|
|
     |
||
 |
     |
 |
|
The purpose of this chapter is to lay a solid foundation for the econometric analysis to follow in the next chapter. The descriptive findings on outcome15 are more intuitive than the much more complex econometric analysis; a good understanding of the basic outcomes presented in this chapter will help the reader understand the econometric analysis. A note of caution is in order, though: differences between participants and non-participants cannot be attributed to Employment Benefits and Support Measures (EBSM) until the econometric models control for outside influences. This modeling will be presented in the next chapter. 8.1 Key OutcomesThis section examines the major outcomes of interest to labour market programs such as EBSM: time spent working, time spent in school and unemployed; earnings; use of income support; and educational achievement. Time Spent Working, In School and UnemployedFor the longitudinal econometric analysis to follow, it is necessary to establish preprogram and post-program labour market activity. The survey asked participants and non-participants to account for their time spent working, unemployed, and in school since 1995. The results are portrayed in Table 8.1. Before EBSM, non-participants spent more time than participants working and less time in school. Pre-program time spent unemployed (and not in school) was somewhat higher for participants. For participants, time spent unemployed held steady over the four years. Non-participants showed a similar pattern although time spent unemployed was somewhat higher in 1997 and 1998 than it had been earlier. During 1997, when the survey sample was participating in EBSM, participants' time spent in school increased markedly. Participants were significantly less likely to be employed than non-participants before the program. But this wasn't because they were more apt to be unemployed; rather they were more likely to be in school.
The next chapter will isolate the impact of the program, but for now the influence of being in the program can be removed by looking at how those who finished participating in EBSM in 1997 spent their time in 1998. Chart 8.1 shows that participants out of the program in 1997 spent about 31 percent of 1998 unemployed, 64 percent working, 3 percent in school and 2 percent in school and working. Non-participants spent 32 percent of 1998 unemployed, 65 percent employed, 1 percent in school and 2 percent working and in school. The only statistically significant difference between the groups was time spent in school.16 
EarningsTable 8.2 shows the average annual earnings of participants and non-participants from 1992 to 1998. Average pre-program earnings were low, never reaching $20,000 for any subgroup in any year (note that those with zero income are excluded). By component, Job Creation Partnerships clients had consistently earned the least. Until 1996, Self-Employment clients earned the most. Non-participants consistently earned more than participants. The familiar "Ashenfelter Dip," where earnings fall just before participation in a labour market program, is in evidence for participants in every component, especially Self-Employment. By 1997, Self-Employment clients, most of whom were attempting to get their business going, had the lowest earnings. Participants reported mean gross earnings of $17,599 from all jobs in 1998; non-participants earned $19,572, a difference of $1,973. The difference did not quite reach statistical significance, however.17 The jump for participants' gross earnings from 1997 to 1998 is certainly encouraging: it was over twice as large as the rise for the comparison group.18 From Table 8.1 it is clear that participants worked only 6 percent more time in 1998 than in 1997 on average, so most of the increment must be attributable to better-paying jobs. It is possible however that some of the earnings gains are subsidized since many were still participating in 1998. But if the analysis is limited to EBSM clients who completed their intervention before 1998, the picture looks better: mean 1998 gross earned income for participants was $20,241, $669 more than non-participants. This difference however falls short of significance.19 In 1998, Employment Assistance Services clients still earned less than they had preprogram. Every other EBSM group earned more in 1998 than they had before the program.
Income Support from GovernmentData in this section looks at UI/EI20 histories of EBSM participants and non-participants since 1992 (Table 8.3). Prior to EBSM, about 45-55 percent of participants had been on UI at some time during each year between 1992 and 1995. That proportion rose to about two-thirds during 1996 and about three-quarters by 1997. Non-participants started off about the same as participants, but had a much greater tendency to be on UI by 1995. In 1996, the two groups were again about equal. In 1997, non-participants were much less likely to have been on EI than participants. The proportions of participants on EI fell by 1998, but this is not necessarily due to EBSM, since the fall was just as dramatic for non-participants. Differences by component are also apparent in Table 8.3. Employment Assistance Services clients were consistently the least likely, Job Creation Partnerships the most likely to use UI prior to participation. The most notable drop in EI use after the program was experienced by Self-Employment clients. This may reflect the entitlement rules for EI in the sense that self-employed individuals are generally not entitled EI. The percent of Job Creation Partnerships clients receiving EI in 1998 is high considering their program earnings were not insurable.
