Is Job Satisfaction U-shaped in Wages?

Size: px
Start display at page:

Download "Is Job Satisfaction U-shaped in Wages?"

Transcription

1 Is Job Satisfaction U-shaped in Wages? Andrew Brown*, Andy Charlwood, Chris Forde*, and David Spencer* * University of Leeds, University of York, UK

2 Background Job satisfaction and well-being at work increasingly important areas of interest: Concern at policy level with promotion of knowledge jobs and better work-life balance Availability of large-scale data sets on different facets of job quality Convergence of research on job quality in social sciences: e.g. labour economists have looked to colonise terrain of other labour fields

3 Main argument Much of the literature takes a subjective utility approach to job satisfaction Economics of happiness (e.g. Layard, Clark) We will argue for an alternative, objective approach to job satisfaction data We uncover a U-shape relation between reported job satisfaction and wages: we argue that this result can only be properly explained using the objective approach We bring out essential policy implications

4 Objective perspective on job satisfaction A. Sen s capabilities approach offers an objective perspective on well-being in general F. Green (2006) has applied this to the case of well-being at work and job satisfaction Well-being at work defined by the capabilities and functionings afforded to job holders What the job enables an individual to do or be E.g. wages, benefits, prospects, security, autonomy, creativity, skill level, effort level

5 Objective perspective on job satisfaction Job satisfaction remains an important facet of well-being at work But in the objective tradition, job satisfaction is not utility gained from a job Instead job satisfaction is internally related to objective job aspects hence itself qualitatively complex

6 Measurement of job satisfaction For objective approach (Green 2006), reported job satisfaction in social surveys (on Likert scale) reflects 2 things: True job satisfaction Norms and expectations of respondents regarding jobs

7 True vs. reported job satisfaction For the same level of true job satisfaction, then: Lower norms and expectations lead to higher reported job satisfaction Higher norms and expectations lead to lower reported job satisfaction Reported job satisfaction may not accurately measure true job satisfaction!

8 Contrast with subjective approach The subjective utility approach also stresses the role of norms and expectations But the subjective approach does not recognise the distinction between reported and true job satisfaction Our empirical work substantiates and illustrates the policy importance of this subtle contrast

9 Empirical work We uncover a robust U-shaped pattern of reported job satisfaction in wages across British employees Based on data from WERS and BHPS Hitherto only seldom remarked upon (or puzzled over) in secondary literature

10 U-shape relation 3.7 Figure 1a: Satisfaction with sense of achievement, Workplace Employment Relations Survey 1998 Satisfaction with sense of achievement (1=very dissatisfied, 5=very satisfied) Wage quintile (1=low est, 5=highest)

11 U-shape relation Figure 1b: Satisfaction with sense of achievement, Workplace Employment Relations Survey Satisfaction with sense of achievement (1=very dissatisfied, 5=very satisfied) Wage quintile (1=low est, 5=highest)

12 U-shape relation Figure 1e: Overall satisfaction, British Household Panel Survey Overall satisfaction (1=not satisfied at all, 7=completely satisfied) Wage quintile (1=lowest, 5=highest)

13 U-shape relation 5.36 Figure 1f: Overall satisfaction, British Household Panel Survey 2004 Overall satisfaction (1=not at all satisfied, 7=completely satisfied) Wage quintile (1=lowest, 5=highest)

14 U-shape relation Figure 1g: Satisfaction with work itself, British Household Panel Survey Wage quintile (1=lowest, 5=highest)

15 U-shape relation Figure 1h: Satisfaction with work itself, British Household Panel Survey 2004 Satisfaction with work itself (1=not at all satisfied, 7=completely satisfied) Wage quintile (1=lowest, 5=highest)

16 U-shapes in WERS2004 Basic regression (P-values in parenthesis) Full regression with controls (P-values in parenthesis) Satisfaction facet Coefficient on log wage squared term Coefficient on log wage Coefficient on log wage squared term Coefficient on log wage term Sense of achievement ** (0.017) * (0.076) *** (0.001) *** (0.001) Influence *** *** ** ** (0.039) (0.031) Pay *** *** ** (0.02) (0.752) Initiative *** *** * * (0.053) (0.07) Training *** *** ** (0.02) (0.674) Security *** *** * (0.068) (0.123) Work itself *** ** ** ** (0.006) (0.02) (0.014) (0.031) Composite satisfaction *** *** *** (0.001) ** (0.028)

17 A general explanation for the U- shape We suggest explanation in terms of norms and expectations Low earnings: low norms and expectations matched by actual job Medium level earnings: High norms and expectations not matched by job High earnings: High norms and expectations matched by job Given this explanation, only top third of earners have high job expectations which are actually met by their jobs

18 Subjective interpretation A utility interpretation of the general explanation of the U-shape relation would be that those in low paid jobs are truly satisfied in their jobs, just as are those in high paid jobs: Low norms and expectations of low-paid cause high job satisfaction But this has potentially counterintuitive policy implications: it suggests that low paid jobs are truly conducive to high employee job satisfaction and therefore that these jobs should be encouraged!

19 Objective interpretation For the objective approach, job satisfaction is truly high only in the high earnings category: Low paid jobs are poor quality on many objective dimensions (not least pay itself) Poor current and prospective objective job quality leads to low norms and expectations of the low paid Such low norms and expectations conceal from surveys the low job satisfaction of low earners Disjuncture between reported and true job satisfaction for low earners

20 Qualitative evidence Existing qualitative research supports objective interpretation: Low paid tend to have both objectively low job quality and low norms and expectations (Lloyd et al. 2008, Anderson et al. 2006, Commission on Vulnerable Employment 2008) Low norms and expectations explain high reported job satisfaction of low paid (Tomlinson 2005, Walters 2005, MacKenzie and Forde 2009) Low paid satisficing as opposed to being truly fulfilled in their jobs

21 Conclusion So, is job satisfaction U-shaped in wages in Britain? The robust statistical U-shape is not decisive Answer depends on concept of job satisfaction and its measurement: YES, if true and reported job satisfaction are equivalent (subjective approach) NO, if there is disjuncture between reported and true job satisfaction (objective approach)

22 Conclusion Conception of job satisfaction matters in terms of interpretation and understanding of job satisfaction data A subjective approach may conceal from social surveys the extent of low job quality in Britain, due to the chronically low norms and expectations of the low paid An objective approach is essential in order to guide policy towards decent work

23 Further work More qualitative evidence required to develop and substantiate our argument Need to develop appropriate survey questions to discriminate between satisficing and true job satisfaction