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

a) Labour Market Outcomes

Employment Rates

Exhibit E.1 presents various employment outcomes for participant and comparison group survey respondents. Overall, these results suggest that some advantage may exist for participant claimants compared to comparison group claimants.

The first panel of Exhibit E.1 shows that a large majority of both the participant and comparison groups indicated that they have had a job since the end of the program or reference date. The incidence of employment at some time in the post-program period was high for all program respondents and ranged from a low of 75 percent of Employment Assistance Services (EAS) participants to a high of 97 percent of SE respondents. For claimants, participants were more likely than comparison group respondents to indicate that they have been employed at some time in the post-program period (86 versus 81 percent respectively), although the opposite was true when we compared reachbacks for these two groups (81 versus 88 percent respectively).

A review of employment rates according to respondents' socio-demographic profiles (not shown) indicates that the younger the participant and the higher his or her education and household income, the more likely he or she was to have found a job following the intervention.

View Exhibit E.1

Employment Stability

Data on two measures of employment stability were collected in the participant and comparison group surveys: the proportion of respondents who have worked for 12 consecutive weeks following the end of the program or reference date, as well as the number of employers they have had since that time.

The second panel of Exhibit E.1 shows that the majority of participants in all programs had worked 12 consecutive weeks since the end of their intervention, with the highest incidence occurring among Self-Employment, Targeted Wage Subsidies (TWS) and Feepayer participants (95, 82 and 82 percent, respectively) and the lowest incidence occurring among Training and EAS participants (69 and 59 percent, respectively). Once again, while participant claimants were more likely than comparison group claimants to have been employed for 12 consecutive weeks (76 versus 64 percent, respectively) the opposite was true of reachbacks (71 versus 76 percent, respectively). The incidence of respondents having worked 12 consecutive weeks since the end of their program also rose with level of education (not shown).

Results for the second measure of employment stability, number of employers since the end of the program or reference date, is presented in the third panel of Exhibit E.1. Although there was little variation between participants in different programs on this measure, SE and TWS participants were somewhat more likely than participants overall to have had only one employer (91 and 85 percent, respectively) and to have had fewer employers on average (1.1 for each versus 1.4 overall). The employment pattern of comparison group claimants seems to have been slightly more stable than that of participant claimants. Comparison group claimants were more likely than their participant counterparts to have had only one employer (74 versus 60 percent, respectively) and to have had a lower mean number of employers (1.1 versus 1.4 respectively).

Interesting sub-group differences based on respondents' socio-demographic profiles (not shown) also revealed that number of employers in the post-program period declined with the respondents' age and the number of dependants, and rose with household income and for single respondents.

Employment Status Outcomes

Participant survey respondents were asked about their employment status at two points in time following their intervention: at one week following the end of the program and at the time of the survey. Overall, these results show a positive shift in employment between these two times for Employment Benefit and Support Measures (EBSM) participants, with the largest positive shifts in employment occurring for full-time year-round jobs. These findings suggest that some employment gains which may be attributed to the EBSMs tended to persist.

A comparison of the first and second panel of Exhibit E.2 reveals that, by the time of the survey, employment rates had risen somewhat among participants relative to their first post-program week (57 versus 50 percent, respectively). Although the incidence of full-time seasonal employment fell by the time of the survey (from 14 to 10 percent), this was more than offset by the incidence of full-time year round employment, which rose from 20 percent in the first post-program week to 26 percent at the time of the survey. Further, the incidence of respondents who were unemployed and looking for work fell between the first post-program week and the time of the survey (37 to 29 percent, respectively). This drop in the unemployment rate is attributable to both increased employment rates for most participant groups, as well as increases in the proportion of respondents in all groups, except SE, indicating that they were out of the labour force. The rates of full-time year-round employment rose with education level and fell with age, and the rates of official unemployment rose with age and fell with education level and with household income.

The only exception to the rise in employment rates among participants occurred for Job Creation Partnerships (JCP) and TWS participants. The decrease in employment among JCP and TWS respondents was due to the loss of full-time seasonal jobs between the first post-program week (25 and 21 percent respectively) and the time of the survey (11 and 14 percent respectively). Another notable shift in the proportion of respondents in different types of jobs occurred among Feepayers, where the proportion of respondents in full-time year-round employment rose considerably from 22 percent in the first post-program week to 40 percent at the time of the survey. This latter finding highlights the fact that, for some programs, a longer gestation period is required before an accurate assessment of program impacts can be made.