About one-quarter of the participants and non-participants had been on UI for at least part of every year from 1992 to 1996 (Table 8.4). Only 5 percent of non-participants but 17 percent of participants were never on UI during these years. Regarding participants, the average number of weeks in receipt of UI held fairly stable until it jumped in 1996 and again in 1997, before dropping in 1998. Although both groups (participants and non-participants) experienced the 1998 decline, the decline was steeper for participants.
Mirroring the mean weeks on UI/EI are the mean benefits received by year (Chart 8.2). The two groups are similar before 1996, but diverge radically after that. The slope for participants is steeper than that for non-participants between 1997 and 1998. 
As reported earlier, Employment Assistance Services clients were the least likely of all EBSM participants to use UI prior to participation. But, as the next table shows, they were consistently the most likely to have relied on social assistance before participating. In general, around 10 percent of EBSM clients had been on social assistance during each year from 1993 to 1997. About a fifth of the participants and 14 percent of non-participants had been on social assistance at some point during the five years previous to 1997. At some time in 1998, 13 percent of participants but only 5 percent of non-participants were on social assistance.21 This is a continuation of the pre-program pattern of social assistance use, although the divergence between the groups grew after the program. Limiting the analysis to those who completed EBSM in 1997, 11 percent of participants received social assistance in 1998, still twice as high as the proportion for non-participants.22 The econometric analysis will help determine if this is a consequence of participating in EBSM.
EducationParticipants and non-participants were very different in terms of highest level of education completed (Chart 8.3). Non-participants tended more to the extremes than participants, having been more apt to drop out of school before finishing high school, but also more likely to complete university. The large difference in proportion completing college may be the result of Purchase of training, since most courses were held at community colleges (and 74 percent of participants who completed community college were Purchase of training clients). 
Knowing that it would be hard to render a definitive verdict on the program's impact on education (because of the lack of pre-program data), the evaluators posed a further question to survey respondents to get a sense of whether the program might have influenced the propensity to upgrade their education. It turns out that 23 percent of EBSM participants had gone back to school, college or university since taking part in the program. About 8 percent were in a course at the time of the survey, and another 6 percent had successfully completed a course. Participants took a wide variety of courses including GED, computers, automotive repair, hospitality, and electronics. By comparison, 24 percent of non-participants had taken further education or training since the start of 1997. When one considers that the question covered a two-year period for all non-participants but 59 weeks for the typical participant, it appears that participants were more inclined to upgrade their skills. About 70 percent of the non-participants who had taken a course had paid for their own training; the remainder were enrolled in government training programs. Most of those paying for their own course (79 percent) had taken a skills training course at a community college or private training school. Others went back to school (12 percent) or university (12 percent).23 The typical non-participant course ran for 17.6 weeks. Thus nearly one-quarter of the comparison group experienced a significant training intervention in the last two years. 8.2 Outcomes by ComponentThis section examines client outcomes in each EBSM component. Employment Assistance ServicesEmployment Assistance Services include a range of services and Human Resources Development Canada (HRDC) programs to help clients prepare for, find and keep jobs. HRCCs or third party agencies contracted by Human Resources Centres of Canada (HRCCs) may deliver Employment Assistance Services. Of those Employment Assistance Services clients responding to the survey, 67 percent received services from an HRCC, 18 percent from a third party, and 15 percent from both.24 Third party agencies included employment resource centres (36 percent), social services offices (14 percent), and private training schools (8 percent). Chart 8.4 displays the Employment Assistance Services received by clients. In general, third party agencies provided more services than did HRCCs. For example, HRCC clients were unlikely to get access to volunteer work, whereas nearly 21 percent of third party clients had such access. The only type of service that most HRCC clients received was a workshop on job search skills. 
A key element of EBSM is the development of an action plan to get back to work. Often Employment Assistance Services clients did not develop such plans according to survey data. The previous graph shows that only 39 percent of HRCC clients and 62 percent of external service delivery clients received help to plan their strategy to get back to work. Fewer still developed an action plan: only 19 percent of HRCC Employment Assistance Services clients and 42 percent of third party clients did so. This conforms with the findings presented in Chapter Four: most low-need clients who use Employment Assistance Services do not develop an action plan. Employment Assistance Services clients may also avail themselves of various services without the active assistance of HRCC or third party staff. Chart 8.5 reveals that 80 percent of HRCC clients and two-thirds of those going to third parties for Employment 
Table 8.6 lists the mean grade given by Employment Assistance Services clients to the services they received. Most marks are above average: overall, HRCC clients gave a Band third party clients gave a B to Employment Assistance Services (not a significant difference).
