Directorates
·	Director General's Office
·	Data Development
·	Strategic Evaluations
·	Partnership and Continuous Evaluations
Publications
·	Evaluation Reports
·	Evaluation Briefs
·	Labour Market Development Agreement Series
·	Lessons Learned Series
·	Evaluation Tool Kit Series
·	Evaluation Forum Newsletter
·	From Theory To Practice

The empirical analysis presented in this paper provides evidence that the inclusion of jobs entailing fewer than 15 hours per week and the move to an hours-based EI system has evoked a significant change in the hours of work distribution. When the raw data are presented so as to difference out both the effect of the redesign of the Labour Force Survey and seasonal effects, a clear tendency away from very short-hours jobs is apparent. Except in seasonal industries, there has been a substitution towards jobs that require fewer than 35-40 hours per week. Most of the change in seasonal industries is away from jobs requiring fewer than 40 hours per week to jobs that require more than 40 hours per week. Parameterizing the EI program changes into a hazard function model generated estimates that attributed virtually all of the changes in the hours distribution in the first half of 1997 to a behavioural response to the incentives created by the EI program. Further econometric analysis provided formal evidence supporting the evidence from the raw data that the pattern of changes is different in stable and unstable sectors in a way that is consistent with the behavioural interpretation.

These findings have several implications:

1. Evidence that the share of jobs that require fewer than 15 hours per week is declining suggests that removing the distinction between weeks of work involving more or fewer than fifteen hours has eliminated a significant distortion in some employers' hours decisions.

2. When workers adjust their working schedules in these ways, they mitigate some of the adversity that a reduction in program eligibility and benefit entitlement would otherwise entail.

3. Evidence of a trend to very long hours jobs in seasonal industries may reflect a reduction in those industries in work-sharing arrangements that had in the past been designed by workers and firms to qualify a large number of individuals for UI benefits.

4. Evidence that the share of 15-30 hours per week jobs in unstable sectors is declining relative to longer hours jobs suggests that a new distortion has been introduced into some hours decisions. This new distortion arises because under the old weeks-based system, benefit rates were based on earnings, but eligibility and entitlement rules were neutral with respect to hours worked per week, so long as at least 15 hours were reported. Under the new hours-based system, a week of part-time work is not insured as fully as a week of full-time work. The implications of this new distortion should be considered by researchers and policy analysts interested in the broader economic and social implications of hours of work decisions.