Canada Education Savings Program: Summative Evaluation Report

Main findings

Key findings

1.1 Limitations

It will not be until 2015 that CESG beneficiaries who were born in 1998 will reach the age of 17 and begin attending a PSE institution. They will be the first age cohort that could fully benefit from CESG payments to a subscriber’s RESP for the 17 years that program funding is paid and have the full amount of RESP assets possible. Thus, the full impact of the CESG on some aspects (e.g. RESP assets) cannot be measured yet. In addition, the impact of RESPs on financing PSE and reducing student debt also will not be possible to measure for an even longer period of time.

Another limitation is that some of the lines of evidence use dated Statistics Canada surveys. ^{Footnote 5} However, most of the findings from these surveys were corroborated with more recent data sources, including data from the CESP administrative database, a survey conducted by Evaluation during the 2013-2014 fiscal year, and analyses involving linked Canada Revenue Agency (CRA) T1 income data and CESP administrative data up to and including the 2012 taxation year.

Finally, although random sampling was used for the 2013-2014 Evaluation survey, the survey data are un-weighted and the results should not be extrapolated to the entire population. However, since the regional pattern of response for the actual sample is very close to the distribution of the sample that would be expected, this provides some confidence to the reliability of the random sampling approach.

2.1 RESPs and CESP incentives

RESPs were introduced in 1972 and allow contributions to grow tax-free until beneficiaries attend a PSE institution. Although there is no longer an annual contribution limit, there is a lifetime contribution limit per beneficiary, which has been $50,000 since 2007. Once a beneficiary begins attending a PSE institution, RESP funds can be withdrawn and are paid out as contribution withdrawals (also referred to as Refund of Contributions or ROCs by the Canada Student Loans Program (CSLP)) to the subscriber and Education Assistance Payments (EAPs) to the beneficiary depending on the details of the RESP contract. ^{Footnote 6}  EAPs include accumulated investment earnings in the RESP and government education savings incentives. RESP contributions are withdrawn tax-free. However, EAPs are taxable to the beneficiary attending PSE (since many PSE students have little or no income EAPs are often withdrawn tax-free or at a low tax rate).

The CESP encompasses three specific measures: ^{Footnote 7}

1. the CESG,
2. the A-CESG and,
3. the CLB.

They are each discussed in turn. ^{Footnote 8}

The Basic CESG was introduced in 1998 and provides a grant of 20% on the first $2,500 of annual RESP contributions for children until the end of the calendar year during which they turn 17. ^{Footnote 9}  Unused grant room can be carried forward. ^{Footnote 10}

The A-CESG came into effect on January 1, 2005. The A-CESG amount depends on the net family income of the beneficiary’s primary caregiver(s):

• If net family income ^{Footnote 11} was below $43,953 in 2013, the A-CESG was 20 cents for every dollar on the first $500 of annual contributions into an RESP (i.e. a maximum of $100); and
• If net family income was between $43,953 and $87,907 in 2013, the A-CESG was 10 cents for every dollar on the first $500 of annual contributions into an RESP (i.e. a maximum of $50).

It should be noted that the unused A-CESG room cannot be carried forward (in contrast to unused CESG grant room).

The CLB was introduced in Budget 2004. To be eligible for the CLB, the beneficiary’s primary caregiver(s) must be receiving the National Child Benefit Supplement and the child must be born on or after January 1, 2004. To receive the CLB, an individual must open an RESP and apply for the CLB, but contributions are not required.

The amount of the CLB is equal to the sum of the following amounts, and can add up to a lifetime maximum of $2,000 per child:

• $500 for the first year of eligibility for the National Child Benefit Supplement, up to and including the child’s 15^th year; and
• $100 in each subsequent year, up to and including the child’s 15^th year. ^{Footnote 12}

2.2 Objectives of CESP measures

The CESG was introduced in Budget 1998 along with enhancements to the Canada Student Loans Program (CSLP) and the creation of the Canada Millennium Scholarship Foundation as part of the Canadian Opportunities Strategy. The strategy proposed action on many fronts, including promoting access to PSE by helping students in financial need cope with rising costs and helping families save for their children’s education in RESPs.

Part of this strategy aimed to address the challenge of “encouraging families to save early for their children’s education”. As stated in the Education Savings Act, the objective of the CESG is to encourage the financing of children’s PSE through savings from early childhood in RESPs. And, according to Budget 1998 (page 13), “…to ensure families can better afford higher education for their children by providing stronger incentives for saving through the new CESG…”

As mentioned in the introduction, the government response to the low RESP participation by low- and middle-income families was the introduction of the A-CESG and CLB. The basic objective of the A-CESG is to strengthen financial assistance for low- and middle-income families who want to save for the PSE of their children. The CLB was implemented with a similar objective to the A-CESG – to help modest-income families to start saving early for the PSE of their children – but without requiring contributions.

3.1 PSE costs

To examine the relevance of the CESP, it is important to outline the context in which it operates. PSE credentials are becoming increasingly important in the labour market. As tuition fees and other PSE-related costs reach several thousands of dollars per year, most parents are expected to pay some portion of the PSE costs of their children (based on the CSLP needs assessment), which can be easier if parents save for PSE. The evaluation acknowledges that not all low-income families should save for PSE, as immediate financial demands (for food, shelter, etc.) can alone be overwhelming when these families face financial hardship. ^{Footnote 13} In this overall context, different levels of government provide financial support for PSE students through different means such as loans, grants, and tax incentives (in addition to providing incentives for parents to save for the PSE of children via the CESP).

Although just one element of total PSE costs, total required fees (tuition and additional fees ^{Footnote 14} ) for full-time domestic undergraduate university students in Canada ^{Footnote 15} were $3,884 in 2000-2001 and estimated at $6,253 in 2013-2014. ^{Footnote 16} This represents a 61% increase, twice that of inflation. ^{Footnote 17} Although average public college tuition fees are not as high as university tuition fees, they increased at roughly the same rate. ^{Footnote 18} In addition to outpacing the rise in prices in the economy, the growth in undergraduate university and college fees has also significantly outpaced the rise in net family income, albeit less so for families in the highest income tercile.  Thus, rising tuition and additional fees appear to justify the need for the CESP to encourage PSE savings, although this need is clearly lower for those in the highest income tercile. ^{Footnote 19}

Accounting for living expenses as well, TD Canada Trust estimated that the overall cost of pursuing a four-year undergraduate degree starting in 2011 was around $80,000 (for students living away from home all four years). ^{Footnote 20}  This estimate is similar to estimates contained in the Actuarial Report on the Canada Student Loans Program (CSLP), where average total student expenses (which includes tuition fees, books, shelter, food and transportation) for 2010-2011 were estimated to be $16,100. ^{Footnote 21}  These expenses are projected to first surpass $20,000 annually in 2018-2019 and to eventually hit $36,400 in 2035-2036 (projections based on annual increase ranging from 2.8% to 3.8%). These findings are also in-line with results from the 2013 CESP survey, where parents with children under 18 years of age expected an annual cost of roughly $20,000 per year once their child went onto PSE. ^{Footnote 22}

To help cover rising PSE costs, many students (particularly those from low- and middle-income families) might have to rely increasingly on employment income (either while in study or prior to PSE) and student loans and grants. ^{Footnote 23} Not surprisingly, this will have an impact on student debt levels which, according to a 2013 Bank of Montreal Student Survey, are already expected to be $26,297 for current graduating students. ^{Footnote 24} Similarly, a 2012 Canadian University Survey Consortium survey noted that graduating students reported an average total debt of $24,579. ^{Footnote 25} Given projected PSE costs in the future, student debt levels are expected to continue to increase – although this will be somewhat mitigated as more students with RESP savings enter colleges and universities and as parental incomes and starting salaries increase over time.

