The economic crisis has demonstrated how critical informed and effective decision-making is for

April 2014 RESEARCH BRIEF: Financial Education and Account Access Among Elementary Students Prepared by Corporation for Enterprise Development (CFED) for U.S. Department Treasury Are classroom education and in-school ing effective? The economic crisis has demonstrated how critical informed and effective decision-making is for economic well-being consumers, families and communities. This has intensified pressure on policymakers to design effective strategies to enhance capability general public, and focus is increasingly put on schools as delivery mechanism to teach education to children and youth. At least 36 states have personal finance standards that are required to be implemented as part high school curricula. Of se, 13 require a course in personal finance be taken and five include testing student knowledge as a condition graduation. However, research on effectiveness high school education courses is not promising. Studies on impact education requirements on economic behavior later in life is mixed at best, and no studies use rigorous experimental methods. Some observers have suggested that in fact high school is too late to be teaching education; instead personal finance needs to be taught earlier beginning as early as elementary school to better achieve stronger economic outcomes. Children (and adults) ten show evidence learning more from education if it is experiential and relevant to ir lives. In context education, combining education with access to an account with a institution may provide one such opportunity for experiential learning. The account could provide with a testing ground to practice skills y were learning in classroom, improving overall learning and retention curriculum. A number schools have added student or credit union branches on site to take deposits and encourage basic management, and although institutions have been encouraged by FDIC and NCUA to pursue partnerships with schools, it s estimated that only a small number schools participate in in-school savings programs. With support from U.S. Department Treasury, Corporation for Enterprise Development (CFED), Center for Financial Security at University Wisconsin-Madison (CFS), and OpportunityTexas partnered with local school districts and institutions to explore impact education and access on elementary-age children. The Assessing Financial Capability Outcomes (AFCO) project consisted field studies with 4 th and 5 th grade in two school districts Eau Claire, Wisconsin and Amarillo, Texas during 2011-2012 and 2012-2013 academic years. This study fers first rigorous test impact approximately five hours classroom-based education and access to a or credit union branch in school, both alone 1 Council for Economic Education. (2012.) Survey States 2011: state economic and personal finance education in our nation s schools. New York: Council for Economic Education. 2 For a review se studies, see Mandell, Lewis. (2009). Two Cheers for School-Based Financial Education. New York: The Aspen Institute Initiative on Financial Security. 3 OpportunityTexas is a joint venture Center for Public Policy Priorities and RAISE Texas. 1

and in combination, on : Financial knowledge Financial behavior, i.e., opening and using savings accounts Attitudes towards saving and institutions Key findings include: Classroom education improved knowledge and attitudes. After receiving five hours classroom-based education, demonstrated greater knowledge concepts, and se knowledge gains persisted one year later. Attitudes towards saving and institutions also improved: more believed it is easy to save money and that s fered services that are useful to m. In-school branches increased account ownership and improved attitudes toward ing. Access to an in-school or credit union branch significantly increased number who opened accounts and how actively se accounts were used. In-school branches also improved attitudes about saving and institutions. Pilot Design This study began in 2011-2012 academic year with 4 th and 5 th grade in Eau Claire, Wisconsin and continued in next school year with new 4 th grade class. In second year pilot, a second site was added with 4 th graders in Amarillo, Texas Independent School District (ISD). FIGURE 1: DEMOGRAPHICS OF PILOT SITE COMMUNITIES 4 Eau Claire, Wisconsin Population, 2010: 65,883 Median household income, 200-2011: $42,226 Uned Households, 2009: 5.4% Economically Disadvantaged Students, 2011-2012: 41% Amarillo, Texas Population, 2010: 190,695 Median household income, 200-2011: $44,69 Uned Households, 2009: 10.3% Economically Disadvantaged Students, 2011-2012: 6% 4 Sources: US Census Bureau State & County QuickFacts; CFED estimates uned households derived from a model based on 2009 FDIC Survey and 2005-2009 American Community Survey (ACS); Texas Education Agency, Performance Reporting and Wisconsin Department Public Instruction, Wisconsin s Information Network for Successful Schools. 2

