ABSTRACT. Roed and Zhang use a flexible hazard rate model with unrestricted spell duration and calendar time effects to analyse a dataset including all Norwegian unemployment spells during the 1990s. Starrin and Larsson's study contributes to the understanding of women's unemployment with emphasis on individual reactions in terms of their significance and context, and it was carried out within the methodological tradition of grounded theory. Raaum and Roed remark that labor market conditions at the time and place of potential entry into the labor market are shown to have a substantial and persistent effect on adult employment prospects.