3.2 Alignment with federal government priorities

The department reiterated the importance of the CESP in the 2013-2014 Report on Plans and Priorities by underlining the importance of “…reducing barriers to education by providing financial assistance to individuals as well as incentives to save for a child’s post-secondary education”. In January 2014, the Minister of State reiterated the importance of RESPs and the CLB by mentioning that “Our government recognizes that access to post-secondary education is vital, not only for a young person’s future, but also for Canada’s long-term growth and prosperity. An RESP, supplemented with the Canada Learning Bond, helps modest-income families make post-secondary education a reality for their children”. ^{Footnote 26} These and other recent references to the CESP and RESPs clearly show that education, a knowledge society and saving for PSE continue to be a priority for the federal government.

3.3 Efficient incentives to increase savings

The effectiveness of grants and bonds in motivating people to save in general is discussed below. This is an important discussion because it speaks to the relevance of the CESP and whether the proper incentive(s) to invest in RESPs are being offered.

Theory suggests that incentives which promote general savings could result in three strategies (or a combination of the three): (i) individuals divert savings from another savings vehicle in order to take advantage of the incentive (i.e. no net increase in total savings); (ii) individuals reduce their consumption to take advantage of the incentive, leading to an increase in overall savings; or (iii) individuals save less of their own money because the incentives enables them to reach the same savings goal without contributing as much of their own money. The two current approaches under the CESP to increase PSE savings are matching individual contributions (CESG) and providing lump sum amounts to account holders (CLB). A third possible approach (default participation that has been tested in other jurisdictions) is also discussed.

Although not specifically related to PSE, the literature on retirement savings indicates that matching contributions had positive but modest effects on participation (e.g. a 25% matching contribution was associated with a 5% participation increase). ^{Footnote 27} Madrian (2012) summarized the results of matching contributions by indicating that “a matching contribution increases savings plan participation and contributions, although the impact is less significant than the impact of non-financial approaches [automatic enrollment, simplification, planning aids, reminders, etc.]”. Although matching contributions are shown to increase savings participation, the effect of matching contributions on the savings rate (i.e. amount saved) was usually found to be small and not always statistically significant.

The matching threshold appears to have a greater impact than the matching rate. For instance, a matching contribution rate of 25% on a higher maximum level of contributions (e.g. $5,000) was associated with higher savings than a matching contribution rate of 50% on a lower maximum level of contributions (e.g. $2,500). ^{Footnote 28}

Benjamin and Smart (2011) examined the effect of RESPs/CESP on savings behaviour and their results indicated an increase in RESP balances since the introduction of the CESP, but there was little evidence that overall net financial assets increased among eligible households. In another study, Benjamin and Smart (2012) reiterated this by indicating that “the main effect of the CESP may be to induce substitution between tax-preferred assets, rather than to increase overall household saving or educational attendance.” ^{Footnote 29} Nonetheless, they conclude that “the RESP Program may have its greatest impact in serving to educate households to plan ahead for the financing of university”.

The former Child Trust Fund (CTF) in the United Kingdom offers an example of lump-sum contributions, as well as default participation (i.e. automatic enrolment if an account was not opened within a year). Launched in 2005, the CTF provided a universal benefit (£250 or £500) for children at birth and when children reached seven years of age. Families were eligible to initially contribute up to £1,200 annually (increased to £4,000 later on), and savings grew tax free. ^{Footnote 30}

An evaluation of the CTF ^{Footnote 31} indicated that the majority of parents (78%) opened a CTF account by themselves (including 67% of low-income parents) but only about 37% of CTF accounts received an individual contribution. Where accounts were opened automatically by the government, only 9% received individual contributions.

Finally, an experimental study ^{Footnote 32} from Oklahoma showed that those who had a 529 plan ^{Footnote 33} (an education savings plan operated by a state or educational institution designed to help families set aside funds for future college costs) opened for them with an initial deposit, matching incentives, program materials and regular statements were more likely to have received contributions than children who did not benefit from these advantages. ^{Footnote 34} However, evidence was mixed regarding the effect on actual amounts saved. It was shown that the program mainly benefits higher-income families. ^{Footnote 35}

The U.K. experience with the CTF and the experimental study in Oklahoma suggest that automatic enrolment (which currently requires income-testing consent on behalf of potential participants) would significantly increase participation in the programs examined, but for the CLB the issue is slightly different as CLB receipt does not require any contributions or savings. The U.K. experience with the CTF suggests that if fully automatic CLB enrolment was feasible/practical (which it currently may not be due to the requirement of income-testing consent on behalf of potential participants), it could significantly increase RESP participation, but it may not necessarily affect individual RESP contributions – thereby not having much impact on encouraging PSE savings using RESPs. These findings are supported by the literature findings from Madrian & Shea (2001), Connelly and Kohler (2004), and Choi, Laibson, and Madrian (2004). Thus, it would not be the most efficient way of encouraging PSE savings using RESPs.

The literature presented here suggests that the current federal approach of grants and bonds provides an appropriate incentive for low-income families to save for PSE. In the case of the CLB, the financial situation of many of these families might preclude them from making RESP contributions. As evidence, the Survey of Household Spending (SHS) study showed that families in the lowest income category spent about 82-85% of their annual family income on basic necessities (i.e. food, shelter, clothing and transportation) compared to roughly 60% and 42% for families in the middle- and highest-income categories, respectively. With such a high percentage of income being spent on the basic necessities by low-income families, their likelihood of generating RESP savings on their own is significantly lower. Nevertheless, others such as grand-parents, relatives, or friends may contribute into the RESP account of children from low-income families and enable them to receive the grants and bonds.

4.1 RESP participation and savings

The cumulative level of RESP assets has increased from $2.4 billion in 1997 to $40.5 billion in 2013 (refer to Table 1). ^{Footnote 38} RESP asset levels in 2013 increased by $4.9 billion from 2012.

Up to the end of 2013, the cumulative amount of all CESP payments paid into RESPs was $8.5 billion ($7.7 billion in CESGs, $298 million in A-CESGs and $499 million in CLBs).

Since the introduction of the A-CESG, the annual amount of A-CESG payments has increased from $7 million in 2005 to $54 million in 2013, while annual CLB payments have increased from $450,000 in 2005 to $101 million in 2013. Combined with CESG payments of $728 million in 2013, $883 million was disbursed via the CESP in 2013.

Table 2 provides some insight into the share of eligible children under 18 years of age who have ever received the CESG – in 1998, 9.7% of all eligible children had received the CESG at least once. By 2013, the share of eligible children who had received the CESG at least once had increased to 47.1% (i.e. of the roughly 6.9 million children under the age of 18 in Canada in 2013, 3.26 million had an RESP and had received the Basic CESG at least once). ^{Footnote 39}  Almost 2.5 million of these 3.26 million beneficiaries, or 75.3%, made a contribution and received the Basic CESG in 2013. Of these, 860,000 (or about 35%) received the A-CESG.