Purchase of trainingThe Purchase of training component provides training courses to clients to improve job skills. Almost 85 percent of Purchase of training clients surveyed asserted they could not have taken the course without the financial support of the program. Chart 8.6 presents the occupation for which Purchase of training clients were training. Sciences, engineering and health, as well as construction were the most popular fields. Asked how closely related their current job is to the job training they received, 58 percent of Purchase of training participants said very closely related, 17 percent said somewhat related and 26 percent said not at all related. 
The typical course lasted for 20.2 weeks, but the length ranged from 1 day to over 100 weeks. The median course length was 9 weeks. Most of the courses were full-time: 77 percent ran for 30 or more hours per week. On average the courses were 35 hours per week. Multiplying the number of weeks by the number of hours per week produces the mean course length: 707 hours. Some 44 percent of the courses included a work placement component. Approximately 43 percent of Purchase of training respondents were determined to be able to contribute to the cost of their training and 40 percent actually did so. The average course contribution was $1,171; but this figure is skewed by nine clients who paid $4,000 or more for their course. The median contribution was $400. Only 11 percent of the trainees had not finished their course by the time of the survey. The main reason given ( by about 40 percent) was that they were still in the course while the remainder had not finished the course for various other reasons. Purchase of training clients seemed impressed with nearly every aspect of the training they received. Overall, they gave the course a B+. In fact, most every aspect was graded B+. Only the monthly income available was marked lower.
Targeted Wage Subsidies and Job Creation Projects25Chart 8.7 shows that client placements by the type of organization mirrors the objectives of each of these work experience programs. Over 80 percent of Targeted Wage Subsidies placements were with private sector employers, whereas 44 percent of Job Creation Partnerships placements were with non-profit community agencies. 
Types of work done for placement employers also differed widely between Targeted Wage Subsidies and Job Creation Partnerships clients (Chart 8.8). Targeted Wage Subsidies jobs were most likely to be in the service, sales or clerical fields. Job Creation Partnerships jobs were chiefly in construction, services or administration. This conforms to the type of projects funded through Job Creation Partnerships. 
The typical client worked 39 hours per week on the placement job. There was no difference between Targeted Wage Subsidies and Job Creation Partnerships. Two-thirds of the work experience clients worked 40 hour weeks. Only 4 percent worked for less than 30 hours weekly. The vast majority of participants in these two components stayed on the job for the entire period of the placement: 84 percent for Targeted Wage Subsidies and 86 percent for Job Creation Partnerships. The most often given reason for leaving before the placement ended was finding a job with another employer (45 percent); other reasons included fired or laid off (11 percent); disagreement with employer (9 percent); moved (5 percent); and disliked the job (5 percent). In many cases, the wage subsidy worked as intended.26 Targeted Wage Subsidies clients had a much higher probability of receiving such an offer, however: 62 percent of Targeted Wage Subsidies participants versus 38 percent of Job Creation Partnerships participants were offered a job after the sudbidized placement ended.27 This variance is to be expected given the different nature of the two components, although both are supposed to subsidize "sustainable" jobs. Reasons respondents offered for not being offered a permanent job by the subsidized employer differed between the components: over two-thirds of the Job Creation Partnerships clients, but only 38 percent of the Targeted Wage Subsidies clients said there were no available positions or no money to hire. Other reasons given by Targeted Wage Subsidies clients for not being offered a permanent job included company closed (17 percent); seasonal employment (10 percent); not interested in working for the employer (9 percent); and don't know (12 percent). Other reasons given by Job Creation Partnerships clients included company closed (6 percent); seasonal employment (6 percent); and not interested in working for the employer (5 percent). Unfortunately, in many cases, the job did not turn out to be permanent. Of those who were retained by their employer after the subsidy, 58 percent were no longer at the same job at the time of the survey. Job Creation Partnerships clients, less likely than Targeted Wage Subsidies clients to be kept on in the first place, were also more apt to have left or lost that job: only 34 percent hired after the subsidy expired were still working with the employer in early 1999. Just under half the Targeted Wage Subsidies clients (48 percent) were still with the placement employer at the time of the survey.28 One-third of work placement clients were on temporary lay-off, so it is probable the picture would be better in the summer. Also, about a quarter of those who left the job went to what they perceived to be a better work opportunity. In the end, then, only 19 percent of all Job Creation Partnerships clients and 35 percent of all Targeted Wage Subsidies clients were still with the placement employer in January 1999. On average, Job Creation Partnerships clients retained after the subsidy but who subsequently left, were employed for 13 unsubsidized weeks. The corresponding figure for Targeted Wage Subsidies clients was 23 weeks.29 Table 8.8 presents the satisfaction ratings of Targeted Wage Subsidies and Job Creation Partnerships participants with various aspects of their intervention. For the most part, these participants rated their work experience good to very good. Guidance on services available after the placement was rated somewhat lower. There were no significant differences in satisfaction between the two groups.
