Search

This thesis uses data from the Labour Force Survey (LFS) and the Health Survey for England (HSE) to examine the impact of disability on labour market outcomes in the UK. The analysis documents the extent of, and examines the reasons for, the gap in employment and earnings between disability groups. In particular, it attempts to separate the effects of observable differences in characteristics, unobserved productivity differences and discrimination. Unobserved productivity differences are found to be an important influence on employment and earnings. As such, the existing evidence, which ignores this influence, overestimates discrimination against the disabled. Relative to the non-disabled, disabled workers are concentrated in part-time and self- employment. The analysis examines if this concentration is due to marginalisation of the disabled, or if disabled workers have different preferences for non-standard work driven by the need to accommodate disability. The concentration of the disabled in part-time employment is found to be predominately driven by differences in preferences. Amongst males, preferences are also an important explanation for the concentration in self-employment. Estimates of the impact of self-reported disability on labour market outcomes have been criticised due to the potential influence of measurement error and justification bias. The analysis uses more objective health information in the HSE to instrument self-reported disability in a labour market participation model. Self-reported information is found to underestimate the impact of disability, which suggests measurement error is important. The employment provisions in the 1995 Disability Discrimination Act (DDA) were intended to improve the labour market outcomes of the disabled. Data from the LFS indicate that, after controlling for characteristics, the employment gap between the disabled and non-disabled narrowed in the post-DDA period. In contrast, analysis based on a difference in difference procedure and data from the HSE (1991-2004) does not support a positive influence of the legislation.