learn$ave Project: Final ReportTable 8.2learn$ave Present Value Cost-Effectiveness, by ProgramGroup, All Program Group Participantslearn$aveonlylearn$aveplusImpact on enrolment in education5.5 7.5<strong>and</strong> training overall (%) 1Cost per Additional PersonReceiving Education <strong>and</strong>Training ($)70,168 54,966Impact on enrolment in an education7.8 10.8program (%) 1Cost per Additional PersonEnrolled in an EducationProgram ($)Source:Note:48,902 37,952Calculations based on learn$ave site staff time studies, ProgramManagement Information System, accounting records,baseline survey, <strong>and</strong> three follow-up surveys.1Enrolment estimates differ somewhat from those presentedin Chapter 7, as the latter were based on only educationstream participants.discussion of how the costs vary by saving stream, seeBox 8.3.To properly evaluate the cost-effectiveness of learn$ave,a reference or benchmark is needed. As with the abovecomparisons of cost-economy <strong>and</strong> cost-efficiency, oneway to do this is to compare learn$ave’s cost-effectivenessto other programs with similar purposes. Regrettably,insufficient information was available to enable this comparison.While a cost-efficiency comparison was madewith the Employment Insurance Part II employment<strong>and</strong> skills training measures, the available data did notpermit us to calculate a true estimate of cost-effectivenessfor this comparison program. Indeed, estimates of true(incremental) impact of any existing programs are rarelygenerated.Another comparison method is to examine the increasein average participant annual earnings needed to coverthe financial cost of encouraging them (including theparticipants identified as having a windfall benefit) toenter into education. 15 This is not a traditional way ofassessing cost-effectiveness but it at least provides abasic benchmark (breakeven point) to place learn$aveIDA program costs into context. Since a typical learn$aveparticipant was about 33 years old at the time ofenrolment in the project (see Chapter 3), the averageparticipant would have 32 more earning years if retirementage is assumed to be 65. Using an annual discount15 Even when the annual increase in earnings can cover the financial cost, however,cost-effectiveness does not imply there is a net social benefit. There are other costs besidefinancial ones, including the opportunity cost of lost earnings while in school. However, if thepotential annual additional increase in earnings resulting from the intervention is lower thanthe financial cost, it would be more cost-effective to simply transfer an equivalent amount ofmoney from the government to the participants since the program is not likely to be viable.See Table G in Appendix G for more on this.Box 8.3Differences in costs by saving streamIn addition to considering differences in the costs for the two differentprogram groups <strong>and</strong> for two different types of education activity(programs <strong>and</strong> courses), the study of the learn$ave costs also examinedwhether there were differences in costs between the education <strong>and</strong>micro-enterprise streams. Micro-enterprise stream participants wereallowed to use their earned matched credits on small business start-up<strong>and</strong>/or education <strong>and</strong> training, while education stream participantshad to use them for education or training only. Not unexpectedly, theeducation participation rate as well as learn$ave’s impacts on educationwere lower among the micro-enterprise stream.Based on the results in Table 8.3, we conclude that the cost-efficiencyof learn$ave was marginally better for the education streamthan for the micro-enterprise stream. Specifically, the costs peractive-saving-participant, per participant-dollar-saved, <strong>and</strong> per participantreceiving-education(rows 2 through 4) were somewhat lower for theeducation stream than for the micro-enterprise stream.Moreover, the results would suggest that learn$ave could be morecost-effective in encouraging self-employment among micro-enterprisestream members than encouraging education among education streammembers. As indicated in Chapter 7, learn$ave increased the incidenceof self-employment by about 25 percentage points <strong>and</strong> 15 percentagepoints for the learn$ave-only <strong>and</strong> learn$ave-plus groups, respectively.This means that the costs for each new participant who entered intoself-employment were $14,901 for learn$ave-only (per participant costsof $3,658.14 / 0.25) <strong>and</strong> $27,166 for learn$ave-plus ($3,968.91/ 0.146). This means that, to cover the financial cost of encouragingparticipants to enter self-employment, learn$ave-only group membersinduced into self-employment would have to earn less than educationstream members induced into education or training, i.e., they would haveto earn only $943 more a year for 32 years from their self-employmentjob