Unlike their U.S. counterparts, Canadian unions, governments, corporations and workers appear to be receptive to the debate on work-sharing programs. Canadian workers' desire to change working hours has been highlighted by Statistics Canada surveys. But corporate initiatives and union and government involvement are still in their early stages in both the public and the private sectors.

One factor which emerged from the first Quebec experiments in this area, namely, the importance of voluntary participation in such programs, is instructive for other firms considering taking this route (see Huberman and Lanoie 1999). What would still be most helpful for these firms is information on the profile of workers likely to sign up for these programs.

On the basis of the survey on work schedules and working conditions (1995), Drolet and Morissette (1997) maintain fhat those wishing to reduce their work week are professionals, managers and workers in the natural and social science sectors, have high income and education levels, as well as considerable seniority, hold permanent positions and already work long hours. Interestingly, those wishing to increase their working hours have a very different profile, so that redistribution of hours across the two groups is improbable.

The objective of our research is to establish the profile of workers who accept to reduce their working hours in a work-sharing program. In particular, our findings will enable us to identify similarities and differences between the profile of workers who intendto accept a reduction in work hours, as highlighted by the national surveys, and that of workers who have actuallysigned up for a reduced work hours program. Our study sheds further light on the profile of these workers by including employmentrelated characteristics which the national surveys have not considered.

Our study is based on a survey at the Jonquiere plant of the Alcan aluminium company. This survey was conducted in 1996 on nearly 1300 employees. With these data, we estimated the probability to sign up for the work-sharing program. We considered two groups of independent variables : (1) job characteristics and (2) characteristics of the worker (socio-demographic factors and variables which express attitudes toward leisure).

Our results reveal that, unlike the national surveys, human capital variables (wages and education), marital status, sex and the presence of young children play no role in determining the probability of participating in a work-sharing program. On the other hand, age and seniority have a downward impact on this probability. Furthermore, what is more revealing is the fact that the attitude toward leisure (measured by the habit of banking overtime and the number of days of absence), as well as the characteristics of the jobs held by individuals (compressed timetable, possibility of doing overtime, office work, jobs related to operations and services) are the main factors determining individuals' choices with respect to reduced work hours.