The number of A-CESG beneficiaries (i.e. those receiving an additional 10% or 20% grant in a given year) increased from 120,000 in 2005 to 860,000 in 2013. As a proportion of the entire population of A-CESG eligible children, this represents an increase from 2.7% in 2005 to 17.4% in 2012. ^{Footnote 41}

It is important to note that the percentage of A-CESG eligible children receiving the A-CESG would have been higher if all A CESG eligible children who received contributions into their RESPs in a given year had received the A-CESG. Additional data analysis examining families with children using the CESP administrative data linked with CRA T1 income data (representing 85% of all RESP expenditures ^{Footnote 42} ) indicated that 33.8% of A-CESG eligible families who contributed into an RESP in 2012 and received the CESG did not receive the A-CESG. ^{Footnote 43} The three main reasons for this are: (i) the RESP subscriber did not apply for the A-CESG at their financial institution; and/or (ii) some financial institutions did not offer the A-CESG ^{Footnote 44} ; and/or (iii) some of these children received the A-CESG in the RESP of another relative outside their household. In 2012, 30.9% of A-CESG eligible families who made an RESP contribution were not registered for the A-CESG.

The annual number of CLB beneficiaries has increased from 800 in 2005 to 384,100 in 2013. The cumulative number of CLB beneficiaries has increased to 604,566 in 2013 and the share of all CLB-eligible children who have ever received the CLB has increased from less than 1% in 2005 to 29.4% in 2013. About 23% of CLB-eligible families with an RESP in 2012 did not receive the CLB in 2012, as they were not registered for it at their financial institution or children received the CLB in the RESP of another relative outside their household. ^{Footnote 45} The program’s efforts to promote A CESG and CLB since 2005 have helped to significantly reduce the number of eligible RESP subscribers who did not apply for the A-CESG and/or CLB. Moreover, a new streamlined application form was introduced in 2013 that encourages new subscribers to apply for all CESP incentives (subscribers now have to specifically opt-out of the incentives they do not wish to be considered for).

4.2 Savings patterns in RESPs

The number of children receiving RESP contributions has increased each year since 1998. In 2013, 75.3% of RESP beneficiaries under 18 received a contribution, as not all children with RESPs receive contributions every year. ^{Footnote 46}

In 2013, the average RESP contribution (among those with a contribution in that year) was $1,497 compared to $1,202 in 1998. ^{Footnote 47} However, average annual RESP contributions adjusted for inflation (in $2006) have declined over this period – from $1,436 in 1998 to $1,331 in 2013 (refer to Graph 1: Average Annual RESP Contributions from 1998 to 2013). Further analysis using the linked CRA-CESP data (i.e. data examining RESP subscribers living with their children and representing 85% of all RESP expenditures) indicated that average contributions adjusted for inflation have declined for all income groups since 2004. ^{Footnote 48}

Families receiving only the Basic CESG increased their annual RESP contributions from $1,399 in 2006 to $1,676 in 2013 (or from $1,399 to $1,491 adjusted for inflation). By comparison, the average for those receiving the Basic CESG and the A-CESG changed little, going from $1,170 to $1,181 (or decreasing from $1,170 to $1,050 adjusted for inflation), while the average for those receiving the CLB changed from $947 to $1,039 in 2013 (or decreased from $947 to $923 adjusted for inflation).

Receipt of the CESP incentives and RESP contributions also fluctuate significantly by other factors such as a child’s age, province, parental education and financial literacy.

Text description of graph 1

Graph 1 shows average annual RESP contributions (in 2006 constant dollars) for all CESP beneficiaries, basic CESG recipients, A-CESG recipients, and CLB recipients during each year from 1998 to 2013. Graph 1 is based on a 10% random sample of CESP administrative data of recipients under 18 years old and includes only those with an RESP contribution for that year.

	All CESP Beneficiaries	Constant dollar - 2006 / Basic CESG Recipients	A-CESG Recipients	All CLB Recipients	CLB Recipients
1998	1436	1436	N/A Not Applicable	N/A Not Applicable	N/A Not Applicable
1999	1705	1705	N/A Not Applicable	N/A Not Applicable	N/A Not Applicable
2000	1533	1533	N/A Not Applicable	N/A Not Applicable	N/A Not Applicable
2001	1455	1455	N/A Not Applicable	N/A Not Applicable	N/A Not Applicable
2002	1400	1400	N/A Not Applicable	N/A Not Applicable	N/A Not Applicable
2003	1372	1372	N/A Not Applicable	N/A Not Applicable	N/A Not Applicable
2004	1388	1388	N/A Not Applicable	N/A Not Applicable	N/A Not Applicable
2005	1386	1409	1172	656	889
2006	1366	1399	1170	886	947
2007	1402	1453	1192	896	976
2008	1388	1455	1167	859	972
2009	1369	1457	1131	818	958
2010	1384	1491	1123	781	939
2011	1340	1468	1075	737	901
2012	1327	1474	1055	732	901
2013	1330	1489	1049	741	923

Source: 10% random sample of CESP administrative data of recipients under 18 years old. Averages include only those with an RESP contribution for that year

4.3 RESP use by income group and contributors

Analysis using CRA T1 income tax data linked to CESP administrative data (the data examined only RESP subscribers living with their children and therefore represents 85% of all RESP expenditures) showed that the 2012 RESP take-up rate varies significantly by net family income, from 25.2% for families with net family income below $25,000 to 70.1% for families with net family income over $125,000. ^{Footnote 52}

4.3.1 RESP use by family income level

Each year from 1999 to 2012, the number of families with RESPs increased. Graph 2 examines families with children by income group and shows that the number of families with RESPs increased every year in each income group.

In parallel, more families in the two lowest income categories (of Graph 2) are receiving the A-CESG (as shown in Table 2). The study using linked CRA-CESP data indicated that the RESP take-up rate among families with an income below the lowest A-CESG threshold increased from 7.8% in 1999 to 30.3% in 2012 (a four-fold increase), while the RESP take-up rate among families with an income between the two A-CESG thresholds increased from 15.6% to 47.3% (a three-fold increase), and the RESP take-up rate among families above the highest A-CESG threshold increased from 26.5% to 63.8% (over a two-fold increase). ^{Footnote 53} However, although RESP take-up rates continue to increase, Graph 3 shows that visually, the gap in RESP take-up rates between these three groups of families is not narrowing, even after the introduction of the A-CESG and CLB. ^{Footnote 54}

Text description of graph 2

Graph 2 shows the number of families with RESPs for all CESP beneficiaries, basic CESG recipients, A-CESG recipients, and CLB recipients for each year during the 1999 to 2012 period. Graph 2 is based on a 1% sample of families living with children (CRA T1 income tax data linked with CESP administrative data with 545,274 observations from 1999-2012).

Sources: 1% sample of families living with children (CRA T1 income tax data linked with CESP administrative data with 545,274 observations from 1999-2012). This sample of families living with children represents 85% of CESP expenditures. * Annual A-CESG thresholds are used, which are also CCTB thresholds. For years before the introduction of the A-CESG, CCTB thresholds are used.

