Essays on Labour Supply and Health of Older Workers

The thesis studies three issues: how a change in the expected social pension affects older workers’ labour supply decisions under a life-cycle framework; what are the mental and physical health consequences of formal retirement; and how health deterioration affects labour supply transitions of older informal workers. Previous studies on health and labour market transitions focus on older formal workers from developed countries. This thesis adds to the current literature by studying informal workers in China who tend to work until much older ages or until they are physically incapable. They suffer from low levels of income and savings, and limited health insurance coverage for old-age support. Data from the China Health and Retirement Studies are used for all three chapters. The first chapter, ‘Social Pension and Labor Supply Responses of Older Informal Workers’, utilises the community-level difference in the years of introducing the New Rural Social Pension (NRSP) programme in rural China to identify the effect of individual NRSP participation on their probability of working and actual hours of working. We use a trivariate, non-linear random-effects model and adopt life-cycle framework to guide the empirical specification that focuses on heterogeneous responses from participants in different age groups. We find that male participants overall do not change their working probability, while female participants are less likely to work if they are above 50, although the effects are only marginally significant. For participants staying in the labour force, male workers who are below the pension eligible age of 60 significantly increase their weekly working hours by 5.86% to 7.67%, with the effect mainly coming from non-agricultural workers. The New Rural Social Pension (NRSP) programme is both a basic social security programme that pays monthly basic benefits to age- eligible participants, and a retirement pension programme for age-ineligible participants who are not covered by employees’ pension programme. Given that the NRSP is one of the largest social security programme around the world in the terms of enrollees and financial input, evaluating how effective it has been in changing individual behaviours and welfare is important and provides policy implications on developing countries that are reforming or expanding their non-contributory public transfer programmes to cover the larger population of informal workers. The second chapter, ‘Gender Difference in the Retirement Effect on Cognitive Functioning and Depression Risk’, utilises the different compulsory retirement ages for blue collar and white-collar workers to identify both the short-run effect of transitions into retirement and the cumulative effect of retirement years on formal retirees’ mental health. We estimate both dynamic and non-dynamic panel data IV models. We focus on the bivariate random effects IV model and find that transition into formal retirement increases men’s scores in cognitive tests by 30-50%, but reduces women’s scores in the mental intactness test by 52.3%. The positive short-run effects of retirement disappear with the number of years in retirement for men and the negative effect reverses for women. There is no significant retirement effect on either men’s or women’s depressive risk or physical well-being. The third chapter, ‘The Dynamic Relationship between Depression, Physical Health, Labour Market Exits and Entries of Older Informal Workers’ studies the sequential causality between mental and physical health and labour market transitions of older informal workers, while also identifies the roles of state dependence, observed SES variables, and unobserved individual heterogeneity in explaining health and labour supply outcomes. Using a trivariate dynamic cross-effects model, we find that depression increases the probability of male nonworkers re-entering the labour force, mainly by taking agricultural works, in the subsequent period. Physical health deterioration predicts a higher risk of depression in the subsequent period, larger for men than for women. Depression does not have any effect on men’s reported number of physical problems in the subsequent period though increases women’s, supporting an asymmetric effect between physical health and mental health. Physical health declines predict an exit from the labour force, with sectoral and gender-specific difference. There is strong and significant statedependence for the number of physical health problems and labour market status, but not for the risk of depression, similarly for men and women. The paper provides empirical evidence that adding mental health into studying the relationship between health and labour supply transitions can better explain the variation in health effects and the mechanisms behind the labour supply adjustments after health shocks.