While re were small differences between pilots in two sites, both used a staggered design in which classroom- or teacher-based cohorts were randomly selected to receive education during study, and or received it after study was completed. All were given an assessment or survey prior to start education program ( study period ) and again after program ended (before control group teachers began teaching program). The assessment included a 13-point literacy quiz and or questions to measure attitudes, beliefs and experiences with issues. The education program used in both Eau Claire and Amarillo was adapted from Council on Economic Education s Financial Fitness for Life curriculum, drawing primarily from lessons addressing savings, decision-making, and money management. 5 The lessons were delivered by classroom teachers in five or six lessons approximately 45 minutes each. So that teachers felt prepared to teach material to and to standardize education treatment, y received training on lessons and were provided with all lesson materials. In addition, approximately half schools in each school district had access to in-school or credit union branches. In Eau Claire, in-school branches had already been launched prior to study s implementation by Royal Credit Union (RCU) in six 13 elementary schools. RCU s School $ense program allows children to open a joint savings account with a parent or guardian and make deposits and withdrawals at school. The in-school branches are open once a week at lunchtime, and are encouraged to set savings goals and can earn small prizes for depositing. In Amarillo, Happy State Bank (HSB) operated Kids Bank branches in three Amarillo s 36 elementary schools prior to implementation this pilot. In addition, HSB and Amarillo ISD agreed to expand in-school ing program to an additional 15 randomly selected schools for a total 18 schools with access to Kids Bank branches during study period. HSB uses a similar model to RCU s School $ense program to operate ir Kids Banks: open a savings account, eir jointly with a parent or guardian or in ir name only, and can make deposits at school. Each time a student makes a deposit into ir account, he or she receives a small prize from treasure chest as an incentive to save. The Kids Banks were open once a week during study period, eir before school, at lunch or after school. In Eau Claire, in-school s were well-established, and 69% were ed prior to pilot. However, in Amarillo, this pilot provided first exposure to accounts in school for majority. In order to increase number who had HSB accounts during study period, pilot team randomly selected within ed schools to receive a $25 seed deposit if y opened a Kids Bank account. Approximately half in ed schools were selected to receive incentive through drawings held in ir math classrooms. 6 Implementing study for a second year in Eau Claire created an opportunity to follow up with initial group 4 th graders as 5 th graders, after all had received education. Students were administered assessment a third time, approximately a year after y received education, allowing an examination persistence any effects observed in first year. 5 Adjustments were made to curriculum and student assessment in Amarillo so that y covered majority standards for grade 4 in new K-8 personal curriculum standards Texas legislature passed in 2011. Texas school districts are not required to teach new standards until 2014-2015 academic year, so this pilot provided an early opportunity to test and get feedback on new standards. 6 Drawings held in classrooms were used to distribute incentive so that process selecting incentive fer recipients was transparent to and ir families. The $25 incentive was not funded by U.S. Treasury Department but was provided through private funds raised by OpportunityTexas through Amarillo Area Community Foundation. 3

deposits at school. Each time a student makes a deposit into ir account, he or she receives a small prize from treasure chest as an incentive to save. The Kids Banks were open once a week during study period, eir FINDINGS FROM THE ASSESSING FINANCIAL CAPABILITY OUTCOMES (AFCO) YOUTH PILOT before school, at lunch or after school. Key Findings In Eau Claire, in-school s were well-established, and 69% were ed prior to pilot. However, in Amarillo, this pilot provided first exposure to accounts in school for majority. In order to increase number who had HSB accounts during study period, pilot team randomly selected within ed schools to receive a $25 seed deposit if y opened a Kids Bank account. Approximately half in ed schools were selected to receive incentive through drawings held in ir math classrooms. 6 Knowledge Gains: Financial education appears to produce a large improvement in knowledge; on 13-point assessment, who received education improved by an average two points.additionally, with accounts show stronger effects in terms changes in learning. However, having a account was not randomly assigned, persistence so any effects observed that in already first year. had accounts or opened accounts during study are likely different than without Key a Findings account. No evidence was found suggestion that a branch in school has a direct impact on Knowledge Gains: Financial education appears to produce knowledge gained. Implementing study for a second year in Eau Claire created an opportunity to follow up with initial group 4 th graders as 5 th graders, after all had received education. Students were administered assessment a third time, approximately a year after y received education, allowing an examination a large improvement in knowledge; on 13- point assessment, who received Attitudes toward Saving and education Financial improved by Institutions: an average two points. Financial Additionally, with accounts show stronger education produced a modest increase in all perceptions effects in terms changes in learning. However, having a how easy it is to save money and an account even was larger not randomly increase assigned, in how so much believe s fer services that already that had are accounts useful or to opened m. accounts While during effect education was stronger, study are likely re different was than also evidence without a that account. No evidence was found suggestion that a having access to a or credit branch union in school has branch a direct impact had on an impact on attitudes. knowledge gained. Students Banked: Based on school-based Attitudes toward variation Saving and Financial access Institutions: to in-school Financial education ing, produced a modest likelihood increase in all a student perceptions how easy it is to save money and an even larger increase how much believe s fer services having a account was highly that are useful correlated to m. While with effect presence education was stronger, in-school re was also evidence services: that having access 64.3% to a in schools with in-school branches or credit union were branch ed, had impact compared on attitudes. to only 39.9% in schools without branches. Additionally, we estimate that number ed increased slightly during study period after received education Students Banked: and Based when on school-based variation had a in access to or in-school credit ing, union likelihood in ir a student school. having These a account was highly correlated with presence in-school services: 64.3% schools with in-school effects were found predominately branches in Eau were Claire, ed, compared where to only 39.9% in schools may without have branches. had a better Additionally, understanding we estimate that number ir ability to open accounts. ed increased slightly during study period after received education and when had a or credit union in ir school. These effects were found predominately in Eau Claire, where may have had a better understanding ir ability to open accounts. Account Use: Data on student accounts suggest that in-school ing access is related to more actively using accounts, although Account Use: sub-sample Data on student accounts suggest with that account in-school ing data access was is somewhat related to limited. more actively There was some evidence that weekly deposits using accounts, in although Eau Claire sub-sample were higher with following account data was somewhat education, limited. There but was some evidence results were not statistically significant that and weekly deposits average in Eau deposit Claire were data higher following was a relatively education, noisy but results measure, were not meaning statistically significant re was a great deal variation and in average deposit deposit data was activity. a relatively noisy measure, meaning re was a great deal variation in deposit 6 Drawings activity. held in classrooms were used to distribute incentive so that process selecting incentive fer recipients was Incentives Boost Take-up: In Amarillo, transparent to fer and ir families. The $25 $25 incentive was not resulted funded by in U.S. an Treasury 18% Department increase but was in provided through becoming ed, showing Incentives private that even Boost funds raised a Take-up: by Opportunity modest In Amarillo, Texas incentive through fer can Amarillo boost $25 Area incentive Community account resulted Foundation. take-up. in an 18% increase During planning becoming ed, This showing change in that assessment even a score modest is close incentive to a full can standard boost deviation account unit take-up. (a 0. effect During size), which planning is quite large this by pilot, most standards. school this pilot, school administrators expressed worry that, particularly from low-income families, administrators expressed worry that, particularly from low-income families, would withdraw incentive as soon 3 would withdraw incentive as it as was soon deposited as in it was account. However, this was not a common occurrence, according to account data on deposited in account. However, that received this was seed not FIGURE 3: CHANGES IN OUTCOMES OVER TIME FIGURE 3: CHANGES IN OUTCOMES FOR THE EAU CLAIRE a common occurrence, according deposit. to account TREATMENT GROUP OVER TIME data on that received seed deposit. Persistence over time: The 1.50 follow-up Persistence over time: The results followup survey with in Claire Eau were Claire remarkably were survey with in Eau 1.30 1.10 remarkably positive. We positive. surveyed We surveyed a year after y had received a year after y 0.90 education, and each outcome had received we looked at education, and each 0.0 improved between baseline and follow up and outcome we looked at Baseline First Follow-up Second Follow-up stayed at those elevated levels improved a year between later. baseline This Quiz Score (13pt) Student ed is encouraging evidence and follow persistence up and stayed at Spend money immediately (5pt) Easy to save (5pt) benefits education. those elevated levels a year Saving is for adults (5pt) Banks useful to you (5pt) later. This is encouraging evidence persistence benefits education. This change in assessment score is close to a full standard deviation unit (a 0. effect size), which is quite large by most standards. 8.50 8.00.50.00 6.50 6.00 5.50 5.00 FIGURE 2: FINANCIAL QUIZ SCORE Baseline Follow-up No Financial Education (605) Financial Education (98) 4 Implications Insight #1: Classroom-based education works. Financial education in schools, even in small amounts, does appear to increase knowledge and capability. These are encouraging findings for proponents classroom-based education. Positive results were found after integrating five or six short lessons into classroom instruction. However, training and support for teachers was critical to

