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

2.1 What is Work Sharing?

As noted above, Work Sharing was introduced to reduce the number of temporary layoffs by sharing the available work among all of the workers of the Work Sharing unit. Participating employees are partially compensated for their lost hours of work with EI benefits.

How does the program work? In a highly simplified format, suppose that a company has 100 workers and it must reduce output by 20 percent temporarily, either because of a recession or other events. To do so, it could lay off 20 workers or reduce the normal working week from 5 to 4 days for all 100 workers. With the Work Sharing program, HRDC encourages companies and their employees to choose the second option.

By choosing Work Sharing, both the companies and employees involved are expected to benefit. Companies retain their skilled employees minimizing hiring and training costs; employees receive EI benefits for the days of lost employment. Specifically, if each worker normally earns $500 per week, participating employees will receive $400 per week from their employer and $55 from EI (55 % of $100). As a result their weekly earnings will be reduced by $45.

2.2 Why Work Sharing?

Q-1: Is the rationale of the Work Sharing program still relevant and why?

Both the occupational and industrial structure of Canada's economy and Canadian labour markets have changed significantly over the last 25 years since the Work Sharing program was first introduced. Nonetheless unemployment remains a continuing economic challenge and policy concern for Canadians. Work Sharing was introduced as a program to help both companies and employees to adjust to both economic demand and supply shocks. This need to lessen unemployment and the human costs associated with unemployment remains just as much a major policy goal today as it was in the past.

However, it has been often noted that thecomposition and nature of unemployment has been changing. For example, recent research in the United States indicates that temporary layoffs as a share of total layoffs is declining⁸. If the same trend held true for Canada, this suggests a reduced need for the Work Sharing program in time. However, as the analysis shows below, the use of the Work Sharing program is very low given the number of job separations in Canada each year and thus there is still considerable room for the program take-up to increase.

The review of labour market policy literature identified potential benefits and costs of policies and programs to reduce layoffs. There was nothing in the existing scientific literature that would suggest that the rationale for averting layoffs is any less relevant today than it was at the program's inception. A discussion of some of these potential costs and benefits would highlight the continuing relevance of the Work Sharing program.

A voluminous scientific and policy literature shows that unemployment is associated with a wide range of lifestyle challenges such as increased stress, aggression, alcoholism and reduced physical activity. The economic literature also shows that the longer a worker is unemployed, the more likely will her job skills deteriorate and the tougher it will be to find new employment. If workers participate in Work Sharing as opposed to being laid off, they will earn a much higher income and also avoid many of the socio-economic and psychological stresses associated with a period of unemployment. It is more likely that they will maintain their job-specific skills than in the case of prolonged layoff.

A potential drawback of the Work Sharing program is that it may give workers a false sense of security. Had the worker been laid off, she might have immediately begun looking for new work or began training courses. However, under Work Sharing, since the worker expects to return to full working hours, she may delay looking for new employment or commencing training programs. Therefore, work sharing programs could lead to a less responsive labour adjustment process. This is part of the reason why the Work Sharing program is mainly intended to be used in the case of temporary layoffs, since in the case of a permanent layoff, the worker should almost certainly begin looking for new work immediately.

For a company, the primary benefit of the Work Sharing program is that it can keep its skilled work force intact. Some workers who are laid off may no longer be available for recall when the firm is ready to return to full output, as they will have found other work. That increases a firm's rehiring and training costs, which tend to increase further if the affected workers are more skilled. Indeed, Siedule et al. found that as the skill level of the work unit increased, the firm was more likely to use work sharing as opposed to layoffs.⁹ The firm may also benefit from increased flexibility when the time comes to increase production. This is because employee hours can often be increased more quickly than workers can be rehired or trained.

Another possible intangible benefit for a company is that with Work Sharing decreasing the likelihood of layoffs, employee morale may well increase. This could eventually translate into productivity improvements for companies.

For Canadian society, the primary benefit of work sharing programs is that they help to reduce unemployment. In turn, this would reduce the negative health and social impacts on society that were noted above, as well as what are likely quite significant costs.

Q-2: Are the activities and outputs of the Work Sharing program consistent with its mandate and can they be plausibly linked with the program objectives and intended effects?

One of the challenges in answering this question arises from the fact that a Results-based Management Accountability Framework (RMAF) has never been written for the Work Sharing program. Therefore, the linkages between activities and outputs to objectives and intended effects have never been clearly delineated and few indicators of program success have been established.

In spite of the lack of a program RMAF, the program documentation clearly indicates that averting temporary layoffs is the primary objective of Work Sharing. Indeed the evidence collected in this evaluation suggests that the program operates to a large extent as intended.