Exhibit E.2 also presents the employment status of comparison group respondents. The data show that the overall employment rate among participants was somewhat higher than that of the comparison group at the time of the survey, thus implying some advantage to EBSM participants over non-participants. Participants were also slightly less likely to indicate that they were out of the labour force at the time of the survey, and had slightly lower rates of unemployment (i.e., unemployed and looking for work). Little difference in the proportion of respondents holding different types of jobs at the time of the survey was observed for the participant and comparison groups, although comparison group claimants were much less likely than other groups to be employed full-time year-round at the time of the survey.

View Exhibit E.2

b) Characteristics of Current/Most Recent Job

In this section, we examine the characteristics of respondents' current or most recent job for those respondents who have been employed at some time in the post-intervention period. On average across all programs, we see that only one in five program participants who had a job prior to the program returned to this same job in the post-program period (Exhibit E.3). Training and EAS participants were the most likely to have done so (31 and 29 percent, respectively). Comparison group respondents were much more likely than participants to return to the same job. The incidence of respondents returning to their pre-program job in the post-program period declined with education level.

Panel 2 of Exhibit E.3 shows that little variation existed in the number of hours per week worked by different participants in their post-program job, with the exception of SE participants who reported working an average of 51.4 hours per weeks compared to the 40-43 hour range for the other groups. The average number of hours worked per week was higher among men and declined with education level.

SE and Feepayer participants were most likely to report being employed year-round (75 and 61 percent, respectively) while JCP participants were more likely to report seasonal employment and had the highest rate of casual or contract employment (17 percent). Further, reachbacks were more likely than claimants to report being employed year-round (59 versus 47 percent, respectively for participant; 46 versus 35 percent, respectively for the comparison group). Overall, participants were more likely to be employed year-round than comparison group respondents.

View Exhibit E.3

The highest average weekly earnings were reported by Feepayer, JCP and Training participants ($440, $433, and $419 per week, respectively), and the lowest earnings were reported by TWS and EAS participants ($354 and $333 per week, respectively). Further, comparison group respondents reported higher average weekly earnings than participants. Sub-group differences based on client demographics (not shown) reveal that weekly earnings rose according to education level and household income and were higher among men.

c) Joblessness and Job Search Outcomes

In this section, we present survey results for three post-intervention outcomes: number of weeks jobless (duration), number of weeks looking for work while jobless (intensity), and job search activity. "Jobless" individuals are defined as people who are officially unemployed (i.e., unemployed and looking for work) plus those who are not in the labour force.

Duration of Jobless Spells

The first panel of Exhibit E.4 presents the participant and comparison group survey results for the duration of jobless spells following the intervention, scaled by the time since the intervention or program reference date. On average, participants were not working for 30.9 percent of the time since their intervention. The lowest rates of joblessness were observed for SE (12.6 percent) and Feepayer participants (22.4 percent), and the highest were observed for JCP and EAS participants (44.4 and 39.6 percent, respectively). Participants tended to be unemployed for a smaller proportion of the post-intervention period than the comparison group, and this was true for both claimants (27.2 versus 45.9 percent, respectively) and reachbacks (34.3 versus 38.2 percent, respectively). The contrast in the proportion of time that participant and comparison group respondents were jobless in the post-intervention period suggests that participation in EBSMs conferred an advantage, especially among EI claimants.

View Exhibit E.4

Duration of Job Search

Another potential positive outcome of the EBSMs concerns the number of weeks clients searched for work in the post-program period while jobless. These results were scaled by weeks since the intervention, resulting in a measure of job-search intensity. This also may be considered a measure of interest in finding a job.

As shown in the second panel of Exhibit E.4, participants looked for work for an average of three-quarters (75.4 percent) of the post-intervention period while jobless. Fairly similar proportions (69-81 percent) were observed across EBSMs, apart from the decidedly lower percentage among SE participants (38.6 percent). Participants had higher job-search intensity than the comparison group. Among participants, there was little difference between active EI claimant and reachbacks, but among comparison group member, reachbacks had higher job search intensity than active claimants (67.8 versus 58.6 percent).

d) Utilization of Income Support

Another indicator of the extent to which EBSMs have positively impacted participants is the extent to which participation in the EBSMs has reduced reliance on income support. In this section, we present results regarding participants' post-intervention use of two forms of income support: Social Assistance (SA) and Employment Insurance (EI).

