Does Access to Bank Accounts as a Minor Improve Financial Capability? Evidence from Minor Bank Account Laws

Finance and Economics Discussion Series Federal Reserve Board, Washington, D.C. ISSN 1936-2854 (Print) ISSN 2767-3898 (Online) Does Access to Bank Accounts as a Minor Improve Financial Capability? Evidence from Minor Bank Account Laws J. Michael Collins, Jeff Larrimore, and Carly Urban 2021-075 Please cite this paper as: Collins,J.Michael,JeffLarrimore,andCarlyUrban(2021). “DoesAccesstoBankAccounts as a Minor Improve Financial Capability? Evidence from Minor Bank Account Laws,” FinanceandEconomicsDiscussionSeries2021-075. Washington: BoardofGovernorsofthe Federal Reserve System, https://doi.org/10.17016/FEDS.2021.075. NOTE: Staff working papers in the Finance and Economics Discussion Series (FEDS) are preliminary materials circulated to stimulate discussion and critical comment. The analysis and conclusions set forth are those of the authors and do not indicate concurrence by other members of the research staff or the Board of Governors. References in publications to the Finance and Economics Discussion Series (other than acknowledgement) should be cleared with the author(s) to protect the tentative character of these papers.

DOES ACCESS TO BANK ACCOUNTS AS A MINOR IMPROVE FINANCIAL CAPABILITY? EVIDENCE FROM MINOR BANK ACCOUNT LAWS J. Michael Collins1 Jeff Larrimore† Carly Urban‡ October 2021 Abstract Banking the unbanked is a common policy goal, but should this include access to bank accounts for minors? This study estimates how teenagers’ access to bank accounts affects their financial development. Using variation in state laws, we show policies that permit access to independentlyowned accounts increase account ownership at age 16 through age 19, although by age 24 those young adults are banked at similar rates to teens who grew up in states that do not allow minors to own accounts independently. Teens who had access to independently-owned accounts use fewer high-cost alternative financial services (like payday loans) through age 20—but are then more likely to use AFS, particularly check-cashing services, from age 21 through 24. Using credit records, we show that access to non-custodial accounts has no effects on credit scores in the short-run, but lower credit scores and more loan delinquencies at ages 21 through 24. While these state laws promote financial inclusion for teenagers, the young people who take on accounts may experience negative consequences in the longer run. JEL Codes: D14, D18, G18, G21, G28 Keywords: Unbanked, Financial Inclusion, Bank Regulation, Financial Capability 1 University of Wisconsin-Madison. E-mail: jmcollins@wisc.edu † Federal Reserve Board. E-mail: jeff.larrimore@frb.gov ‡ Dept. of Ag. Economics and Economics, Montana State University and IZA. Address: 208A Linfield Hall PO Box 172920 Bozeman, MT 59717. E-mail: carly.urban@montana.edu The authors thank Alfredo Burlando, Luke Reynolds, Chris Stoddard, Pierre Carl Michaud, and seminar participants at the University of Glasgow, HEC-Montreal, the BBVA EduFin Financial Education Summit in Buenos Aires, and the Economic Self-Sufficiency Policy Research Institute at the University of California-Irvine for helpful comments. Samantha Cleary provided excellent research assistance. Carly Urban’s work on this research was in part funded by the EduFin Center for Financial Education and Capability, an initiative by BBVA; the grant number is 4W7680. The analysis and conclusions set forth are those of the authors and do not indicate concurrence by other members of the research staff or the Board of Governors. 1

1. INTRODUCTION “I just recently got a job and I obviously need some type of savings or checking account but I want it completely out of my parents' reach.'' --Posting on Reddit.com Personal Finance Forum, September 8, 20182 Financial inclusion is a goal for many policies across the world, with intent to engage more people in the formal economy (Shah, Mullainathan, and Shafir 2012; Dupas et al. 2018). A lack of access to basic banking services and the inability to save and borrow at low costs may compound people's financial problems (Bertrand, Mullainathan, and Shafir 2004). Financial inclusion policies aim to expand the market for banking services to currently un- or under-banked consumers (Célérier and Matray 2019; Demirgüç-Kunt and Klapper 2013). Even within well-developed economies, there are some consumers who remain un- or under-banked for much of their lives (Rhine, Greene, and Toussaint-Comeau 2006). For adults in the US, banking access is common, but for teenagers, there are legal restrictions on owning a bank account before the age of 18. While teenagers can work and earn income, in many states they are only allowed to have accounts held jointly with a parent. This study examines the effects of state policies that allow teenagers to own bank accounts in their own name. Why would access to an independently-owned bank account as a teenager benefit young adults? Both custodial and independently-owned bank accounts allow a teenager to manage cash flows, transfer funds and to store savings. Independently-owned accounts allow teenagers to have more control, including