Text description of graph 3

Graph 3 shows RESP take-up rates (in %) by family income level for each year during the 1999 to 2012 period. Graph 3 is based on a 1% sample of families living with children (CRA T1 income tax data linked with CESP administrative data with 545,274 observations from 1999-2012).

Sources: 1% sample of families living with children (CRA T1 income tax data linked with CESP administrative data with 545,274 observations from 1999-2012). This sample of families living with children represents 85% of CESP expenditures. * Annual A-CESG thresholds are used, which are also CCTB thresholds. For years before the introduction of the A-CESG, CCTB thresholds are used. The $25,000 and $125,000 thresholds are adjusted for inflation each year (real $2012).

The growth in RESP take-up rates among low- and middle-income families has, not surprisingly, coincided with an increase in the number of A-CESG beneficiaries (from 120,000 in 2005 to 800,000 in 2012 as shown in Table 3). However, statistical analysis using the linked CRA-CESP data failed to attribute the growth in RESP take-up rates to the A CESG. RESP take-up rates among low- and middle-income families were already increasing prior to 2005 and did not accelerate after the introduction of the A-CESG. The analysis of the linked CRA-CESP data also failed to uncover a significant impact of the A-CESG on RESP contributions. ^{Footnote 55}

However, analysis using CESP administrative data showed beneficiaries eligible and registered for the A-CESG were more likely to receive contributions in their RESP (by over 10 percentage points) than all other RESP beneficiaries. The analysis also showed a $233 increase in annual contribution amounts for those eligible for the A-CESG who had an RESP before the measure was put in place and made the effort to return to their financial institution in 2005 or 2006 to do the paperwork required to register for the A-CESG. ^{Footnote 56} As this small group is not representative of the eligible population, one cannot conclude that the A-CESG increased contribution amounts among the entire eligible population, even though contributions did increase significantly in the group examined.

Further, the CRA-CESP study found that the CLB increased the RESP take-up rate among low-income families. As only children born on or after January 1, 2004 are eligible for the CLB, the analysis compared low-income families with children born between July and December 2003 (i.e. not eligible for the CLB) to low-income families with children born between January and June 2004 (i.e. eligible for the CLB) to examine the effect of the CLB. Results showed that by 2012, RESP take-up rates were 8.3 percentage points higher among the latter group. However, the analysis failed to uncover a significant impact of the CLB on annual or cumulative RESP contributions. The CESP administrative data analysis also showed that the CLB decreased the age at which RESPs are opened by one year, on average.

5.1 Who saves outside of RESPs?

The 2013 CESP survey found that 25.1% of families with RESPs for their children under 18 years old saved for PSE using non-RESP means, while 38.6% of families with children under 18 years old without an RESP did so. Further analyses of the survey data showed that a greater proportion of families with incomes under $45,000 saved for PSE using non-RESP means than families with incomes of $45,000 or more. The analyses also indicated that saving outside of an RESP occurred more frequently among parents without university studies than among parents with some university studies.

The 2013 CESP survey also revealed that the most common methods of saving for PSE outside RESPs were the use of a dedicated savings account in the name of the child (approximately 35%), followed by the use of other types of accounts (close to 19%) and Tax-Free Savings Accounts (TFSA – 15%). Other savings methods included real estate investments and/or the rental income derived thereof, life insurance funds and/or policies, and RRSPs. The case studies produced similar findings.

5.2 Reasons for saving outside of an RESP

The most common reasons for saving outside of RESPs (among RESP subscribers) given in the CESP survey were ‘diversification’ (21%) and ‘easy access to funds’ (17%). Another 7% of families indicated that they had either maximized the annual CESG received (i.e. contributed $2,500 and received the maximum $500 in grants) or that they had contributed the lifetime RESP maximum of $50,000. Almost all surveyed families who provided these answers had an annual income of $90,000 or more. Additionally, some subscribers indicated that RESPs were not able to completely cover the increasingly high costs of university (and that additional non-RESP savings was needed). The case studies also identified flexibility (i.e. accessibility to funds, use of funds for any purpose, and allowing other people to contribute) as reasons for contributing to PSE savings via other means.

Reasons given by non-subscribers for using non-RESP methods to save for PSE included ‘having accessibility to funds’ (17%) and ‘not having gotten around to it [opening an RESP] yet’ (14%). Only 10.8% listed a lack of RESP awareness (and the associated benefits thereof) as the main reason for saving outside of an RESP. The frequency of this answer was even lower among parents with children under 18 years of age (only 5.9%). Thus, it appears as though awareness is not really an issue when it comes to explaining lack of RESP take-up among non-subscribers who are already saving for PSE. Among those not saving for PSE at all, 90% of those surveyed indicated being familiar with RESPs, but a significant portion were still unaware of the benefits of using an RESP, as over 40% said they would now be likely to save in an RESP after the benefits were explained to them.

6.1 Use of RESP savings by PSE students

In 2013, total aggregate RESP withdrawals reached over $2.7 billion, up markedly compared to the $45 million withdrawn in 2000. ^{Footnote 61} Similarly, the number of students withdrawing from RESPs has increased steadily since 1998, from less than 10,000 to 356,916 in 2013. In percentage terms, about 17% of PSE students made an RESP withdrawal in 2013, up from less than 0.3% in 1998. ^{Footnote 62}

Between 1998 and 2013, the average annual RESP withdrawal increased from $3,705 to $7,673 (among those with an RESP withdrawal of at least $1). Graph 4 shows that this increase is mainly due to the increase in contribution withdrawals during PSE, as EAPs (the withdrawal of accumulated investment earnings in the RESP and government grants) have remained relatively stable since 2007 (average Canadian undergraduate tuition fees are also shown as a point of reference).

Text description of graph 4

Graph 4 shows average RESP withdrawals and average Canadian undergraduate tuition fees for each year during the 1998 to 2013 period. Graph 4 is based on a 10% random sample of CESP administrative data and Tuition and Living Accommodation Costs survey data from Statistics Canada.

	EAP Withdrawals	Withdrawal of Contributions	RESP Withdrawals	Average Undergraduate Tuition Fees
1998	2778	4673	3705	3064
1999	2559	4336	3896	3328
2000	2600	4256	4008	3447
2001	2292	4046	3479	3585
2002	2306	4077	3953	3749
2003	2216	4353	4427	4018
2004	2402	4411	4732	4140
2005	2550	4599	5073	4211
2006	2807	4841	5510	4400
2007	3132	5214	6026	4558
2008	3111	5432	6117	4747
2009	3044	5581	6197	4942
2010	3163	5977	6671	5146
2011	3038	6270	6803	5313
2012	3139	6653	7196	5586
2013	3404	6952	7572	5772
Total	2923	6019	5524

Source: 10% random sample of CESP administrative data and Tuition and Living Accommodation Costs survey data from Statistics Canada. Average withdrawals are calculated excluding zeroes.

In 2012, over 45% of students with RESP withdrawals were in their first year of a PSE program, 25% were students in their second year, 16% were students in their third year, and the remaining in further years. Administrative data analyses revealed that in 2013 the average withdrawal was highest if the student was in the first year of a PSE program and declined thereafter. ^{Footnote 63} Further analyses indicated that approximately 70% of students with RESP withdrawals were in university, while 30% were in college or other studies. Not surprisingly, 2013 RESP withdrawals for those studying at university ($7,908) were greater than for those studying at a college or other PSE institution ($6,814).

