Disability, Health and the Labour Market: Evidence from the Welsh Health Survey

Abstract

This study uses data from the Welsh Health Survey to examine issues of disability/health measurement and its impact on labour market outcomes. The data suggest that self-reported disability, general health and summary composite measures of physical and mental health are broadly consistent both in measurement and predictions for labour market outcomes. This consistency in measurement is also demonstrated at the local level but the employment disadvantage associated with disability is found to vary considerably within Wales.

Acknowledgements

The data were collected by NOP Research Group for the National Assembly for Wales, and were made available through the UK Data Archive. We would like to thank the editors and two anonymous referees for helpful and insightful comments on earlier versions of this paper. None of these organisations or individuals bears any responsibility for the analysis or interpretation of the data.

Notes

¹ The same information on physical and mental functioning as used here is however available in Waves 9 and 14 of the BHPS–see Ludbrook et al. (2003) for a use of Wave 9 data.

² This type of analysis is not possible given the absence of controls for education in the data set used here.

³ The extent of this association is striking since the general health ranking is actually one of 21 items that make up the PCS. More specifically, it is one of five elements in the ‘general health’ cluster or domain that in turn is one of four clusters making up the PCS measure, the others being physical functioning, role-physical and bodily pain. For completeness, the MCS domains are vitality, social functioning, role-emotional and mental health.

⁴ The validity of the taxonomy for the PCS measure is indicated by looking at the proportion classifying themselves as disabled for various (banded) PCS scores. Thus, for those whose PCS is below 30, 93.3 per cent indicate they are disabled (almost 100 per cent for those with a PCS below 20). For those with a PCS in the range 30–40, 63.9 per cent report themselves to be disabled, falling to 31.9 percent when PCS is between 40 and 50 and just 6.0 per cent when PCS exceeds 50. The corresponding figures for MCS are 46.0, 36.1, 22.6 and 14.6 percent, respectively.

⁵ The actual measure of the WIMD is the population-weighted average of the combined scores across the electoral divisions comprising each UA.

⁶ However we are not able to control for the differences in the disabled population among areas. Using data from the LFS, Jones et al. (2006b) show that after controlling for personal characteristics there are no significant differences in the probability of employment for the disabled across areas.

⁷ Using the same dataset as here, Skapinakis et al. (2005) report formal tests of the association between mental health and deprivation, and indicate that regional variation was not explained by controlling for personal characteristics.

⁸Beatty & Fothergill (2005) show these five areas, together with Neath and Port Talbot, all appear in the top 20 districts in Great Britain in terms of sickness claimant rates, a finding they attribute to the ‘diversion’ from unemployment to sickness benefits. Their analysis reveals Wales is second nationally in terms of the extent of this diversion, and, with the exception of Torfaen, all the above areas appear in the top 20 districts for this list also. This should be borne in mind in relation to the discussion above, but is not something that can be addressed further in the analysis presented here.