A Picture of the Work and Well-Being of Working Age Carers in Wales : Evidence from the Welsh Health Survey

A Picture of the Work and Well-Being of Working Age Carers in Wales: Evidence from the Welsh Health Survey


Introduction
This paper considers the implications for the Welsh labour market of carers: those individuals who are restricted as a result of caring for someone who is limited by age or disability.Relative to other regions in the UK it is well established that Wales has a higher rate of working-age disability (Jones et al., 2006).Whilst the reasons for this are difficult to distinguish, Wales appears to compare unfavourably in terms of the distribution of industry, the age structure of the population, average levels of education and the prevalence of low income, which are all identified by Smith and Twomey (2002) as important influences on regional variations in disability. 1 Moreover, Wales also has a higher concentration of the population above standard retirement age (20.6% compared with 18.7% for the UK as a whole in 2005 2 ).Therefore the need for care may be anticipated to be greater in Wales, and it is likely that part of this demand will be met by informal care.Indeed, Maher and Green (2004) and Young et al. (2006) both report that Wales is close to the top of the UK regional league table of caring prevalence, and as such this phenomenon is likely to be especially important in terms of its impact on the Welsh labour market.The evidence also suggests carers are especially vulnerable (Heitmueller, 2007), so that understanding their wider situation is an endeavour of some consequence.In particular, we consider the health and well-being of carers in Wales.

Informal care
The adverse effect caring may have on an individual's labour market opportunities has only been recognised by economists very recently in the UK.Heitmueller and Inglis (2004) find that informal carers represent about 14% of the working age population and their labour market participation rates are 8% lower than non-carers, while increasing the number of caring hours is associated with lower carer participation rates. 3This last result confirms the earlier work of Carmichael and Charles (1998) who found that those caring for more then 20 hours each week were less likely to work than non-carers, but that those whose caring involved a more modest time commitment were more likely to do, albeit they worked fewer hours.Carmichael and Charles (2003) suggest there is a gender dimension to the care-giving phenomenon, with female participation rates being more sensitive to caring than otherwise similar males.The effect of informal care is not however, restricted to participation, and carers experience an hourly earnings gap relative to noncarers after controlling for personal characteristics (Carmichael and Charles, 2003;Heitmueller and Inglis, 2004).Importantly, Heitmueller (2007) considers the potential influence of endogeneity between work and care and finds that failure to control for this effect may result in underestimation of the impact of care on the participation decision. 4

The Welsh Health Survey
The 1998 Welsh Health Survey (WHS) contains information from individual self-completion questionnaires for almost 30,000 adults aged 18 and over in Wales (National Assembly for Wales, 2000). 5This cross-sectional survey contains extensive self-assessed information on health and, importantly, is able to identify carers.A carer is defined in this survey as "someone whose life is restricted in some way because they look after a person who is mentally or physically disabled, or who is limited in what they can do by illness or old age".The labour market analysis of this group to date has been modest, but this survey can be used to provide some additional, quantitative evidence on the indirect effect of ill health on the labour market through the channel of caring. 6It is important to note that the analysis here focuses on those of working age.A weighting scheme is applied to render figures representative of the population.

Caring and labour market outcomes
Table 1 details some basic information about both the incidence of caring and the time spent by carers.As can be seen, slightly fewer than 7% of the working age population act in this capacity in the present survey.This figure, which includes caring for the elderly (and not just the disabled) is lower than for example in both Young et al. (2006) and Maher and Green (2004) Heitmueller and Inglis (2004), Young et al., (2006) and Heitmueller (2007), employment rates are lower for those acting as carers, both for men and women.However, this is the consequence of low employment rates amongst those where the intensity of caring is greater than 20 hours each week; in this case the employment rate drops to (just) 46%.Nonetheless, around 55% of carers manage to combine work with caregiving.Among those who do work, 31% of carers work part-time compared to 18% of non-carers, suggesting that for many, this type of work is chosen so as to allow them to (continue to) work.Part-time work has previously been found to be an important 'accommodating device' among the disabled themselves (Jones, 2008), and the WHS suggests this is mirrored for carers.A similar logic has also been proposed for the higher incidence of self-employment among the disabled (Jones and Latreille, 2006), and this too receives some support from the data.

The well-being of carers
An important issue in respect of carers is the impact that such activity might have on health and well-being, and this is examined in typically better among those in work compared to those not in employment, reflecting the strong causal link between health and work capacity.However, controlling for employment status, the health of carers is consistently worse than that of non-carers: almost twothirds of non-carers in employment describe their health as excellent or very good compared with just over half of carers, while at the opposite end of the health scale, carers are more likely to report their health is only fair or poor compared with non-carers. 9Carers are also substantially more likely to be disabled or have a limiting long-term illness themselves, and many thus combine the caregiving function with the management of their own impairment(s).Not surprisingly, carers are also more likely to report feeling 'worn out' all or most of the time.Oswald, 1994;Theodossiou, 1998;and the wider psychology literature reviewed by, inter alia, Bartley, 1994;Owen and Watson, 1995;Murphy and Athanasou, 1999)  10 .
Strikingly, carers are substantially more likely to report feeling sad/depressed than their noncaring equivalents.
A similar pattern emerges when considering other, more short-term, self-reported measures of mental wellbeing (i.e.relating to the last 4 weeks) such as feeling nervous; downhearted and low; inconsolably 'down in the dumps'; and happy.These and the questions in the previous paragraph are in fact part of a battery of so-called 'SF-36 questions' which can be combined to produce physical and mental summary health measures (see http://www.sf-36.orgfor details).These are widely accepted and used by health economists and epidemiologists, have minimum and maximum bounds of 0 and 100 respectively, and mean values of 50.The lower psychological well-being of carers is revealed by comparing mean Mental Component Summary (MCS) scores among the groups, which conform to the same pattern described above, and thus confirm the adverse impact of caring on carers.

