Work and well-being: introduction.

Bryson, Alex ; Freeman, Richard B.

Work absorbs a huge share of the lives of most people. It is the major source of income. How people fare at work, the conditions under which they work and the mechanisms for remedying problems at work impact subjective well-being, labour market behaviour, and mental and physical health. While economic analysis treats work as creating disutility, many people are workaholics for reasons beyond income, and one of the robust findings from the new economics of happiness is that unemployment is one of the major depressants of subjective well-being. The independent Strategic Review of Health Inequalities, headed by Professor Sir Michael Marmot, noted in its June 8, 2009 policy dialogue that "employment conditions influence health both directly and indirectly" and argued for structural interventions at workplaces, organisations and broader policies to make a significant contribution toward healthier work. As the world economy has fallen into a sharp recession, the increase in joblessness, potential delays of retirement due to reduced value of pensions, particularly in the private sector, and the probable intensification of work among those who remain employed can be expected to have major impacts on well-being and social behaviour. It is therefore timely to devote an issue of the National Institute Economic Review to work and well-being.

Our collection of four papers tackles different aspects of the links between work and well-being that illuminate issues of substantial relevance to policy debate. In contrast to most of the earlier literature, that relates measures of life or job satisfaction as the principal indicator of well-being to employment compared to unemployment, the empirical papers in this issue relate diverse measures of mental health and of satisfaction to the quality of jobs and employment compared to retirement. In contrast to the bulk of earlier literature, that looks at cross-section relations between personal attributes and well-being, they use longitudinal (panel) data as their main source of information, which enables them to eliminate unobserved person fixed effects, and thus to determine better the impact of changed incentives on satisfaction or other outcomes.

Ana Llena-Nozal examines the impacts of changes in a person's workforce status on their mental well-being. She looks at changes in mental well-being when people gain employment and how the change in well-being depends on the nature of the work they gain. Her panel analysis for four advanced industrialised economies (the UK, Australia, Switzerland and Canada) shows that the mental health pay-off to employment depends crucially on the type of employment the individual enters. Mental health benefits for inactive individuals are much greater when they move to jobs with greater job security and when they obtain greater job satisfaction from their job. In some cases, but not all, the gains are smaller for those obtaining non-standard jobs than for those moving into standard employment contracts. But she also finds strong persistence in mental health distress. In addition, the analysis uncovers differences in the impact of types of jobs by gender and country, suggesting that there remains more to do to obtain a full accounting of the job-mental health link. Still, the bottom line message is clear; work of the appropriate kind should be considered as part of the toolbox of 'medical' cures for mental health distress.

The standard analysis of satisfaction and employment compares the well-being of the employed to that of the unemployed. Andrew Clark and Yarine Fawaz undertake a different comparison by examining the change in mental well-being using questions about depressive symptoms, strain, inability to cope, anxiety-based amnesia, and so on, when persons move from employment to retirement. Their panel analysis includes eleven countries from a European panel data set and the UK from the British Household Panel Survey. To the extent that individuals choose their retirement voluntarily, marginal analysis suggests that there will be only modest changes in well-being. In fact, they find that on average there is a modest increase in well-being from retirement but they show that there is a great deal of variation across individuals and according to the jobs individuals held prior to retirement around the average level of change. Some workers gain substantially on leaving work, while others experience substantial falls in well-being on retiring, suggesting that they may have preferred to carry on working. They note that persons having high education and good jobs tend to have higher gains in well-being, which they hypothesize results from having higher-valued pensions. Paralleling Llena-Nozal, they find considerable differences across countries and among regions in the United Kingdom. They stress that the groups whose well-being falls the most upon retirement are those whom policy should encourage to keep working. By extension, those whose well-being rose the most from retirement may arguably be working too long.

Decisions about the division of time between work and household activities should depend on the situation of a household rather than solely on that of an individual. Patrik Hesselius considers the value of time off work relative to working when one's spouse is retired and at home. Are persons with a retired spouse more likely to take sick-leave than persons with a working spouse? There are three reasons to expect workers to value time at home more when their spouse is at home than when both persons work and thus to make greater use of sickness absence than persons whose spouse is working. The employed person may obtain greater pleasure from being at home through synchronisation of leisure with their spouse. They may choose to stay home more when they feel ill because there is someone at home to care for them. And they may seek to stay home to care for their retired spouse. Hesselius uses Swedish panel data to explore how old age retirement of one's spouse affects the working spouse's use of sickness absence. Retirement of the male spouse increases female average long-term sickness absence by about a week per year. When the male spouse retires for reasons of disability, the increase in sickness absence rises to two weeks per year for the employed woman. The response of employed men to the retirement of female spouses is modest and not statistically different from zero, while the response of employed men to the disability retirement of female spouses is about one week per year. Without information about how persons on sick-leave allocate their time at home, one can only speculate about the extent to which spouses take sick-leave to care for their retired spouse, be cared for by their retired spouse, or enjoy leisure together. Whatever the importance of these three possibilities, the message of this paper is clear; policymakers need to give careful consideration to household labour supply decisions when devising sickness and retirement policies and should consider the impacts on the healthcare system as well.

Finally, John Treble points to the big differences in the way in which modern economies provide insurance against sickness. He proposes a framework for understanding the policy choices made in terms of optimal absenteeism and optimal insurance provision. His model suggests that an efficient market solution entails a system in which firms fix a wage, an acceptable level of absence and experience-rated insurance such that workers' ability to supply labour reliably is matched to firms' demands for reliable labour. He notes the use of 'sick-leave banks' in some US and UK firms and points out the problems involved when sick-leave is a tradeable commodity. The paper is timely given the revisions currently underway in the US, Sweden and Germany regarding the design of sick-leave policies.

The new studies of the link between well-being and work are important in two ways. They open a new terrain in labour market analysis that should provide a richer understanding of the way work and labour policies affect lives. Second, they bring labour analysis and medical analysis closer together, which directs attention to the use of firm-based human resource policies as part of health policies, particularly regarding mental health.

Alex Bryson, National Institute of Economic and Social Research and Centre for Economic Performance, London School of Economics. e-mail: a.bryson@niesr.ac.uk.

Richard B. Freeman, Harvard and Centre for Economic Performance, London School of Economics.

NOTE

(1) Strategic Review of Health Inequalities in England Post 2010. A Policy Dialogue on the Social Determinants of Health Work Social Protection, Monday 8 June at the Royal College of Physicians.