The review of the Work Sharing agreement files indicates that the vast majority of agreements are negotiated and decided for reasons that seem consistent with the mandate of the program. For example, some 80 percent of agreements show that the applying companies cited economic downturns as their reason for requiring Work Sharing. Likewise, consistent with the program's mandate, there was little evidence that firms providing seasonal employment were accessing the program. Indeed there was very little evidence of repeat use of the program by either companies or individual employees. Further, in virtually all Work Sharing files reviewed, the companies indicated that the economic downturn was expected to be temporary in nature and that, given time and effort on their part, they could restore their business.

2.3 How is Work Sharing Implemented in Canada?

When facing a short-term layoff situation, the employers can apply for the Work Sharing program at the local HRDC office level. In order for employers to qualify for the Work Sharing program, they must satisfy the basic eligibility criteria:

• The employer must have been in business in Canada for two years;
• There must be a minimum of two employees in the unit;
• The shortage of work must be beyond the control of the employer;
• The shortage of work cannot be due to seasonal factors; and
• The employer must have the consent of all employees in the Work Sharing unit either through their union, or if there is no union, through the employees themselves.

To participate, the employer must maintain significant work activity over the life of the Work Sharing agreement. The anticipated hours of work for the participating Work Sharing employee must drop at least 20 percent and, at most, 60 percent from their normal levels (roughly one to three working days for a full-time worker). The minimum duration of a Work Sharing agreement is 6 weeks and the maximum duration is 26 weeks. Under exceptionable circumstances, the Work Sharing Agreement can receive an extension to 38 weeks.

To be accepted, the company must provide a "recovery plan" to HRDC, which explains why the work reduction is unavoidable, and what the employer intends to do to return to normal employment levels. The employer must also show why the layoffs are expected to be temporary and not due to recurring seasonal developments. Note that a company does not apply for all of its employees; the company can apply for only a selected group of employees (for example, employees who work in shipping and receiving or employees who are product assemblers).

In turn, for individual employees to be qualified for EI Work Sharing benefits, they must be qualified to receive regular EI benefits. However, unlike the case for regular EI benefits, the two-week waiting period prior to drawing regular benefits is waived. If these workers are laid off following the Work Sharing program, they are still fully entitled to regular EI benefits, and the normal EI rules apply. In such cases, the employees must then serve the two-week waiting period before collecting EI benefits. The amount of time that they are entitled to EI benefits will not be affected by their prior use of the Work Sharing program.

The following program model (or logic model) roughly identifies the major linkages between activities within the Work Sharing program and the achievement of its outcomes and major impacts. The Work Sharing program structure and operation have not changed to any great extent since the program was first introduced, although the actual EI entitlement and benefit calculation rules have changed with changes to the EI system.

The way the Work Sharing program operates is roughly delineated in the program logic model. The logic model outlines the set of program activities, which in turn produce a set of program outputs that lead to a set of intended effects for individuals and the labour market as well as a set of unintended effects. For example, a set of program activities (e.g. development and implementation of Work Sharing agreements) are carried out by HRDC staff. This leads to various program outputs (e.g. the number of Work Sharing agreements and Work Sharing EI benefit claims). These activities and outputs result in intended effects (e.g. reduced layoffs, reduced individual hardship) as well as unintended effects (e.g. reduced labour mobility, shortening of the work week). It is identifying and measuring the intended and unintended effects that any program must be evaluated.

2.4 The Usage of Work Sharing

Exhibit 2.1 presents the basic statistics of the Work Sharing program between 1990-91 and 2001-02. It is apparent that Work Sharing claims as the proportion of all job separations (as measured through Records of Employment received by the EI program) is quite low even during recessions. More specifically, Exhibit 2.1 shows that:

• In an average year, 36,219 individuals started a Work Sharing claim.¹⁰ However, there is considerable variation over time. The number of participants ranged from a low of 7,683 in 1997-98 to a high of 112,357 in 1990-91;
• The corresponding average annual number of layoffs attributed to a shortage of work was about 3 million a year, and approximately one half of these layoffs resulted in an EI claim. This means that the number of Work Sharing claims is only 1.2 % of all layoffs and 2.4 % of all EI claims attributed to a shortage of work — although these rates were three times as high at the start of the 1991 recession.

Exhibit 2.1 also shows a number of interesting comparisons between Work Sharing claims and regular EI claims due to a shortage of work:

• The duration of Work Sharing claims is just over one-half as many weeks as regular claims (11.8 vs. 21.3 weeks). This is consistent with the Labour Force Survey result that the duration of temporary layoffs is about one-half as long as the duration of all layoffs;
• Work Sharing weekly benefits are about one-third of regular weekly benefits ( $87 vs. $258). Work Sharing participants are unemployed only part of the week and, as a result, just receive partial benefits.