Social Assistance

Only a minority of participants used SA since the end of the intervention (13.1 percent) and on average, SA was used for 7.4 weeks (panels 1 and 2, Exhibit E.5). The lowest rate of post-program SA use was observed for SE respondents (4.1 percent) and the highest occurred among EAS participants (24.5 percent). The highest mean percentage of weeks since the intervention that SA was received was observed for both JCP and EAS participants (13.5 and 13.3 percent, respectively).

Despite only a small minority of participants making use of SA, participants were more likely than non-participants (in particular, comparison group claimants) to make use of SA in the post-program period. The mean proportion of post-program weeks in which SA was collected was also lower among comparison group respondents, and was lower among participant claimants than participant reachbacks (5.9 versus 12 percent, respectively).

Respondents to the surveys who had been on SA were also asked if their employment program had reduced their reliance on income support. For participants, one-third of those who had received SA benefits indicated that their employment program had helped to move them off social assistance and toward employment (panel 3, Exhibit E.5). The proportion of respondents who responded in this way did not vary considerably across different programs, with the exception of JCP participants where fully 52 percent indicated that their employment program had had this effect. Overall, participants were more likely than comparison group respondents to indicate that their program had helped them move to employment and off social assistance.

Employment Insurance

The last two panels of Exhibit E.5 present the incidence and rate of Employment Insurance (EI) use in the post-program/reference date period. Overall, 34 percent of participants indicated that they had collected EI in the post-program period. The highest incidence of EI use occurs among TWS participants (43.9 percent) and the least among SE participants (8.3 percent). The same pattern is observed with respect to the rate of EI use, with TWS participants collecting EI for the highest mean proportion of post-program weeks (19.5 percent) and SE participants collecting EI for the lowest mean proportion (3.7 percent).

Unlike the results pertaining to SA use and consistent with findings from the pan-Canadian study,⁵⁰ EI use among participants and comparison group members implies an advantage for EBSM participants, who were much less likely to use EI in the post-program period. Further, participant reachbacks were somewhat less likely than participant claimants to have used EI in this time (25.4 versus 40.5 percent, respectively). The rate of EI use also shows that participants collected EI for a smaller mean proportion of post-program weeks than comparison group members. Rate of EI use also reveals that reachbacks in both the participant and comparison groups (12.7 and 18.6 percent, respectively) collected EI for a smaller proportion of post-program weeks than claimants in each of these groups (14.5 versus 23.6 percent, respectively).

View Exhibit E.5

e) Summary of Descriptive Analyses

The impacts for different employment outcome indicators can be summarized as follows:

• Employment Outcomes: The survey evidence suggests that participation in EBSMs may confer some modest advantage in terms of obtaining employment. On the other hand, mixed results were observed with respect to employment stability: participants were slightly more likely to have worked for 12 consecutive weeks after the program ended, although there was little difference in the mean number of employers that participants and comparison group members had in the post-program period.
• Employment status outcomes revealed large positive gains for participants in all EBSMs between the end of the intervention and the time of the survey, suggesting that any positive impacts of program participation may grow over time. These positive shifts were larger relative to the comparison group.
• Job Characteristics: Participant wages were lower than those of the comparison group, although few differences existed in the number of hours worked in post-program jobs. JCP and Feepayers reported the highest weekly wages and EAS and TWS participants reported the lowest wages.
• Joblessness: EBSM participants tended to be unemployed for a lower proportion of the post-program period than comparison group members, with the lowest rates of joblessness occurring among SE and Feepayer participants and the highest occurring among JCP and EAS participants. Participants had a higher level of job search intensity (proportion of time looking for work) while jobless than the comparison group. SE participants had by far the lowest job search intensity.
• Utilization of Income Support: EBSM participation has had little impact on participants' use of social assistance, although one-third of participants who had used SA prior to their intervention felt that their employment program had helped them move off SA and toward employment. The survey evidence suggests that there have been positive impacts in terms of lower incidence and duration of use of Employment Insurance compared to non-participants.

Footnotes

50	That study found that a minority of repeat users of EI used EI in the post-intervention period. Authors of this study felt this finding suggested EBSMs had some small impact on reducing reliance on income support.