The average cumulative amount of RESP withdrawals increased from $8,000 for the generation who reached 18 years old in 1998 to $15,500 for the generation who reached 18 years old in 2008. ^{Footnote 64} Of beneficiaries who started withdrawing from an RESP before 2009, 28% made a withdrawal in only one year, 23% in two years, 41% in three or four years, and 7% made a withdrawal in five or more years. ^{Footnote 65}

6.2 Impact of RESPs on PSE funding

Those with RESP savings should be in a better financial position to afford their PSE studies than those without RESP savings, all else being equal. Analyses using the ASETS confirmed this, as did the CESP survey. The CESP survey showed that PSE students with an RESP withdrew an average of $6,830 from their RESP, accessed $9,525 in non-RESP funding (i.e. any other type of financial support – loans, grants, employment income, etc.), and had PSE-related costs averaging $14,094 in 2012. By comparison, students without an RESP accessed $8,958 in non-RESP funding and had PSE-related costs of $12,739 in 2012. Thus, on average, students without an RESP appeared to be short of funds to finance their PSE. ^{Footnote 66}

Based on the above figures, the percentage of PSE-related costs in 2012 that were covered by RESP withdrawals (i.e. ROCs and EAPs) equates to 48.5%, among those who made RESP withdrawals. Respondents provided a lower percentage (37%) when specifically asked “…what percentage of these [PSE-related] costs were covered by the RESP savings that were used.” A similar percentage (36%) was reported in the ASETS study by RESP users. Therefore, it appears that during the time period covered by these studies, RESP savings covered about 40% of PSE-related expenses, in years students had RESP withdrawals.

7.1 Efficiency & economy – Assessment of resource utilization

The objective of the CESG, which comprised 82.4% of all CESP payments in 2013, is to address the challenge of “encouraging families to save early for their children’s education.” A logical question to ask is if there is a need to encourage all families, specifically those with higher incomes, to save early for their children’s education via government grants. The reason to ask this question is because families with higher incomes:

1. Have more financial resources at their disposal and are therefore able to save more money than lower- and middle-income families (as the SHS study demonstrated). Thus, their need for the incentives offered through the CESP is lower.
   1. In addition, they can better afford to pay a part of their children’s PSE costs during PSE studies using their income in those years.
2. Have a higher degree of financial literacy than lower- and middle-income families and are therefore more aware of the costs associated with PSE and the importance of saving early for PSE. ^{Footnote 76}
3. Were more likely to have already been saving for PSE before the CESP was introduced in 1998:
   1. Data from the 1999 SAEP showed that 62.6% of families with children less than 18 years of age and with a household income of $80,000 or more reported having saved for their child’s PSE compared to 18.7% for those with a household income of less than $30,000 (note that RESP take-up was 15% among all families in 1999). ^{Footnote 77}
   2. Given that the 1999 SAEP was completed shortly after the introduction of the CESP in 1998, it is reasonable to assume that for most of these children saving began prior to 1998. Hence, a majority of these families with a household income of $80,000 or more would not have required encouragement from the CESP to save for PSE. Instead, most of these families saving for PSE outside of an RESP likely changed the way they saved and used RESPs to take advantage of the grants being offered via the CESG.
      1. In-depth analysis of the 1999 SAEP confirmed these points by revealing that 60.0% of families with a household income of $80,000 or more and whose youngest child was 17 or 18 years of age had PSE savings in 1999, while this proportion was 58.2% for families with a similar income and with children aged between 0 and 4 years old. However, RESP participation was only 11.8% among the former group while it was 33.6% for the latter group. Due to program rules ^{Footnote 78} , the bulk of children who were 17 or 18 years old in 1999 were never eligible for the CESG.

The above points indicate that families with higher incomes can save, are aware of the importance of saving, and already were saving for their children’s PSE before the CESP was implemented. In addition, as RESP take-up among families above the highest A-CESG income threshold increased to 63.8% in 2012, it appears that much of the increase in RESP take-up since 1999 is due to a change in the way they save for PSE.

Considering this, it may not be necessary to offer them financial incentives via the CESP as an impetus. The reason for this is that the inputs (i.e. grant funding in RESPs) will not really impact outputs (i.e. saving for the PSE of children) to the same extent as for those below the highest A-CESG income threshold, since PSE saving among families above the highest A-CESG income threshold was already taking place on a large scale prior to 1998. Moreover, the CESP survey indicated that in 2013 close to 75% of families with incomes of $90,000 or more indicated that they would still have saved for PSE using an RESP in the absence of the CESP.

That being said, there is likely still value in having the CESP, even for higher-income families, as some of them were likely induced to save more (and more regularly) for their children’s PSE. As well, there is little question that the existence of the CESP in combination with tax changes to improve the flexibility of the RESP (e.g. raising the RESP annual contribution limit per beneficiary from $1,500 in 1990 to $4,000 in 1997, and then removing the limit in 2007) have increased RESP take-up among families at all income levels, as the linked CRA-CESP data analysis demonstrated. However, as some studies have shown, a high percentage of savings generated from the introduction of a new savings incentive are often due to a substitution effect. ^{Footnote 79}

Given the evidence presented, a majority of families with higher incomes would be saving for their children’s PSE regardless of the existence of the CESP – it just so happens that now most of this saving takes place inside an RESP, due in part to the 20% grant on their RESP contributions. Thus, for these families there is little economy because the cost of resources used is much higher than is needed to achieve the expected outcome of saving for PSE.

Aside from efficiency and economy, there is also an issue of equity. The analysis of the merged CESP-CRA databases showed that those with $125,000 or more in family income (adjusted for inflation) received 31.9% of all CESP payments in 2012 even though they accounted for only 17.1% of all families with children less than 18 years of age (refer to Table 3). This amounted to at least $224 million in CESP payments in 2012. Even the introduction of the CLB and A-CESG did not change the proportion of CESP payments going to these families, as they received 31.8% and 32.9% of all CESP payments in 1999 and 2005, respectively. As the percentage of these families increased during this period, from 9.6% in 1999 to 17.1% in 2012, the discrepancy between the percentage of resources allocated to them and their share in the population has narrowed during this period. For families with less than $45,000 in family income, the discrepancy has narrowed as well during this period, as their share of CESP expenditures increased significantly.

Moreover, out of CESP payments made in 2012, 49.6% went to families with a family income of $90,000 or more. Assuming this proportion for 2013, this represents well over $400 million in 2013 CESP payments that were disbursed to these families, of which $280 million were disbursed to families earning $125,000 or more.

Families in one of the two lowest income categories in Table 3 received 24.3% of CESP payments in 2012 while accounting for 36.9% of all families with children less than 18 years old. This amounted to at least $171 million in CESP payments in 2012, up substantially from $37 million in 1999 and $72 million in 2005, representing a 362.2% increase over the 1999 to 2012 period. By comparison, CESP payments increased by 187.2% over this same period for families with $125,000 or more in family income.