Exhibit 2.2, in turn, presents the industrial, regional and demographic characteristics of Work Sharing EI benefits. Most of the percentages below compare individuals with Work Sharing claims as a percentage of all similar individuals with regular EI benefit claims. The majority of participants (67 %) in the Work Sharing program were employed in the manufacturing sector. The incidence of Work Sharing in manufacturing was almost four times as high as in the entire work force (8.4 % vs. 2.3 %). The incidence of Work Sharing was also relatively high in the wholesale trade sector (4.9 %). A possible reason for the heavy concentration of Work Sharing in manufacturing and wholesale trade is that the design of the program targets non-seasonal and relatively skilled permanent type jobs. The workers in these jobs must be substitutable to some extent so that they can share the work through reduced hours.

The incidence of Work Sharing is highest in Ontario (3.6 %), followed by Quebec. These two provinces accounted for almost three quarters of all participants in Work Sharing, which also reflects the concentration of manufacturing in these provinces. The incidence of Work Sharing was lowest in Atlantic Canada (0.7 %).

A number of additional characteristics are also revealed in Exhibit 2.2:

• The incidence of Work Sharing is somewhat higher for men than women (2.6 % vs. 2.0 %). This may well be because the Work Sharing program is most applicable in the manufacturing production-type and distribution-type jobs where men are over-represented. In order for companies to choose Work Sharing as an option to deal with slowdowns, there needs to be a certain amount of substitutability among workers but these workers must have skills that the companies want to preserve. That is to say, the work hours of the work unit need to be interchangeable with the number of workers in the unit. This may be more common in industries where a large number of workers perform similar technically-skilled jobs, such as in manufacturing production. Thus the under-representation of women in manufacturing and thus the Work Sharing program may be due in at least some measure to factors concerning their work functions and occupations.
• Work Sharing use was higher for those with above-average insurable earnings (2.6 % vs. 2.1 %). This may be because the workers participating in Work Sharing programs are retained for their in-house skills and are generally permanent year-round employees. These employees are generally found in the upper range of the industrial wage distribution.
• The incidence of Work Sharing was similar across different age groups, with the exception of individuals aged under 25 (1.7 %). This may be attributed to the fact that younger workers have fewer in-house skills and less experience than their prime-aged (25 to 54) co-workers. Indeed, young workers are usually among the first to be laid off in any production slowdowns.

2.5 Implementation in Other Countries

Q-11: How does Canada's Work Sharing program compare to work sharing policies in other countries?

The Work Sharing program is not unique to Canada. It has also been adopted in different formats in other countries. Canada's version of Work-Sharing is somewhat similar to programs that some American states have (e.g. California, Arizona and Oregon) as part of their unemployment insurance systems, but with noteworthy differences. Where they exist, the details of those state government programs differ from state to state.

Under most, if not all, of the state programs, workers who collect work sharing benefits face decreased benefits, if they subsequently apply for regular insurance benefits (e.g. a dollar for dollar reduction in California). Another difference is that employer insurance premiums in most states, such as California, are experienced-rated — the employer pays a higher premium based on their employees' previous receipt of insurance benefits. In other states, such as Oregon, employers are not required to continue with non-wage benefits (vacation pay, health benefits, etc.) when participating in a Work Sharing program.

Participation in the work sharing programs is comparable between Canada and the United States and relatively low as compared to regular EI claims in both countries.

Several European countries also have Work Sharing programs that are somewhat similar to what is found in Canada. Note that it is important to distinguish between what are truly long term Job Sharing programs in Europe from what would be considered Work Sharing programs in Canada. Many European countries have job sharing programs that are designed to distribute the available work among all employees on a long term basis. Nevertheless, Work Sharing programs in the sense of sharing the available work among the existing employees on a temporary basis are an integral element of labour force adjustment mechanisms in many of these same European countries.

European participation in these programs is much higher than that in Canada. This may be largely due to the fact that labour markets are much more heavily regulated in Europe than in Canada. Countries such as France and Germany have many more government-imposed rules and restrictions on notice and severance pay, as well as government review procedures. These limit the ability of companies to lay off workers, make those layoffs much more costly and make participation in work-sharing schemes a more attractive option when faced with production slowdowns.

Footnotes

8	Erica L. Groshen and Simon Potter, "Has Structural Change Contributed to a Jobless Recovery?" Current Issues in Economics and Finance, Federal Reserve Bank of New York, August 2003.
9	Siedule, Tom, Carole Guest, and Ging Wong, "Economic Activities and the Demand for Work Sharing in Canada," Working Time in Canadian Perspective, Volume 1, Kalamazoo, MI: W.E. Upjohn Institute. 2001
10	A Work Sharing claim is considered to have started in the first week in which benefits were paid.