Self-EmploymentThe objective of Self-Employment is to assist clients with viable business ideas to get their business up and running. Most Self-Employment clients (94 percent) were interested in starting a business before becoming involved in the SE component. Only percent had sought assistance from another source in an attempt to start their own business, though. One in five had been self-employed in the past. Apparently most had a reasonable basis for starting the business: 77 percent had previous experience or training in the field of interest. About two-thirds had taken a training course or workshop to help them start a business. About 95 percent of Self-Employment clients made an initial capital investment in their new business. The median contribution was $9,000. Almost every Self-Employment participant (97 percent) started a business through the program. At the time of the survey, 68 percent were still employed in the business they developed (Chart 8.9), which is similar to the proportion reported by business development agency representatives interviewed. The typical business had been in operation for about a year and a half. Interestingly, the amount of the initial investment had no bearing on whether the business survived. 
However these new businesses seem to be struggling judging by their mean monthly profit of $1,390 and the fact that 24 percent reported no profit in a typical month. The highest monthly profit was $10,000. Median monthly profit was only $600. Seven in ten of the new businesses employed only the Self-Employment participant when the survey took place. One business had six full-time employees and two others had five full-time employees besides the owner. In total, Self-Employment businesses had generated 91 jobs — 51 of them full-time positions — besides providing employment for the entrepreneurs. This will understate the true job creation power of Self-Employment since the survey only reached 54 percent of Self-Employment clients. A good guess is the other 46 percent were not as successful since many had moved or were otherwise unreachable. Still it is likely that Self-Employment has produced over 100 direct jobs besides the owners' positions. Sixty percent attributed their success in starting their business at least in part to Self-Employment while 40 percent thought they would have been successful in establishing their own business without the help of Self-Employment. Those who presumed they would have been successful without Self-Employment tended to operate more successful enterprises: all were still in business at the time of the survey, and they created most of the jobs. About half of those who believed they could have started the business without Self-Employment said they were able to start it sooner because of Self-Employment. Asked what the program did for them that they could not have done for themselves, Self-Employment clients most often mentioned the financial support they received from EI while they attempted to get the business going (Chart 8.10). Skills and knowledge to develop a business plan was the most frequently stated active intervention of Self-Employment. 
The 32 percent of SE participants whose business had not been successful were asked how long their business lasted, why it failed and what they would do differently if they had it to do again. On average, the failed businesses lasted for 10 months, with a range of 2 to 19 months. Almost no one was able or willing to answer why the business was no longer operating. They were more forthcoming about what they would do differently if they had the chance to start again. About 14 percent reported they would change locations; another 14 percent would save more money before starting; 11 percent would do more research or better planning; 8 percent would take more business courses; 8 percent wouldn't start again; 14 percent didn't know; and the rest gave other lessons. A priori, one would predict that those whose businesses succeeded would be more approving of the program than would those whose businesses foundered. That is most certainly the case as Table 8.9 demonstrates. Overall, the ratings for Self-Employment were the lowest of the EBSM components.
8.3 Post-program ActivitiesBecause the performance measures focus on the period immediately after completing the intervention, it is important to determine what clients did upon completion. Table 8.10 lists the responses of survey respondents. For the program as a whole, 65 percent were employed immediately after their participation, either because they continued working with their placement employer, continued self-employment, or found a job with- another employer. Another 25 percent began searching for work. There were large differences across EBSMs, most of which are predictable given the nature of each EBSM component. Over three-quarters of Self-Employment participants were self-employed after participation. About 58 percent of Targeted Wage Subsidies and 42 percent of Job Creation Partnerships clients continued working for their placement employer. Over 43 percent of the Employment Assistance Services clients continued their job search. One finding was unexpected: 34 percent of Purchase of training clients continued working for their placement employer (these were apprentices). Earlier it was reported that 44 percent of the Purchase of training courses included a work placement component: with a few exceptions, these clients were apprentices. Apparently, many of these clients impressed their placement employer enough to gain a permanent job there.
Those who were not working or looking for work immediately after the program were asked if they had searched for a job since then. Almost 70 percent said yes. This ranged from 55 percent of Purchase of training clients to 85 percent of Job Creation Partnerships clients. Of those who ended the program looking for work or who subsequently began looking for work, 68 percent had found paid employment by the time the survey took place. Factoring in those who found jobs immediately after EBSM, 11 percent of all EBSM participants who tried to find a job since finishing their intervention had thus far failed to get one. As for non-participants, 68 percent got a job when their EI benefits ceased (in most cases it is likely that EI benefits ended because they found a job). About 14 percent continued looking for a job, and most of the rest withdrew from the labour market. Of this last group, 43 percent eventually began a job search. Close to two-thirds of those who looked for a job since EI ended had found one by the time of the survey. That means about 11 percent who wanted a job hadn't found one by the time of the survey — the same proportion as participants.