keeping their earnings safe from account custodians (“completely out of my parent’s reach”). A non-custodial account also puts the onus on the teenager, as opposed to his or her parent, to manage money, creating opportunities for young people to learn financial capability. For example, after experiencing a penalty or fee, the minor may learn the importance of monitoring account balances and paying bills on time. Minor-owned accounts create a more direct relationship between the youth and the account provider than a custodial account, where that relationship is mediated by an adult. Finally, to the extent that teenagers have a positive first experience with the formal banking sector, they may have stronger preferences for using formal financial institutions in the future (Brown, Cookson, and Heimer 2019). Early life experiences could have lasting effects on how people develop attitudes about financial services (Malmendier, Tate, and Yan 2011). A direct relationship with financial services institutions at a young age may influence preferences in the longer run (Alan and Ertac 2018). However, there are also reasons that independent account ownership at young ages without parental guidance could have negative repercussions. Younger, first-time banking customers tend to have accounts with low balances, and incur costs both for financial institutions and consumers (Porteous 2015). Young adults are also more susceptible to making financial mistakes (Agarwal et al. 2009) and negative experiences could affect their banking preferences (Christelis, Dobrescu, and Motta 2020). Some regulators and banking institutions may see the risks and costs of bank accounts in the hands of immature consumers as being too great. It is important that bankers, and bank regulators understand the costs and benefits of minor-owned accounts for teenagers. 2Source: https://www.reddit.com/r/personalfinance/comments/9e9yfm/im_a_teen_17_who_wants_to_open_up_a_bank_account/ 2

Determining the causal effect of account ownership is challenging because there is selection by teenagers into bank account ownership. For example, minors with more affluent parents may be more likely to both have bank accounts and have better financial outcomes later in life. Bank account ownership is likely correlated with unobserved characteristics, such as parental motivation, generating omitted variable bias in any estimates of how owning an account is associated with other outcomes. This study overcomes the identification challenges by using changes in state laws that permit access to non-custodial bank accounts for people under age 18 in order to estimate how access to accounts is related to financial behaviors in young adulthood. Federal policies in the US only allow individuals age 18 or older to own bank accounts. Minors only have access to a custodial account if their parent or guardian is a co-owner until they turn 18. However, some banks are chartered under state laws rather than federal charters, and some states have changed their laws allowing state-chartered banks to permit minors to have bank accounts in their own name as young as at age 15. Our research design uses variation in state minor bank account regulations among state-chartered banks to estimate changes in access to accounts for minors, and subsequently, study the impacts of access to accounts as a youth on financial behaviors later in life. Using these state laws in a difference-in-difference framework, we use two datasets—the Survey of Income and Program Participation (SIPP) and the Federal Deposit Insurance Corporation (FDIC) National Survey of Unbanked and Underbanked Households—to estimate the effect of teenagers having access to an account on their having a checking or savings account, as well as using alternative financial services, as a young adult. We also use this strategy to examine credit behaviors among young adults using the Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP). This study estimates the causal effect of minor account access on financial behaviors to answer several research questions. First, do state laws allowing minor-owned accounts increase the likelihood that teenagers are banked? Second, state laws allowing minor-owned accounts result in greater financial inclusion among young adults beyond their teenage years? Finally, does access to independently-owned accounts impact young adults’ financial behavior including their use of credit and alternative financial services? This research contributes to three strands within the broader literature on banking access and household finance. First, we contribute to research examining the effects of access to banking while young on downstream financial behaviors. Brown and colleagues studied the effects of growing up on American Indian reservations with banking services (compared to those without) on credit use in adulthood (Brown, Cookson, and Heimer 2019). Their findings show that access to banking institutions while young accelerates the take up of credit and improves credit scores. We examine a different policy, state regulations of minor-owned accounts, among a broader population of young people. Second, we build on the literature on access to banking and the take up of accounts.3 Much of the existing research on bank access has focused on the location of branches. Célérier and Matray, for 3 There has been additional research regarding banking regulation in different contexts. Sun and Yannelis (2016) find that banking deregulation increases college-going due to credit constraints. Popov and Zaharia (2019) show that intrastate banking deregulation increased female labor force participation. Further, Gerardi, Rosen, and Willen (2010) and Hacamo (2020) discuss deregulation in the mortgage market. 3

