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

1.1 Background

Over the past 15 years, the pace of economic restructuring in Canada has quickened. While this has created opportunities for some workers, it has increased the vulnerability of others. Older workers more often fall into the latter group. Compared to displaced workers in other age categories, those 55 years and over tend to have longer durations of unemployment that frequently terminate not with employment but, rather, with an exit from the labour force. For those who do find employment, the average wage loss (relative to the previous job) is larger than for other age groups. Labour adjustment difficulties are particularly acute for workers who have lost jobs in declining sectors in depressed local labour markets.

The Program for Older Worker Adjustment (POWA) was introduced in 1987 to address some of the difficulties faced by older workers who have been laid off from their jobs. POWA is intended to help displaced older workers with poor reemployment prospects by providing compensation to bridge the period between layoff and retirement.

The overall purpose of this study is to evaluate the impacts of POWA on older workers and, in particular, to compare these impacts with those of older workers who did not participate in the program. Evaluation objectives include:

• to provide a profile of program participants and non-participants;
• to assess the underlying program rationale;
• to examine POWA delivery results;
• to determine the labour market impacts of POWA; and
• to determine the non-labour market impacts of POWA.

1.2 Research Issues

The Terms of Reference for the study identified four broad issue areas for this evaluation. These are:

1. Rationale
2. Program Design and Objectives Achievement
3. Impacts
4. Alternatives

Each of the research issue areas is discussed in more detail below.

(a) Rationale

Rationale issues refer to questions about the relevance of POWA and the overall logic of the program. Is the original rationale for implementing the program still valid? Is there a continued need for the program? Are the program's resources appropriately targeted? The specific rationale issues for the POWA evaluation are:

• Under what conditions is there a need for the government to provide income assistance (annuity) for older workers laid off at pre-retirement age (55-64)?
• older workers, who have lost their jobs at pre-retirement age (55-64) as a result of a major layoff, live independent of government support until they reach the age of 65?

(b) Program Design and Objectives Achievement

Program design issues refer to questions of program operation and the extent to which the parameters of the program affect the achievement of the objectives for which the program is accountable. Evaluation questions under program design and objectives achievement are:

• What are the main objectives of POWA? What are POWA's sub-objectives (e.g., temporary income support, last resort income support for long-service workers involved in a major layoff, labour market adjustment, etc.)?
• Has POWA enabled recipients to live without accessing other types of government income support? Have recipients been adequately supported through POWA?
• Are the criteria for qualifying as a major layoff appropriate in terms of helping those most in need? Do the criteria exclude certain individuals who may need assistance? Does POWA provide assistance to certain individuals who do not require assistance?
• How have POWA's criteria and benefit levels changed over time and what are the implications of these changes to the individual, the industries, and the community?

(c) Impacts and Effects

Impacts and effects are the more general consequences of the program These include both labour market or employment outcomes and non-labour market outcomes. The latter category of impacts and effects will be particularly important in the current context given the objectives of the program and the client group. The evaluation issues identified in the Terms of Reference under impacts include:

• To what extent do POWA recipients differ from the general population of workers who have lost their jobs in pre-retirement age in terms of labour force activities?
• How do POWA recipients continue to contribute to the community, both financially and non-financially?
• What is the current living situation of POWA recipients and to what extent are they satisfied with their current living situation, including their standard of living, their leisure activities, their prospects/expectations, and their labour force activities?
• To what extent has POWA enabled recipients to attain a level of self-sufficiency through employment or other means of non-government income (e.g., pensions, savings, investments, etc.)?
• Do POWA recipients stay in the communities where they worked after POWA benefits begin?

(d) Alternatives

Alternative issues reflect the forward-looking aspect of the evaluation. The issue of alternatives asks whether there are more cost-effective alternatives to the current configuration of the program that would yield the same or better results. Specifically, alternative issues for this evaluation include:

• Are there other cost-effective alternative programs or strategies that might achieve the same objectives as POWA?
• What is the relative cost-effectiveness and cost-efficiency of POWA compared to other types of income support?
• To what extent is it cost-efficient and/or cost-effective to deliver POWA benefits via the private sector (i.e., annuity) rather than through traditional government mechanisms?

Not all of the cost-effectiveness issues could be addressed in this evaluation. Reliable comprehensive cost information was not available from either HRDC's administrative data or from external sources. Annual reports from the insurance companies and the POWA payment file were the two sources of data available for the evaluation of the costs associated with the Program. Based on the annual reports from the insurance companies, it was virtually impossible virtually impossible to determine the real costs a particular annuity plan since unspent amounts could be applied to new annuity contracts. The second source of data (the POWA payment file) was only available from one company and only contained the total payments for an individual in a given year and not the total costs associated with that person since the beginning of the annuity. Given the lack of cost information, more effort was devoted to other research activities such as developing a detailed profile of the participants and the comparison group using the administrative data.

