scale. Robust s tandard errors in parthesis. Source: GSOEP, West Germany Significant levels: ** p < 0.01, * p < Alois Stutzer 28

15 Potential Caveats Sectors differ in other dimensions than job security differences in wage sensitivity to the business cycle control for net hourly wage and household income different pressure at work place over the business cycle control for working time Private and public sector employees differ in innate traits control for personal characteristics, like gender, education, children,... control for individual fixed-effects Self-selection Alois Stutzer 29 Results for West Germany, Dependent variable: Life satisfaction Ordered Probit OLS Coef. Coef. Unemployment Rate (mean adjusted) (0.007)* (0.010)** Public servant (Beamter) (=1) (0.026)* (0.064) Public servant (Beamter) * UR (0.009)* (0.011)** Personal Controls Yes Yes State dummies Yes Yes Year dummies Yes Yes Individual fixed-effects No Yes Clustered Individual State*year No. of observations No. of individuals Chi ** - F-value ** Notes: Dependent variable: life satisfaction on a 11-point scale. Robust standard errors in parentheses. Source: GSOEP, West Germany Personal controls include: working time, working time squared, net hourly wage, hh income, no. hh members, head of hh, ch ildren in hh, gender, age, age squared, nationality, working part time, and marital status. Significant levels: ** p < 0.01, * p < Alois Stutzer 30

16 Robustness of Results Robust to... cluster on state*year restricted to Germans public sector employment instead of public servants excluding Berlin Basic results can be replicated for the US and for Europe Alois Stutzer 31 Is Job Insecurity the Driving Force? Unobservable differences between public and private sector employment over the business cycle might explain the observed empirical pattern How much of the life satisfaction difference can be directly attributed to perceived job security and individual economic situation? Survey questions: What is your attitude towards the following areas - are you concerned? your job security: very concerned, somewhat concerned, not concerned at all (3-point scale) your economic situation: very concerned, somewhat concerned, not concerned at all (3-point scale) Alois Stutzer 32

17 Perceived Job Insecurity across Sectors Alois Stutzer 33 Life Satisfaction and Job Security Coef. Coef. Coef. Unemployment Rate (mean adjusted) (0.010) ** (0.010) * (0.010) (*) Public servant (Beamte) (=1) (0.064) (0.063) (0.062) Public servant (Beamte) * UR (0.011) ** (0.011) ** (0.011) * Job Security: "Very concerned" Reference Group Reference Group "Somewhat concerned" (0.017) ** (0.017) ** "Not concerned at all" (0.019) ** (0.019) ** Economic Situation:"Very concerned" Reference Group "Somewhat concerned" (0.016) ** "Not concerned at all" (0.020) ** Personal controls Yes Yes Yes State dummies Yes Yes Yes Year dummies Yes Yes Yes Individual fixed-effects Yes Yes Yes Clustered Stateyear Stateyear Stateyear No. of observations 105, , ,242 No. of individuals 19,225 19,225 19,225 F-value 63.4** ** ** Notes: OLS-Regressions. Dependent variable: life satisfaction on a 1 1-point scale. Robust standard errors in parenthesis. Source: GSOEP, West Germany Significant levels: ** p < 0.01, * p < 0.05, (*) p <0.1. Alois Stutzer 34

18 Discussion Size of the effect Comparison to becoming unemployed One standard deviation increase in the rate of unemployment (2.5 pp) is about a 10th of the size of becoming unemployed (compare Stutzer and Frey 2003 for Germany) Selection of risk averse individuals into the public sector (Barsky et al for the US and Bonin et al. (2006) for Germany) As risk averse individuals suffer more from insecurity, the reported effect is downward biased - compared to random assignment to two sectors with different unemployment risks Alois Stutzer 35 Discussion Evidence from the life satisfaction gap shows, that for protected public servants, business cycles are less worrying the well-being of private sector employees decreases in recessions partly because of increased job insecurity Should dismissal protection be increased even more? Mainly evidence from Germany with high general protection against dismissal Employers become more reluctant to hire people Decreases chance of finding job when dismissed Alois Stutzer 36