example, show that interstate bank branch deregulation in the 1990s and early 2000s increased the number of bank branches in low-income areas, and in turn that shift increased the likelihood individuals were banked and as well as high levels of wealth later in life (Célérier and Matray 2019). Bank account access may not unambiguously positive, however, if individuals are not provided the knowledge and skills necessary to understand and use the accounts. Third, we broadly contribute to the literature determining the effects of banking regulations on bank account access. Bank access potentially improves the welfare of previously underbanked consumers, while also having positive externalities for communities and financial services providers by expanding the market (Barr 2004). For example, Washington finds that when states require banks offer low-cost bank account options, the proportion of low-income minority households who are unbanked decreases (Washington 2006). If minors have unmet demand for independently-owned accounts, we expect an increase in bank account ownership among young adults after states pass minor bank account laws. However, access for young people may have different implications than financial inclusion for all consumers. In credit card markets, for example, the CARD Act restricts access to credit cards for young adults aged 18 to 20 years old. However, Debbaut, Ghent, and Kudlyak (2016) show the policy reduces independently-owned credit cards and increases the ownership of co-signed credit cards. The authors conclude that blocking direct access to credit card markets delays young people from entering the credit market and raises their future costs of credit. The role of regulations in balancing access for young consumers into financial markets while also protecting them from harm remains an ongoing debate, including the rise of investment and stock trading technologies aimed at young investors (Welch, 2020). Our findings show that state laws allowing minor account ownership increase the likelihood that an individual is banked before age 18. While the effect of early access to independent accounts persists until individuals are age 20, the effect dissipates with age. We find that youth age 18 to 20 who had access to minor accounts were less likely to use alternative financial services, though these effects are not statistically different from zero. More importantly, these effects reverse with age. We show that access to minor-owned bank accounts increases use of costly alternative financial services— particularly check-cashing for 21 through 24-year-olds. Further, while access to minor accounts has no effect on young adult’s credit behavior in the first years of their financial independence, it increases delinquency, and reduces credit scores in subsequent years. The young people who obtain accounts in states with permissive minor-banking laws for state-chartered banks do not benefit from these policies. Young people who are motivated by the opportunity to access banking directly rather than through custodial accounts seem likely to place higher value on independence, perhaps eager to be on their own. This is a unique group of teenagers and young people, including some who are avoiding their parent’s financial problems (Britt, 2016). Accelerating the ability of teenagers to have an account on their own may seem like an effective financial inclusion strategy, but it may not be welfare enhancing in the longer run. To the best of our knowledge, this is the first study to estimate the effects of bank account access, unrelated to firm location, on a targeted cohort of account holders. While our focus is on how young people use bank accounts, we can also rule out that state minor account laws are associated 4

with shifts in the structure of financial services markets. We find no evidence to suggest that allowing state-chartered banks the ability to offer minor-owned accounts substantively changes the supply of state or federal banks in an area. In the next section we provide more details about banking regulations that create the variation we use to identify the effects of bank account access. We then describe the data and methods used to estimate the effects of access to minor banking laws, as well as a series of robustness exercises. We conclude with an overall discussion of these estimates and the implications of this study for policy. 2. BACKGROUND Understanding how youth access to bank accounts improves financial outcomes is an important piece of evidence to better design policies, regulations and financial products that benefit firms and consumers. 