Intra-regional variation
Whereas Young et al. (2006) and Maher and Green (2004) show that the incidence of caring varies among UK regions, it is interesting to note that variation also exists within Wales.Thus, 8% of the working age population act as caregivers in the Neath and Port Talbot UA, compared with just 5.25% in Torfaen.The relationship between disability and caring at the UA level is shown graphically in Figure 1.This reveals there to be a positive correlation between the incidence of carers in the working age population and the disabled population by UA (r=0.461,p-value=0.031),but also substantial variability, with some areas having relatively high proportions of carers and low disability rates (for example, Powys) and vice versa (for example, Torfaen).This may be expected given the multitude of factors affecting the care decision including the closeness of family ties, availability of formal care and the economic opportunities available to individuals in the area.However, the correlation between the 2000 Welsh Index of Multiple Deprivation 11 (WIMD) and the proportion of working age carers however, at 0.427 indicates that informal care is significantly more prevalent in more disadvantaged areas 12 .Whereas previous work has highlighted the role of ill health as a driver of low rates of economic activity in some of the most disadvantaged areas in Wales (Blackaby et al., 2004), it is evident too that policymakers need to recognise the associated implications in terms of caring and the limitations these may imply.Indeed, this is likely to be associated with the acute incidence of workless households identified as a particular cause of concern in some of these areas (Blackaby et al., 2004)   for more than 20 hours each week.

Conclusion
Caring at this intensity is found to be negatively associated with labour market participation, consistent with the findings from the small existing literature.Working age carers are also found to experience worse health outcomes than non-carers, and in particular in relation to mental health/well-being.Both of these suggest that informal care imposes significant costs on those who perform such a role.Moreover, intra-regional examination indicates the incidence of caring is generally higher in more disadvantaged areas.It is in these areas where the impact of caring on labour market outcomes and well-being may be particularly acute.
Given the contribution made by carers to the UK economy (estimated at £22 billion in 1999 (Laing andBuisson, 2003, cited in Heitmueller andInglis, 2004)), together with the wider social impact of caring, it would seem there is a strong case for additional policy measures to support carers.The Government has recently established an 80% employment target (see next feature article by Victoria Winkler) and much of the focus of policy has been on encouraging the participation of individuals on incapacity benefit.Clearly however, progress towards this target could also be enhanced by facilitating labour market participation by carers.Flexibility in employment appears vital, and this has been recognised by giving carers the legal right to be able to request flexible working under the Work and Families Act 2006.Moreover, the provision of reliable and affordable support services to aid informal carers is fundamental in creating the opportunity to combine work with care and improving the well-being of carers.Given the population is ageing and the demand for care is accordingly likely to increase, recognising and responding to the work and welfare implications of informal care is likely to be even more important for the future.
Latreille, School of Business and Economics, Richard Price Building, University of Wales Swansea, Singleton Park, Swansea, SA2 8PP, UK.Tel: +44 (0)1792 295168.Fax: +44 (0)1792 295872.E-mail: p.l.latreille@swansea.ac.uk 1 Senior (1998) using the 1991 Census, found observable factors such as working environment, lifestyle, deprivation and area demographics under-predicted the observed level of illness and hence argues cultural factors may also be an explanation for the higher rates of disability in Wales.Carmichael, F and Charles, S., (2003)

2
Authors' calculations based on Office for National Statistics data in

Table 3 .
The interrelationships among caring, health/well-

A Picture of the Work and Well-Being of Working Age Carers in Wales: Evidence from the Welsh Health Survey Melanie
K. Jones and Paul L. Latreille * School of Business and Economics, University of Wales Swansea being and employment are complex, and Table 3 accordingly reports separate figures in respect of a number of measures of health/well-being by carer and employment status.In the existing literature, Doran et al. (2003) identify a concentration of poor health among carers, with only 56% of carers (not restricted to working age) self-reporting good health compared to 70% of noncarers.A similar ranking is evident in the WHS.As Table 3 reveals, health is

Table 2 : Employment Outcomes By Carer Status Incidence
Source: Welsh Health Survey, 1998.Working age population only.Data are weighted.

Working Age Carers And Total Disability By Unitary Authority Carers as proportion of working age population Disabled as proportion of total population
It should be noted that this is not explicable in terms of the fact that carers are typically older: within each of the age bands in Table 1, carers typically report worse health than non-carers.10 Note however, that causality can run in both directions depending on the characteristics of work itself (e.g.stress).The relationship is also confounded by factors such as age, gender, education, location, social capital/networks, etc.That unemployment impacts adversely on various measures of mental wellbeing, including parasuicide however, is widely accepted.Journal of Epidemiology and Community Health, Vol.48, pp 333-337.Blackaby, D., Latreille, P., Murphy, P., O'Leary, N. and Sloane, P., (2004) Identifying Barriers to Economic Activity in Wales Part II: A Survey of the Economically Inactive in Three Areas of Special Interest, report for Welsh Assembly Government.Carers UK (Policy Brief).Carers UK, (2005) Facts About Carers, Carers UK (Policy Brief).Carmichael, F. and Charles, S., (1998) The Labour Market Costs of Community Care, Journal of Health Economics, Vol.17, pp 645-795. 9