19 Concluding Remarks Policy oriented future research on individual subjective well-being Intrinsic rewards from work What institutions support the realization of the gains from sorting and matching? What institutions allow controlling rent-seeking of the public bureaucracy? Unhappiness from unemployment What institutions reduce the psychic costs of business cycles? Different effects of incidence and duration of unemployment Alois Stutzer 37 Full specification, West Germany (1/2) Ordered Probit OLS Unemployment Rate (mean adjusted) (0.007)* (0.010) ** Public servant (Beamter) (=1) (0.026)* (0.064) Public servant (Beamter) * UR (0.009)* (0.011) ** Working time (0.002)** (0.002) ** Working time2/ (0.002)** (0.003) ** Net hourly wage, ln (0.015)** (0.020) ** HH income, ln (0.015)** (0.021) ** No. of HH members1/ (0.024)** (0.034) ** Children in HH (0.016) (0.020) * Head of HH (0.016) (0.032) ** Female (=1) (0.017) Age (0.004)** Age2/ (0.005)** (0.008) * Years of education, ln (0.036)(*) (0.076) Alois Stutzer 38

20 Full specification, West Germany (2/2) Nationality Foreign (Non Euro Area) Reference Group German (0.023)** Euro Area (0.031)** Working Full Time Reference Group Reference Group Part Time (0.024)(*) (0.026) Marital Status Married Reference Group Reference Group Separated (0.034)** (0.049) ** Single (0.020)** (0.027) * Divorced (0.024)** (0.037) Widowed (0.051)** (0.096) ** Constant (0.333) ** State and Year dummies Yes Yes Individual fixed-effects No Yes Clustered Individual Stateyear No. of observations No. of individuals Chi ** - F-value ** Notes: Dependent variable: life satisfaction on a 11-point scale. Standard errors (in parenthesis) clustered at the level indicated in the specification. Source: GSOEP, West Germany Significant levels: ** p < 0.01, * p < Alois Stutzer 39 Data Set: United States General Social Survey (GSS), repeated cross-section for 20 years 14,704 individuals who are full-time working 1,222 in public administration 13,482 in private sector Subjective well-being Taken all together, how would you say are things these days - would you say that you are very happy, pretty happy, or not too happy? (3-point scale) --> dummy = 1 if very happy Public employment industry codes Unemployment: State level from BLS/Haver Analytics Alois Stutzer 40

21 Results for US, Public Private All Coeff. s.e. Coeff. s.e. Coeff. s.e. State unemployment rate (mean adjusted) (0.043) (0.010)* (0.010)* Public Administration (0.040) Public Administration * UR (0.021) ( * ) Personal controls Yes Yes Yes State fixed effects Yes Yes Yes Year fixed effects Yes Yes Yes Clustered State*year State*year State*year N ,482 14,704 Chi Notes: Probit specifications. Dependent variable: dummy coded as 1 for an individual with the highest stated happiness: "very happy." Control variables include gender, age, age squared, race, no. childeren, income quartile, hh size, education, marital status, and size of town. Standard errors are robust to heteroskedasticity. Source: GSS, United States Significant levels: ** p < 0.01, * p < 0.05, (*) p <0.1. Alois Stutzer 41 Data Set: Europe Eurobarometer (EB), Repeated cross-section for 13 European countries 1 67,617 individuals 40,412 in private sector 27,205 in public sector Subjective well-being 1 Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Norway, Portugal, Spain, UK On the whole, are you very satisfied, fairly satisfied, not very satisfied, or not at all satisfied with the life you lead? (4-point) public employment public administration and nationalized industries Unemployment: national unemployment rate from OECD Alois Stutzer 42

22 Results for Europe, Variable Private sector Public sector All Unemployment rate (mean adjusted) (0.012)* (0.019) (0.013) Public sector (0.021) ( * ) Unemployment rate *Public sector (0.005)* Personal Controls Yes Yes Yes Year dummies (6) Y e s Y e s Y e s Country dummies (13) Y e s Y e s Y e s Clustered State*Year State*Year State*Year N 40,412 27,205 67,617 Chi ** ** ** Notes: Source Eurobarometer Dependent variable: Life satisfaction from 1 not at all satisfied to 4 very satisfied. Coefficients of Ordered probit regressions. Robust standard errors in parenthesis. Control variables are gender, age, age square, marital status, education, size of town. Level of significance: * p < 0.05, ** p < Alois Stutzer 43