Implications Insight #1: Classroom-based education works. Financial education in schools, even in small amounts, does appear to increase knowledge and capability. These are encouraging findings for proponents classroom-based education. Positive results were found after integrating five or six short lessons into classroom instruction. However, training and support for teachers was critical to success this pilot, and schools seeking to replicate se outcomes would need to be willing to invest time and resources necessary to provide similar support to ir teachers. Insight #2: In-school branches may be an effective way to introduce capability to children. This study provides some evidence that do learn more when given an opportunity to apply ir learning. As one teacher in Amarillo observed, with accounts seemed to be more engaged in lessons as account made what y were learning relevant. Access to in-school ing resulted in more with accounts, more actively using ir accounts, and improved attitudes towards saving and institutions. Partnerships between local institutions and school districts are a promising strategy to teach skills to children early in life, particularly when in-school ing programs can be combined with a education curriculum. Insight #3: Financial institutions could use guidance on children s savings marketplace. The field children s savings initiatives is growing rapidly, and institutions are eager to serve this expanding marketplace. With this expansion, re is a growing need for guidance on how institutions can design or modify products to serve market at scale. Providing accounts for children can be complicated. Banks ten run up against barriers, perceived or real, that hinder ir ability to open accounts for children. A complex patchwork federal and state rules and regulations make it difficult for s to understand questions like: What documentation is required to open an account? Can a third party (e.g., a school or nonprit) open accounts on behalf children? If so, is a Social Security Number or individual tax account number necessary? Financial institutions would be served by having a deeper and clearer understanding what is possible and permissable in this marketplace. Insight #4: Even A Modest Incentive Can Boost Account Ownership. The results from providing with a $25 incentive to open an account in Amarillo suggest that even a modest incentive can be effective in encouraging account take-up. A modest initial deposit has also proven helpful in or child savings programs as a way to generate excitement and interest in accounts among and parents alike. While this research is an important contribution to literature on education and access for youth, re remain a number unanswered questions about how to best incorporate personal instruction and access to ing services into schools, and about long-term impact se interventions. Future research should examine long-term outcomes associated with se interventions as well as test strategies for increasing ir impact and effectiveness. 5