2.1 Laws Regulating Account Access Before Age 18 Financial institutions in the US are generally chartered at the national or state level. Each state is left to implement regulations for state-chartered institutions, including whether or not teenagers under 18 years of age can own an account without an adult co-owner. The result is that financial institutions operating in the same market may have different regulators and have to adhere to different rules (VanHoose 2017). In the absence of state legislation, state-chartered firms adhere to the federal policy for nationally-chartered institutions, which only allow accounts for minors using a joint account with a parent or guardian (a custodial account). Joint accounts operate like a minorowned account, where the young person can make deposits and withdrawals but is not an independent owner and the co-owner can also make deposits, withdrawals, or approve or not approve certain account actions. Custodial accounts are set up by an adult to be used for the benefit of a minor, but an adult owns the account at least until the child reaches age 18. The experiences of a young person in interacting with financial services are not likely to be as significant with these types of accounts.4 We collected data on minor account laws for this study (see Appendix A, Table A.1). Panel A of Figure 1 maps the states where teenagers are allowed to have independent accounts at statechartered banks, as well as the first year in which this was legally permitted. These laws begin in the early 1900s. By 1985, 19 states had laws allowing minors ages 15 and older to have their own checking or savings accounts, and this number grew to 29 states by 1999 and 45 states by 2017. We rely on variation over time in state minor banking law polices across states from 2000 to estimates the effects of state laws in this study. Since these policies only apply to state-chartered banks, it is important to see that state-chartered banks exist uniformly across states. In Panel B of Figure 1, we map every ZIP code where there is at 4 There are several additional legal issues related to accounts for minors. For example, there are regulations to protect minors from having their assets abused by an adult, as well as to protect them from their own potentially uninformed decisions (Rhine, Greene, and Toussaint-Comeau 2006; Office of the Office of the Comptroller of the Currency 2017). 5

least one state-chartered bank based on bank-level data from the FDIC from 2017. This map suggests that indeed, state-chartered banks exist in nearly all populated areas.5 Our thought experiment is as follows: a 16-year-old wants to open a checking account and wants the money to be hers and hers alone. Depending on the state and year, she is either able, or not able, to open the account on her own. That difference allows us to identify the effects of access to accounts. 2.2 Costs and Benefits of Minor Accounts There are several potential benefits and costs of having an independent account before turning 18. Most generally, there is a body of research that shows that access to basic financial services has positive effects on account holders. For example, Nguyen shows that bank branch closings decrease small business lending, with the effect most concentrated within six miles of the closed branch (Nguyen 2019). Berger and Seegert (2020) study the limitations on access to bank accounts among marijuana firms, finding the inability to access accounts harms the regulated businesses. Additionally, bank access is viewed as an important aspect of global anti-poverty efforts (Burgess and Pande 2005) and access to basic banking services can become a pipeline to more sophisticated credit and other financial products (see, for example, Agarwal et al. 2021). But a separate question is whether or not minors benefit from access to an independently-owned, non-custodial account. Teenagers can still get a bank account in states without minor-owned account provisions for state-chartered banks, but will have to share the management of these accounts with a parent or guardian. Custodial accounts could be viewed as having “guard-rails” that benefit teenagers. Or custodial accounts could limit teenagers from becoming more financially independent and capable young adults. Survey data show that young adults (age 18 to 22) are among the most likely to report they could not come up with $2,000 in the event of a financial emergency within the next month. They are also more likely to use high-cost non-bank borrowing, including payday loans, auto title loans, rent-toown stores, and pawn shops (FINRA Investor Education Foundation 2016). At the same time, young adults are the most likely age group to be unbanked, defined as not having checking, savings, or similar accounts used to deposit funds, store cash and make payments (Hayashi, Minhas, and others 2019; Addo 2014).6 Owning their own checking account may motivate teenagers to open bank accounts, and having these accounts may benefit them as they deposit paychecks, transfer funds, and smooth consumption over time. Previous research shows that young adults who live closer to financial institutions have stronger credit profiles later in life, at least in the unique context of banking access around Native American 5 FDIC Statistics on Depository Institutions Report as of September 2018 shows 1,159 total nationally-chartered institutions with depository insurance, and 4,318 state-chartered institutions. Nationally-chartered firms have about twice the assets of state-chartered firms, however. At the same time, NCUA reports 5,480 credit unions; of these 3,608 have national charters, with both types of charters holding roughly equal assets. 6 These types of transaction-based accounts are distinct from college savings or child savings accounts used to save for a specific, restricted purpose. See (Sherraden et al. 2013) for a discussion. 6