For lower-income families, it appears to be efficient to help them save for PSE, as not many were saving for PSE prior to 1998. The evaluation acknowledges that it is more difficult for most families with an annual income of under $90,000 to save for their children’s PSE. Nevertheless, as a high number of these families do save for PSE, it is essential to underline the contribution of the CESP. Furthermore, it should be noted that even low-income families without savings can open an RESP and receive the CLB.

Several lines of evidence highlight the contribution of the CESP in helping to finance PSE savings in RESPs among middle- and lower-income families:

1. The CESP survey revealed that 67.6% of families with an annual income of less than $45,000 with an RESP for a child indicated that the CESP encouraged them to start saving earlier than they otherwise would have, while this proportion was 58% for families with an annual income of $45,000 to $89,999.
2. The percentage of families with less than $30,000 in household income and with PSE savings increased from 19.5% in 1999 (SAEP) to 45.7% in 2008 (ASETS), and declined slightly to 44.2% in 2013 (SAEP). ^{Footnote 80} Over the same period, the percentage of families with $30,000 to $49,999 in household income with PSE savings increased from 36.9% in 1999, to 56.5% in 2008, and fell to 53.4% in 2013. ^{Footnote 81}
3. The number of middle- and low-income families with children receiving the CESG has increased significantly since 1998 (see Section and continues to increase.
   1. The introduction of the CLB has led to a significant increase in RESP take-up among the lowest income families (although there was no evidence to attribute the growth in RESP take-up rates to the A CESG, as RESP take-up rates among families with lower incomes were already increasing prior to 2005 (refer to Section 4.3.1)).
4. Based on the SAEP/ASETS, the average RESP value (in constant dollars) for families having under $30,000 in household income increased from $3,812 in 1999 to $7,230 in 2012, while the average RESP value for families having $30,000 to $49,999 in household income increased from $3,800 in 1999 to $6,500 in 2012.

7.2 Efficiency of program delivery

This section examines the efficiency of the delivery model of the CESP. In terms of the delivery of the CESP (from an administrative standpoint), there appear to be some issues with the delivery of the A-CESG. The study using the merged CESP-CRA data indicated that 34% of A-CESG eligible families who made RESP contributions in 2012 did not receive the A-CESG in 2012 (likely either because their financial institution did not offer the A-CESG or because eligible families did not apply for the A-CESG). Similarly, about 23% of CLB-eligible families who had an RESP did not receive the CLB (for the same reasons as the A-CESG as outlined in Section 4.3). Although the government introduced a new form in 2013 that has addressed this issue for newly-opened RESP accounts, the problem still exists for many families who opened an RESP before 2013.

It is important to note that RESPs are not directly administered by the Government of Canada (aside from the CRA’s role in registering the plans). The CESP is delivered through an alternative service delivery arrangement with financial institutions, banks, mutual fund companies and scholarship foundations. The only CESP program delivery costs are salary and non-salary dollars pertaining to federal-level employees (other administrative costs such as IT, call centres, etc. are not included). The file review indicated that CESP operating costs rose from $3.7 million to $8.5 million from 1998 to 2013. Dividing the CESP operating budget by total CESP payments indicates that operating costs were 2.5% of total CESP payments in 1998 and 1.0% in 2013.

A comparison between CESP delivery efficiency and that of other ESDC strategic outcomes and programs was conducted in the file review. The results showed that CESP operating costs were comparable with programs falling under the Income Security strategic outcome (0.1%) and Social Development strategic outcome (2.1%) and with other programs such as Employment Insurance (9.7%) and the CSLP (3.1%). ^{Footnote 82} However, using this approach to measure program delivery efficiency is complicated by the fact that the CESP delivery model achieves considerable savings with its alternative service delivery model where financial institutions and group providers play a significant role in program delivery without cost to the Government of Canada. ^{Footnote 83}

Therefore, although the CESP appears to be among the most efficiently delivered ESDC programs using this metric, it should be kept in mind that the CESP is delivered via a unique alternative service delivery model (some of whom charge fees to account holders to cover their delivery costs – which does not impact the Government of Canada’s program delivery efficiency) compared to programs delivered exclusively by ESDC. ^{Footnote 84} This makes comparisons to other programs in which the Government of Canada plays a greater role in service delivery more difficult.

8.1. Recommendations

1. Explore ways for funds to more effectively reach families with the greatest need for assistance and encouragement to save for their children’s future post-secondary education.
2. Complement outreach efforts on promoting awareness with promoting understanding of the CESP savings incentives in order to support increased participation among all Canadian families.

Relevance: Continued need for the program

Relevance: Alignment with Government Priorities

Relevance: Alignment with Federal Roles and Responsibilities

Performance: Achievement of Expected Outcomes

Performance: Demonstration of Efficiency and Economy

ESDC (2012D), “Literature/File Review for the Summative Evaluation of the Canada Education Savings Program (CESP)”

The purpose of the literature and file review was to provide an overview of the information already available in the public domain and internal ESDC documents, completed surveys and research papers, and then compile the information (from over 70 documents) in order to provide evidence for a series of evaluation questions. This review demonstrated that much information was already available in the literature.

ESDC (2012C), “Impacts and Effects of the CESP on Family Savings: A Study Using the Access and Support to Education and Training Survey (ASETS)”

This study attempted to provide evidence on the impacts and effects of the CESP on family savings by using the 2002 Survey of Approaches to Educational Planning (SAEP) and 2008 ASETS, both of which were administered by Statistics Canada. The analysis was restricted to families with children under 18 years of age. The analysis was provided for different levels of income in order to compare the impacts of the CESP across income groups. The report provided information on families who had PSE savings and those which did not have PSE savings. The results were based on a sample of 7,545 parents with children aged 0-17.

ESDC (2012G), “PSE Students and RESP Savings Use: A Study Using the ASETS”

This study focused on the impact of the CESP on the affordability of PSE studies. The report also used data from Statistics Canada’s ASETS in order to determine what share of PSE funding came from RESPs and to what extent the CESP is improving the affordability of PSE. The selected sample used in this study was made up of 6,327 youth aged 18 to 24 years old.

Gray & McDonald (2012), “An Evaluation of the RESP program based on the National Graduate Survey”, Prepared for ESDC

This report used the 2000 and 2005 NGS of Statistics Canada, which surveyed students from public PSE institutions (universities, colleges and trade schools) who have graduated or completed the requirements for degrees, diplomas or certificates. The survey was designed to determine, for instance, the extent to which graduates of PSE programs were successful in obtaining employment; the relationship between the graduates' programs of study and the employment subsequently obtained; the graduates' job and career satisfaction; and the rates of under-employment and unemployment. The report included both descriptive statistics and regression analyses. The first part of the report focused on the incidence of RESPs as a funding source for PSE, while subsequent parts examined the impact of RESP funding on student loans.

ESDC (2012A), “Saving for PSE: Findings from the Canadian Financial Capability Survey”

The CFCS is a survey that was conducted by Statistics Canada in 2009, with the objective of shedding light on Canadians’ knowledge, abilities and behaviour concerning financial decision-making (i.e. how Canadians understand their financial situation, the financial services available to them, and their plans for the future). The study derived from the survey (sample size of 4,637) examines savings for PSE (including RESP savings) by individuals who are financially responsible for children under the age of 18 years old. The main purpose of the report was to identify the characteristics of PSE savers and measure the RESP take-up rate by socio-demographic characteristics.

