Inferring disability from post-injury employment duration

The paper examines the extent to which occupational-injury-induced permanent impairment translates into work-related disability. While most permanently impaired workers return to their time-of-accident employer, job and wage, these workers then experience high initial rates of turnover. This turnover can be seen as a manifestation of the dynamics of information, as the time-of-accident employer re-employs the impaired worker, then 'tests' his/her post-injury productivity in order to resolve initial uncertainty regarding the impairment's impact. These turnover patterns can, accordingly, be utilized as a source of disability inference. A statistical model based on the notion of sequential testing is derived and confronted with data reflecting the post-injury labour market experiences of permanently impaired workers.