reservations (Brown, Cookson, and Heimer 2019).7 Turning to credit cards, Debbaut, Kudlyak, and Ghent (2016) find that the CARD Act substantially limited credit card access to those under age 21. The young adults who came of age before the CARD Act obtained their own credit cards sooner, had lower default rates, and ultimately had higher credit scores in their 20s. These studies suggest minor-owned accounts could have benefits for teenagers. For young people who are afraid that their parents or guardians will meddle in their finances, the prospect of having a custodial joint account may be especially problematic. These youth may have more to gain from being able to open independent bank accounts allowed under minor bank account laws. While there is scant research on the incidence of parents abusing their minor's financial accounts, one review is suggestive that, at least for some teenagers, the ability to have an account separate from their parents would be valuable (Postmus et al. 2020).8 Another benefit of access to minor-owned accounts for teenagers is that access to these accounts may facilitate the acquisition of financial literacy that persists into young adulthood. Two out of three young adults lack basic financial literacy (Lusardi, Mitchell, and Curto 2010; Lusardi and Mitchell 2014). The process of how people acquire financial knowledge is complex, ranging from peer norms and parental influences to formal education and learning by doing (Lusardi and Mitchell 2014). Cross-sectional data from the Programme for International Student Assessment (PISA) study by the Organisation for Economic Co-operation and Development show that youth who report having a bank accounts also show higher levels of measured financial literacy (Federal Deposit Insurance Corporation 2014; Jappelli 2010).9 Prior studies show that account ownership could facilitate financial learning. Jamison, Karlan, and Zinman (2014) conducted an experiment randomizing financial education only, financial education and account access, and account access only. They found that financial account access has some positive effects independent of other interventions, although there is stronger evidence that bank accounts and financial education are complements. From the institutional perspective, offering accounts for teenagers may be an attractive financial product for banks to offer mainly as a strategy to engage multiple generations of account holders, as parents with accounts have teenagers who open accounts (Lewis and Bingham 1991; Lewis 1982). Indeed, studies of the demand for youth accounts show parents and children often have accounts at 7 While these are promising findings in terms of the effects of bank access, the setting of that study is specific to Native American areas, which are unusual financial markets that are not readily generalizable to the estimated outcomes of young people's access to minor-owned accounts overall. 8 For example, one post in 2018 included the following statement: “I just recently got a job and I obviously need some type of savings and/or checking account but I want it completely out of my parents' reach. They've borrowed money from my siblings...not paying [it] back.” https://www.reddit.com/r/personalfinance/comments/9e9yfm/im_a_teen_17_who_wants_to_open_up_a_bank_acco unt/ 9 Students in many states already receive formal financial education in high school. Prior studies generally find positive effects of recent financial education graduation requirements on credit and other financial behaviors (Urban et al. 2020; M. Brown et al. 2016; Stoddard and Urban 2020; Harvey 2019). However, these courses do not always provide experience. Accounts may combine experience with formal learning in a complementary way to lead to greater financial capability and economic self-sufficiency. 7