ESDC (2012B), “Impacts and Effects of the CESP on Family Savings: A Study Using the Survey of Household Spending (SHS)”

The Statistics Canada SHS was designed to collect annual information about expenditures for consumer goods and services, changes in assets, mortgages and other loans, annual income, as well as information about dwelling characteristics and household equipment. The SHS is conducted annually from January to March of each year. For this study, the Public-Use Micro Files (PUMFs) were used, spanning from 1997 to 2009. Each year of the SHS was compared to ensure continuity in the variables. The focus of this study was mainly on household savings in general. It first examined whether families (especially low-income families) had adequate financial resources to generate savings and how they prioritized their spending. Then, it attempted to understand the possible impacts of the A-CESG and CLB on the net savings amounts and rates of eligible families.

Finnie & Wismer (2012), “Assessing the Use of Registered Education Savings Plans (RESPs) using Youth in Transition Survey (YITS) & Post-Secondary Education Participation Survey (PEPS)”, Prepared for ESDC

In this report, both the YITS (cohort A) and the PEPS were used for the analyses. The YITS-A follows a representative sample of youths starting in 2000 who were born in 1984. For this report, Cycles 3 and 4 (2004 and 2006) were used, as it is in these cycles where respondents (aged 18 to 21) had their first PSE experiences. The PEPS was conducted in February and March 2002 on youths between 18 and 24 years of age to collect information on education programs, student loans, and other PSE-related information. Using these surveys, this paper sought to answer a number of questions regarding savings behaviour of parents and how these were related to access to PSE and how PSE students funded their participation. 

ESDC (2013A), “Registered Education Saving Plans: Participants, Take-up and Trends”

This report also used the administrative databases to examine issues related to participation in the RESP program. Using a 10% random sample for the 1998 to 2011 period of 452,837 beneficiaries (and 288,461 subscribers), the study looked at the number of RESP participants, RESP take-up and trends from 1998 to 2011. It also examined the profile of RESP participants and looked at the age at which RESP accounts are opened.

ESDC (2013B), “Registered Education Savings Plans: Saving Habits”

This report used the administrative databases to examine issues related to saving habits of RESP users. A 10% random sample for the 1998 to 2011 period of RESP beneficiaries from the CESP administrative database was used (452,837 beneficiaries in the sample). The report mainly examined annual RESP contributions, lifetime savings and the effects of the A-CESG and CLB on saving.

ESDC (2013C), “Registered Education Savings Plans: Canada Learning Bond Recipients”

This paper examined the RESP saving behavior of low-income families that have received the CLB at least once. The paper used a 10% random sample of RESP beneficiaries from the CESP administrative database and examined the period between 2005 and 2011. Since the report focused on CLB recipients the sample was restricted to those recipients, which accounted for 41,963 observations. The report attempts to answer if CLB recipients were contributing their own resources to RESPs, and the extent to which families entered and exited low-income status.

ESDC (2013D), “Registered Education Saving Plan Withdrawals”

This report assessed the impacts and effects of the CESP on the savings of families. It attempts to answer how many students are withdrawing from RESPs and what is the trend and what portion of PSE costs are covered by RESP savings. The paper uses a 10% random sample of 483,334 RESP beneficiaries between 1998 and 2012, of which 113,226 have withdrawn from their RESP during their PSE studies.

R. A. Malatest & Associates (2013), “Case Studies with Low-Income Families”, Prepared for ESDC

The case studies involved conducting a total of 104 in-depth interviews with parents from low- and middle-income families. Low-income families were defined as families with a total annual household income of less than $40,000, while middle-income families were defined as families with a total annual household income of between $40,000 and $82,000. ^{Footnote 85} These interviews were equally divided between subscribers (i.e. those who have opened an RESP), and non-subscribers (drawn via random-digit telephone dialing). The primary purpose of the case studies was to document and assess the reasons for which some low-income families choose to save for PSE while others choose not to. In addition, the case studies provided evidence on the extent to which outside family members, relatives or friends are contributing to the RESPs of low-income families.

ESDC (2014D), “CESP Survey: Technical Evaluation Report”

A survey was conducted by R.A. Malatest and Associates with 2,015 families with children less than 18 years of age and children 18 and older who were attending a PSE institution, with a mixture of families with and without RESPs. The purpose of the survey was to gather information on four main issues: (i) why some families save for PSE using RESPs while others do not; (ii) the characteristics of those who save using RESPs versus those who do not; (iii) the impact of parental aspirations on RESP savings and children’s PSE participation; and (iv) the impact of RESP savings on PSE costs and student loan use. The survey was mainly conducted by telephone (or through the Internet) and had an average duration of approximately 20 minutes.

ESDC (2014C), “Examples of the Impact of RESP Withdrawals on Student Loan and Grant Amounts”

This report provided an initial examination of the impact of RESP withdrawals on student loan and grant amounts by presenting hypothetical examples that students could be confronted with when making RESP withdrawals. To determine the possible impact of RESP withdrawals on student financial aid for these hypothetical examples, Evaluation used the “Student Financial Assistance Estimator”, an online tool that it is provided by the department to help students obtain an estimate of the student financial aid (student loans and grants) that they might receive. ^{Footnote 86} Although these hypothetical examples do not measure the importance of the interaction in terms of dollar value, they clearly demonstrate the strong interaction between the CSLP and CESP.

ESDC (2015), “Analysis of the Canada Education Savings Program Participation and Expenditures for Different Income Groups”

The study examined how participation and program expenditures vary by income group, how the introduction of the A-CESG in 2005 affected the RESP participation of lower-income families, and examined if RESP contributions affect RRSP contributions. The analyses are based on a random sample of families living with children under 18. The sample was based on linked data from the CRA T1 and CCTB files, and CESP administrative data and used a 1% sample of families who are in the CCTB database between 1999 and 2012. These are families living with children under 18 and who are primarily responsible for these children and registered for the CCTB.

Internal studies

• ESDC (2011A), “Evaluability Assessment for the Summative Evaluation of the Canada Education Savings Program (CESP)”.
• ESDC (2012A), “Saving for PSE: Findings from the Canadian Financial Capability Survey”.
• ESDC (2012B), “Impacts and Effects of the CESP on Family Savings: A Study Using the Survey of Household Spending (SHS)”.
• ESDC (2012C), “Impacts and Effects of the CESP on Family Savings: A Study Using the Access and Support to Education and Training Survey (ASETS)”.
• ESDC (2012D), “Literature/File Review for the Summative Evaluation of the Canada Education Savings Program (CESP)”.
• ESDC (2012E), “Evaluation Framework for the Evaluation of the Canada Education Savings Program”.
• ESDC (2012F), “Evaluation Methodology for the Evaluation of the Canada Education Savings Program”.
• ESDC (2012G), “PSE Students and RESP Savings Use: A Study Using the ASETS”.
• ESDC (2013A), “Registered Education Saving Plans: Participants, Take-up and Trends”.
• ESDC (2013B), “Registered Education Savings Plans: Saving Habits”.
• ESDC (2013C), “Registered Education Savings Plans: Canada Learning Bond Recipients”.
• ESDC (2013D), “Registered Education Saving Plan Withdrawals”.
• ESDC (2013E), “RESP Contributions by Foreign-Born Parents: Using the Access and Support to Education and Training Survey (ASETS)”.
• ESDC (2014A), “Canada Education Savings Program (CESP) and Family Savings: Interim Evaluation Report”.
• ESDC (2014B), “Registered Education Saving Plan Contributions above the Maximum Grant Limit”.
• ESDC (2014C), “Examples of the Impact of RESP Withdrawals on Student Loan and Grant Amounts”.
• ESDC (2014D), “CESP Survey: Technical Evaluation Report”.
• ESDC (2015), “Analysis of the Canada Education Savings Program Participation and Expenditures for Different Income Groups”.