in increased earnings, while learn$ave-plus participants would haveto earn almost twice that: $1,719 per year.rate of 5 per cent for a 32-year period, learn$ave participantswould have to earn an additional $4,440 annuallyfor learn$ave-only participants <strong>and</strong> $3,478 annually forlearn$ave-plus participants to cover the financial costsof learn$ave. These are substantial amounts <strong>and</strong> wouldmean large annual increases over participants’ baselineaverage annual income of $11,000. For those participantswho invested their learn$ave dollars in education ortraining courses, previous research suggests the returnswill be small <strong>and</strong> unlikely to meet this threshold test. Asa result, we conclude that learn$ave is not likely to be a98 | Chapter 8 <strong>Social</strong> <strong>Research</strong> <strong>and</strong> <strong>Demonstration</strong> <strong>Corp</strong>oration
learn$ave Project: Final ReportTable 8.3learn$ave Present Value Cost-Economy, Cost-Efficiency, <strong>and</strong> Cost-Effectiveness ($), by Saving Stream <strong>and</strong> Program Group, AllProgram Group ParticipantsEducation StreamMicro-Enterprise Streamlearn$aveonlylearn$avepluslearn$aveonlylearn$aveplusCost-EconomyCost per Program Group Member 3,660 3,971 3,658 3,969Cost-EfficiencyActive participants as a proportion of all participants (%) 81.7 85.5 74.9 79.0Present Value Cost per Active Participant 4,480 4,643 4,885 5,026Average peak learn$ave savings 934 991 886 946Present Value Cost per Dollar Saved 3.92 4.01 4.13 4.20Proportion of participants who enrolled in education/training (%) 1 88.0 89.7 68.1 73.5Present Value Cost per Participant Enrolled in Education or Training 4,159 4,428 5,375 5,401Proportion of participants with self-employment (%) … … 66.7 56.7Present Value Cost per Participant with Self-Employment … … 5,485 6,995Cost-EffectivenessImpact on enrolment in education or training (percentage points) 6.6 8.2 -4.0 1.5Cost per Additional Person Enrolling in Education or Training 55,872 48,303 N/A N/AImpact on enrolment in an education program (percentage points) 1 9.1 12.6 -2.3 0.3Cost per Additional Person Enrolled in an Education Program 40,260 31,537 N/A N/AImpact on self-employment (percentage points) … … 24.6 14.6Cost per Additional Person in Self-Employment … … 14,901 27,166Source: Calculations based on learn$ave site staff time studies, Program Management Information System, accounting records, baseline survey,<strong>and</strong> three follow-up surveys.Note: … Self-employment outcomes not cost-effective or appropriate for the education stream.1Enrolment estimates differ somewhat from those presented in Chapter 7, as the latter were based on only education stream participants.cost-effective way to increase participation in coursesamong low-income adults in Canada.However, learn$ave does show better promise, from acost-effectiveness perspective, as a way to encourage lowincomeadults to enrol in education or training programs.Education program participation among all participantswas 7.8 per cent higher among the learn$ave-only group(compared to the control group) <strong>and</strong> 10.8 per cent higherfor the learn$ave-plus group. The higher impacts oneducation program enrolment translate into greater costeffectivenessfor education program outcomes than foreducation <strong>and</strong> training outcomes overall: $48,902 for thelearn$ave-only group <strong>and</strong> $37,952 for the learn$ave-plusgroup. To cover these costs, learn$ave-only participantswould need to receive $3,095 a year in increased earnings,<strong>and</strong> learn$ave-plus participants $2,402 a year. Again,research indicates that returns to education are greatestwhen the education is a program that leads to a postsecondarycertification or qualification. 16 To the extentthat learn$ave participants were encouraged or limitedto using their accounts to fund education programs, thecost-effectiveness would likely be much higher.Costs by immigrant statusGiven that such a sizeable subset of the learn$averesearch sample were recent immigrants to Canada, thestudy of learn$ave cost-efficiency <strong>and</strong> cost-effectiveness16 Based on 2006 Census data on 2005 median earnings for full-time earners 25-64 yearsof age by highest level of educational attainment (Statistics Canada, 2009), the returns(increase in earnings) to a high school certificate (relative to not having a high schoolcertificate) are $5,374; <strong>and</strong> the returns to a post-secondary education (relative to a highschool certificate) are: $2,593 for a trade certificate; $5,534 for a college certificate;<strong>and</strong> $18,645 for a Bachelor’s degree. These figures are not adjusted for any systematicdifferences between graduates at different levels of education.<strong>Social</strong> <strong>Research</strong> <strong>and</strong> <strong>Demonstration</strong> <strong>Corp</strong>oration Chapter 8 | 99