Do Those Unemployed Longer Work Longer?

The analysis of the relationship of the average durations of employment and unemployment spells across labor force participants over an extended period is used to study the long-run structure of unemployment. Based on this relationship, how the unemployment rates of individuals vary with their average durations of unemployment and employment and with their frequencies of unemployment spells can be found. These relationships are derived graphically and algebraically, and are estimated, controlling for differences in age, sex, labor market experience and degree of seasonality of employment, for a sample of workers experiencing frequent unemployment spells over an extended period in labor force. The sample is drawn from a data base giving the week-by-week labor market experience over an eight-year period (1972-79) for Canadian workers claiming unemployment insurance. With the exception of young inexperienced workers, a positive relationship was found between the average durations of employment and unemployment experienced by these individuals: those who are unemployed longer also tend to work longer. As a result, those with shorter average durations of unemployment tend to have the higher frequencies of unemployment spells and the lower individual unemployment rates. Hence, the counter intuitive relationship arises that those workers with higher turnover rates have the lower individual unemployment rates among this group of frequently unemployed individuals. It is suggested that this observed pattern of employment and unemployment is consistent with many of these workers making voluntary ( if not precise) choices about the proportion of time they want to be employed and unemployed over the longer run given the labor market incentives they face, particularly those provided by the unemployment insurance system.