1.3 Evaluation Methodology

The methodology for this evaluation was based on multiple lines of evidence. Initial preparatory research included:

• literature and documents review. This review involved an examination of the literature on the displacement experiences of older workers and the international experience with respect to adjustment programs for older workers. A thorough review of program documentation was also conducted. A description of POWA, including a program logic model, was developed.
• key informant interviews. Several key informant interviews were conducted with program staff at headquarters. These interviews provided an important program context for the evaluation and information on administrative data sources.
• pre-survey focus groups. During the research design phase, two focus groups were conducted with POWA participants. The findings from these groups were used in the design of the survey questionnaires.

The two most important data sources for the evaluation were surveys of POWA participants, non-certified participants and a comparison group of non-participants, and an analysis of administrative data. These methods are described in more detail below.

(a) Survey of Participants

For the purposes of this study, POWA participants were defined to include those who received benefits from the POWA program (i.e., were certified) and were laid off between 1989 and 1993¹. In total, 1,050 interviews were completed with a random sample of the approximately 10,000 POWA participants who were laid off during these years.

(b) Survey of Non-Certified Participants

A small survey was conducted of individuals who were eligible for POWA (i.e., they appeared on the POWA administrative database) but did not participate in the program. Three categories of non-certified participants were surveyed: 1) 150 interviews were conducted with "no shows", that is, individuals who were invited to attend a POWA information session but did not show up because they were ineligible, had found other employment or were receiving relatively high pension benefits; 2) 100 interviews were conducted with individuals in the "zero benefit" category which includes those who met the program eligibility criteria but who, after deductions for other income sources, had their POWA benefits reduced to zero; and 3) 51 interviews were conducted with individuals who had withdrawn from the program.

(c) Comparison Group Survey

In order to ensure the comparability of the non-participant sample, a comparison group sample was created in three steps. The first step involved identifying mass layoff firms with characteristics similar to firms participating in POWA during the period under study. Firms that had experienced permanent layoffs of 20 employees or more in any calendar year between 1989 and 1993 were included in the initial sample. Second, those firms who satisfied the initial criteria were further screened based on community size. The POWA program determines firm eligibility based on size of layoff in relation to size of the community. Firms in the comparison group that did not meet the layoff size and community size criteria simultaneously were excluded². Third, laid-off workers from eligible comparison group firms were selected. These workers had to be between 55 and 64 years of age at the time of layoff. To ensure their comparability to the POWA participant group, the telephone survey also contained a number of screening items based on program eligibility criteria. These items excluded respondents who were not Canadian citizens or landed immigrants, had worked fewer than 13 years in the 20 years prior to their layoff and whose layoff was temporary.

(d) Administrative Data Analysis

Three types of administrative data were analyzed for this evaluation study. The administrative data analysis began, first, with an examination of the full POWA applicant database. This database included information on over 30,000 applicants to the program who were laid off between 1986 and 1995. The analysis was used to profile POWA participants and to examine program take-up.

Second, administrative data linkages were established between HRDC administrative data files, the POWA applicant database and the comparison group respondents to the telephone survey³. HRDC administrative data files, specifically the Status Vector, Record of Employment and T1 files were used to provide important historical and outcome information for POWA participants and the comparison group, including, for example, utilization of unemployment insurance and total income/earnings⁴. Finally, a POWA payments file was examined in the administrative data analysis. This analysis provided information on program benefits and the impact of changes introduced to the program on the benefits structure.

1.4 Report Organization

This report contains five additional chapters. Chapter two provides a brief description of POWA. A profile of POWA participants is presented in Chapter three. Chapter four describes program delivery results. Final outcomes of the program are provided in Chapter five. Conclusions are presented in Chapter six.

---

Footnotes

1	There were two groups of respondents in this survey. The majority of POWA participants (950) surveyed were selected from the population of participants having a geographical match to layoffs in the comparison group. A second smaller sub-group of cases (150) included participants from layoffs without a geographically matched layoff in the comparison group.
2	Due to vagaries in the estimation of community size, sub-sample of small layoff size cases was included to moderate their underrepresentation in the core comparison group sample.
3	A potentially broader comparison group of older layoff workers available from the administrative data was rejected for the analysis because there was no information to ensure the comparability to POWA participants (e.g., employment history, permanent layoff). As a result, administrative data for those in the telephone survey only were used.
4	The Status Vector file includes information on utilization of unemployment insurance. Record of Employment file has data on employment experience and job separations. The T1 file is based on taxation information.