For those who did not have a job upon completing their EBSM intervention, but had found one subsequently, it took an average of 11 weeks to find that job. It took non-participants 12 weeks to find a job after their benefits ended (not a significant difference). In total, 84 percent of participants and 71 percent of non-participants had been employed since completing their intervention/benefits.30 Participants had worked at more jobs post-program than had non-participants (Chart 8.11). On average, participants have had 1.2 jobs, non-participants 0.9 jobs since EBSM/EI.31 
Of participants who left their first post-program job, 53 percent had secured another; the comparable figure for non-participants was 48 percent. About 39 percent of participants and 37 percent of non-participants who left their first post-program job immediately began working at a second (Chart 8.12). Most of the rest began looking for a new job. Most participants (65 percent) and non-participants (69 percent) in this position hadn't found a second job by the time of the survey: about four in ten of those who hadn't found a second job hadn't looked for one. Of participants who left their second post-program job, 59 percent found a third. 
Primary Results MeasurementsAn important objective of the survey was to validate primary employment results indicators — how many EI clients have returned to work and the resulting unpaid EI benefits. Only 24 percent of EBSM clients were coded as employed according to the HRIB results data from the same period as the evaluation covered. The findings from the survey indicate this figure is far too low. About two-thirds of EBSM participants were employed immediately after their intervention ended. The employment indicator is defined as EI claimants who become employed before the end of their benefit period and is calculated by looking at the EBSM clients who had 12 consecutive weeks of reduced benefits defined as 25 percent or less of entitlement. Survey results show that about seven in ten participants (68 percent) were employed for at least 12 consecutive weeks following participation in EBSM.32 The first job after EBSM accounted for almost all of this; 6 percent reached the 12 week plateau on their second post-program job. Since virtually all this occurred within one year of leaving EBSM, 67 percent of EBSM clients were employed for 12 consecutive weeks within one year of participating in EBSM. Chart 8.13 presents the proportion of participants with at least 12 consecutive weeks of employment by EBSM component. Self-Employment and Targeted Wage Subsidies clients were most successful in this regard. 
On average, EBSM participants had completed participation in EBSM 58.9 weeks prior to the survey and had worked 43.1 weeks during that time. Employment Assistance Services, Job Creation Partnerships and Purchase of training clients averaged around 40 weeks of post-program employment; Targeted Wage Subsidies averaged 51 weeks, Self-Employment 45 weeks.33 The jobs were 40-hour positions for the most part. 8.4 Current SituationA crucial test of success is what participants are now doing as compared to those who did not participate. Are participants more likely to be working than non-participants? Are they less likely to be on social assistance? Are they less likely to be on EI? Are Self-Employment clients still operating their own businesses? Does participation lead to a greater tendency to upgrade one's education or job training skills? Chart 8.14 provides preliminary answers to these questions. The next chapter will control for other factors that may have lead to these results for a more definitive picture. Before correcting for possible outside influences, it appears that participation in EBSM led to a significantly higher probability of upgrading one's education and training.34 On the other hand, current labour market status is not so positive: EBSM participants were somewhat less likely than non-participants to be employed35; more likely to be unemployed and looking for work36; and more apt to be on EI.37 
As Table 8.12 shows, there is preliminary evidence that Self-Employment was more effective than the other options at removing people from EI and at getting them into a paid job. About 68 percent of Self-Employment participants were self-employed at the time of the survey. Purchase of training clients were at the other end of the spectrum, being least likely to have a job and most prone to rely on EI; Purchase of training clients were also most likely to be in training or upgrading courses.
The reason participants offered most for being unemployed was that there were no jobs available (Chart 8.15). 
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
![[To Top]](/web/20060127000436im_/http://www11.hrdc-drhc.gc.ca/edd-img/top.gif){kind=small}
![[Previous Page]](/web/20060127000436im_/http://www11.hrdc-drhc.gc.ca/edd-img/back.gif){kind=small}
![[Table of Contents]](/web/20060127000436im_/http://www11.hrdc-drhc.gc.ca/edd-img/content.gif)
![[Next Page]](/web/20060127000436im_/http://www11.hrdc-drhc.gc.ca/edd-img/next.gif){kind=small}