Unemployment and changes in the age composition of the workforce in Britain
This thesis considers the linkages between the age composition of the population and the incidence of unemployment. The first two chapters investigate the macro implications of demographic change, the second two focus on the age-variation in experience of unemployment at the micro level. It is well known both that the probability of being unemployed varies with age and that, thanks to large fluctuations in the birth rate over the last half century, the age composition of the labour force has undergone profound change. We employ a shift-share analysis to identify the role played by shifts in the composition of the labour force in determining the behaviour of the unemployment rate. If workers acquire human capital as they age, demographic change implies a shift in the distribution of skills across the workforce. Drawing on an established theoretical model, we re-examine the evolving mismatch between the demand for, and the supply of, different skills in the labour market, and the role it played in determining the aggregate unemployment rate over the recent past. If we want to understand the extent to which it is the same individuals who are unemployed through time, we have to look beyond the aggregate unemployment rate. We therefore focus on the distribution of unemployment across individuals - and in particular across age groups - when we aggregate across their separate spells, and the implications a concentration of unemployment on a small number of individuals might have for wage setting behaviour. If past experiences of unemployment scar individuals - increasing their probability of being unemployed in the future - this might explain the persistence in experiences of unemployment we observe. Drawing upon the survival analysis literature, we investigate a particular variant of the scarring hypothesis - that past experiences of unemployment significantly reduce the conditional probability of escaping current spells.