External studies contracted by ESDC

• Finnie and Wismer (2012), “Assessing the Use of Registered Education Savings Plans (RESPs) using Youth in Transition Survey (YITS) & Post-Secondary Education Participation Survey (PEPS)”, Prepared for ESDC.
• Gray and McDonald (2012), “An Evaluation of the RESP program based on the National Graduate Survey”, Prepared for ESDC.
• R. A. Malatest and Associates (2013), “Case Studies with Low-Income Families”, Prepared for ESDC.

Other studies

• Audet, Mathieu and Bele (2011), “Parental Savings for their Children’s Post-Secondary Education: Does Financial Literacy Matter?”, prepared for ESDC.
• Bank of Montreal (2013), “2013 Bank of Montreal Student Survey”, conducted by Pollara.
• Benjamin (2003), “Does 401(k) eligibility increase saving? Evidence from propensity score sub-classification”, Journal of Public Economics 87, pages 1259–1290.
• Benjamin and Smart (2011), “Do RESPs Increase Household Saving?”, Preliminary Draft.
• Benjamin and Smart (2012), “Savings Trends in Canada”, prepared for ESDC.
• Berger, Motte and Parkin (2009), “The Price of Knowledge 2009: Access and Student Finance in Canada”, 4th edition, Millennium Research Series, Canada Millennium Scholarship Foundation.
• Beshears et al. (2006), “The importance of default options for retirement savings outcomes: Evidence from the United States.” NBER, Working Paper #12009.
• Beverly, Clancy and Sherraden (2014), “Testing Universal College Savings Accounts at Birth: Early Research from SEED for Oklahoma Kids”, Centre for Social Development, Research Summary 14-08.
• Canadian University Survey Consortium (2012), “Survey of Graduating Undergraduate Students”, June 2012, Prepared by Prairie Research Associates.
• Choi, Laibson and Madrian (2004), “Plan Design and 401(k) Savings Outcomes”, Written for the National Tax Journal Forum on Pensions.
• Connolly and Kohler (2004), “The impact of superannuation on household saving”, Research Discussion Paper, Economic Research Department, Reserve Bank of Australia.
• Corak et al. (2010), “A Comparison of Upward and Downward Intergenerational Mobility in Canada, Sweden and the United States”, Paper Prepared for the 31st General Conference of The International Association for Research in Income and Wealth, St. Gallen, Switzerland, August 22-28, 2010.
• Dworak-Fisher (2008), “Encouraging Participation in 401(k) Plans: Reconsidering the Employer Match.” U.S. Bureau of Labor Statistics Working Paper 420.
• Dynarski (2004), “Who Benefits From the Education Savings Incentives? Income, Educational Expectations, and the Value of the 529 and Coverdell,” Harvard University, Kennedy School of Government & National Bureau of Economic Research.
• Engelhardt and Kumar (2007), “Employer Matching and 401(k) Saving: Evidence from the Health and Retirement Study.” Journal of Public Economics, 91(10): 1920-43.
• ESDC (2011B), “Summative Evaluation of the Canada Student Loans Program”.
• ESDC (2011C), “Final Report: Linking CSLP-CESP-LAD Data to Analyze Program Interactions”, prepared by CESP and CSLP Directorates.
• ESDC (2014E), “Canada Student Loans Program - Policy Manual”.
• Finance Canada (1998), “Building Canada for the 21st century: The Budget Plan 1998”.
• Finance Canada (1998), “Building Canada for the 21st century: The Canadian Opportunities Strategy”.
• Finance Canada (2004), “New Agenda for Achievement: The Budget Plan 2004”.
• Financial Services Authority of Great Britain (2005), “Measuring Financial Capability: An Exploratory Study”, University of Bristol, Personal Finance Research Centre.
• Gale, Iwry and Orszag (2005), “The Saver’s Credit: Expanding Retirement Savings for Middle- and Lower-Income Americans”, The Retirement Security Project, Number 2005-2.
• Human Resources and Skills Development Canada (2003), “Formative Evaluation of the Canada Education Savings Grant”.
• Human Resources and Skills Development Canada (2009), “Formative Evaluation of the Additional Canada Education Savings Grant and Canada Learning Bond”.
• Human Resources and Skills Development Canada (2014), “Canadian Education Savings Program Annual Statistical Review 2013”.
• Human Resources and Skills Development Canada (2014), News release, “Minister Bergen encourages families to apply for the Canada Learning Bond”, January 23, 2014.
• Informetrica (2008), “Review of Registered Education Savings Plan Industry Practices”.
• Kempson, Finney and Davies (2011), “The Child Trust Fund Findings from the Wave 2 Evaluation”, Personal Finance Research Centre, University of Bristol.
• Madrian and Shea (2001), “The Power of Suggestion: Inertia in 401(k) Participation and Savings Behavior”, Quarterly Journal of Economics, 2001, v116 (4, Nov), pages 1149-1187.
• Madrian (2012), “Matching Contributions and Savings Outcomes: A Behavioral Economics Perspective”, NBER Working Paper No. 18220.
• Mitchell, Utkus and Yang (2007), “Turning Workers into Savers? Incentives, Liquidity, and Choice in 401(k) Plan Design.” National Tax Journal, 60(3): 469-89.
• Office of the Superintendent of Financial Institutions Canada (2012), “Actuarial Report on the Canada Student Loans Program as at 31 July 2011”.
• Organization for Economic Cooperation and Development (2014), “Education at a Glance 2014: Organization for Economic Cooperation and Development Indicators”.
• Scotiabank Group (2014), “Canadian Parents Are Not Confident They Are Saving Enough for Their Children's Post-Secondary Education”, conducted by Nielsen from August 19 through August 27, 2014.
• Statistics Canada (1995-1999), “The Daily: University tuition fees”, Tuition and living accommodation costs at Canadian universities and colleges.
• Statistics Canada (2014), “Survey of Approaches to Educational Planning, 2013”, The Daily, October 29, 2014.
• Statistics Canada (2013), “Tuition and Living Accommodation Costs for Full-Time Students at Canadian Degree-Granting Institutions Survey”.
• TD Canada Trust Education and Finance (2011), “Parents struggle to save for kids’ post-secondary studies”.
• VanDerhei (2010), “The Impact of Automatic Enrollment in 401(k) Plans on Future Retirement Accumulations: A Simulation Study Based on Plan Design Modifications of Large Plan Sponsors”, EBRI Issue Brief, April 2010, No. 341.
• World Bank (2013), “Matching Contributions for Pensions: A Review of International Experience”.