learn$ave Project: Final ReportThe largest driver of cost in learn$ave was the costof the project’s administration <strong>and</strong>, within that costarea, some of the largest component were those labourintensiveservices to check documents for project rulesat entry, enrolment <strong>and</strong> cash out. One important way ofreducing costs would be to tie a savings-based program toexisting means-tested or targeted government programs<strong>and</strong> information systems, as in the case of the CESP, <strong>and</strong>take advantage of the latter’s administrative <strong>and</strong> informationstructures. For example, tying eligibility to theWorking Income Tax Benefit or the National Child Benefitmight reduce the costs of recruitment <strong>and</strong> IDA programenrolment. Similarly, targeting a savings-based programto clients receiving transfer income (such as EI or provincialsocial assistance, the latter of whom our researchresults presented in Chapter 5 showed could save) mightalso reduce the costs of recruiting low-income participantsif administrative databases can be used. Further,the cost of administering credits withdrawals could bereduced if the approved education or training institutionscould claim tuition directly from the program (as is doneunder the Canada Student Loan Program or in certaininternational examples of voucher-based programs)through an electronic funds transfer.102 | Chapter 8 <strong>Social</strong> <strong>Research</strong> <strong>and</strong> <strong>Demonstration</strong> <strong>Corp</strong>oration
learn$ave Project: Final ReportChapter 9Lessons learnedThe learn$ave research <strong>and</strong> demonstration project wasdesigned to test how effective a matched savings incentivewould be in inducing low-income adults to save for <strong>and</strong>participate in adult education <strong>and</strong> training or for startinga small business. The idea was conceived by <strong>Social</strong> <strong>and</strong>Enterprise Development Initiatives based on the assetbuildingconcept of Individual Development Accounts,pioneered in the early 1990s in the United States wherethey have been used to combat poverty by encouraginglow-income people to acquire productive assets rangingfrom a home or a vehicle, to micro-enterprise, education,<strong>and</strong> retirement savings. However, the promise of IDAsin increasing adult education enrolment <strong>and</strong> businessstart-ups, which were the purposes learn$ave focused on,was largely unproven in Canada in 2000 when this projectwas implemented.At the core of IDAs is a matched saving incentive. In thelearn$ave IDA, most participants received, for every dollarthey put aside up to $1,500 in a special account over athree-year period, $3 in virtual credits that could be usedfor only education <strong>and</strong> training or small business start-up.To join the project, participants typically must have hadhousehold income over the previous year that amountedto more than 120 per cent of the Low-income Cut-off, nothave had more than $3,000 in household assets, <strong>and</strong> nothave been a full-time student.The project was delivered in 10 communities representinga mix of large- <strong>and</strong> medium-sized urban areas<strong>and</strong> rural communities. At three of the sites (Halifax,Toronto, <strong>and</strong> Vancouver) the project was implementedusing an experimental design with three different groups:a program group receiving the matched credits only; aprogram group receiving financial literacy training <strong>and</strong>enhanced case management supports in addition to thematched credits; <strong>and</strong> a control group made up of participantsnot receiving any credit or ancillary benefits.To estimate impacts, the outcomes of two groups ofparticipants receiving program benefits were comparedto those of the control group. The control grouprepresented the counterfactual, i.e., what program groupmembers would have done if they had not participated inthis intervention. As individuals were r<strong>and</strong>omly assignedto the groups, they were socio-demographically similar<strong>and</strong> the differences in outcomes between the groupscould then be ascribed to learn$ave. This is considered tobe the best way of measuring incremental impacts of aninitiative, as opposed to the traditional program evaluationapproach of using simple pre- <strong>and</strong> post-comparisonsof outcomes.The learn$ave demonstration project is unique <strong>and</strong>noteworthy in several ways. First, it is one of a h<strong>and</strong>ful ofinnovative social program ideas that have been submittedto rigorous tests in Canada over the last two decades.learn$ave is also the second of two experiments worldwideaimed at evaluating the effectiveness of IDAs. Inaddition, the project provided the opportunity to test thecontribution of financial literacy training over <strong>and</strong> abovethe use of financial incentive to save. Lessons learnedthroughout the life of the project constitute a rich body ofinformation <strong>and</strong> insights for any public authority or nonprofitorganization that is considering the introductionof similar approaches to assist low-income or low-skilledindividuals.Implementation <strong>and</strong> deliveryThe initial recruitment difficulties experienced withlearn$ave indicated that if an IDA program like learn$avewere offered nationwide, take-up would be slow at first.It would take considerable time <strong>and</strong> effort to inform <strong>and</strong>attract the target population <strong>and</strong> build a track record,which is what took place in learn$ave <strong>and</strong> is typically thecase for any new program. A national campaign extollingthe benefits of participation <strong>and</strong> education would havehelped learn$ave reach its target earlier by increasingthe low-income public’s comfort with the concept ofasset-building.The program was attractive only to a small portion ofthe target population. Evidence gathered in preparationfor learn$ave, <strong>and</strong> during the project itself, revealed thatmany low-income individuals may value education, butdo not see it as a viable option for them personally, likelybecause of negative past experiences or life constraints.Thus, a program like learn$ave that promotes educationenrolment would not be appropriate for all low-incomeearners — it should appeal to those already inclinedtoward education. Alternatively, allowing for additionaltypes of asset acquisition besides adult education wouldincrease the appeal of IDAs for the broader targetpopulation.Another lesson learned from the implementation isthat it is important to come to some kind of agreementwith government authorities over the treatment of theIDA proceeds. In the interests of recruitment <strong>and</strong> fairness,it was necessary to negotiate with provincial officials toraise asset limits or otherwise disregard the learn$avematches when computing income assistance levels forparticipants. Similarly, federal tax officials had to beconsulted to ensure that the matched credits were notconsidered as taxable income.<strong>Social</strong> <strong>Research</strong> <strong>and</strong> <strong>Demonstration</strong> <strong>Corp</strong>oration Chapter 9 | 103