the same institution.10 Access to accounts at younger ages may also help young people develop a stronger sense of trust with financial institutions (Brown, Cookson, and Heimer 2019; Horn, Jamison, Karlan, and Zinman 2020). Indeed, trust in banks has been shown to be a key factor that facilitates access to banking and the use of bank accounts (Traweek and Wardlaw 2020). The fact that minors can be allowed to obtain accounts could help young people develop positive perceptions of banks. However, if young people have a negative experience, for example with fees or overdrafts, that could reduce the level of trust that teenagers have with banks. A negative first experience with financial services could result in a shift in longer-run preferences and beliefs (Christelis, Dobrescu, and Motta 2020). In some cases, the minor may have had a better learning experience with their parent using a joint or custodial account. Custodial accounts at least have the role of a parent to deal with problems if they do occur. It is important to recognize that while state-chartered banks may be permitted to offer minor-owned accounts, these regulations to not mandate the quality of the accounts offered. Consumers can be especially sensitive to fees and penalties embedded into checking accounts (Melzer and Morgan 2021). Di Maggio, Ma, and Williams (2020) show that even bank practices like ordering withdraws based on high-to-low amounts rather than chronological order can drive people use alternative financial services. To the extent state minor banking laws result in lower quality, higher-cost financial services, these regulations may have adverse consequences for account holders (Begley and Purnanandam 2021). 3. EMPIRICAL STRATEGY We use a difference-in-difference strategy to identify the effect of minor account laws on account ownership. Our main dependent variable (Y) will equal one if the individual has a bank account (checking or saving) and zero otherwise. We will use variations of the below specification, though we use different datasets to understand different populations. We estimate forms of Equation (1) using a linear probability model (LPM). (1) 𝑌𝑌 {𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖} = 𝛼𝛼0 + 𝛼𝛼1𝑀𝑀𝑀𝑀𝐿𝐿 {𝑖𝑖𝑖𝑖} + 𝛼𝛼2𝑋𝑋𝑖𝑖 + 𝛼𝛼3𝑢𝑢 {𝑖𝑖𝑖𝑖} + 𝛼𝛼4𝐶𝐶𝐶𝐶𝐿𝐿𝐶𝐶𝑤𝑤 {𝑖𝑖𝑖𝑖} + 𝛽𝛽𝑖𝑖 + 𝛾𝛾𝑖𝑖 + 𝛿𝛿𝑖𝑖 + 𝜖𝜖 {𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖} The main coefficient of interest in Equation (1) is the difference-in-difference parameter, , which captures the causal effect of MBL (minor bank law) on whether or not individual i in state s born in year y and responding in survey year t is banked. We control for two individual-level demo𝛼𝛼g 1 raphic characteristics X that are unlikely to be affected by the policy: race/ethnicity and household type i (married couple, unmarried female head, unmarried male head, female individual, male individual, other).11 The model further includes the state unemployment rate at the time the individual was 15, whether or not the individual was exposed to a credit union minor account law,12 survey year fixed 10 See the review in Tank and Tyler (2005.) 11 If we do not control for household type and/or race/ethnicity, our results do not change. 12 While we technically control for the existence of a law allowing minors to have independent accounts at credit unions, only three states changed their laws since 2000—Michigan (2004), Iowa (2007), and Illinois (2012). Thirty-three states already had minor account laws for credit unions before 1999. Since we use state-level fixed effects, this means the control will only be included in specifications when we study downstream outcomes. 8

effects, state fixed effects, and birth year fixed effects. is the error term, and we provide robust standard errors clustered at the state level to account for both heteroskedastic standard errors and the fact that policies are set at the state level. We d𝜖𝜖o { 𝑖𝑖 n 𝑖𝑖𝑖𝑖 o 𝑖𝑖 t } find evidence to suggest that there are economic forces driving the passage of minor bank account laws in states.13 Difference-in-difference (DD) specifications require that the treatment and control groups are parallel in the pre-policy period and the treatment group would have trended similarly to the control group in the absence of the policy. Figure 5 shows suggestive evidence for the validity of the parallel trends assumption by comparing the trends in the pre-period across the treatment and control group for those 18-20 and 18-24. We show these event studies for each of the downstream outcomes: whether or not the individual has an account (Panel A) and whether or not the individual has used AFS (Panel B). In none of the four graphs is there evidence of a trend prior to the start of the policy. Recent developments in the DD model, particularly with staggered rollout of policies, suggests that using states that are always treated as a counterfactual could be problematic (Goodman-Bacon 2021). Thus, we provide robustness exercises in Appendix Table A.3. showing that the results are robust to dropping individuals in states where policies were always in effect. 4. DATA To estimate the effects of MBLs on account ownership and downstream behaviors, we rely upon three main datasets. First, we use the Survey of Income and Program Participation (SIPP) to gauge the first stage: do these laws increase account ownership among those eligible? Second, we determine how long the bump in account ownership lasts using the FDIC Un(der)-Banked Survey. The FDIC data also let us examine how the policy affects use of alternative financial services (AFS). Third, we examine how MBLs affect downstream credit outcomes using data from the Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP). 4.1 SIPP Data We begin with data from the SIPP to estimate the effect of increased access to minor-only accounts on account ownership for minors. While the SIPP is a short panel, we use the data as a repeated cross section of two survey waves: 2014 and 2018. Within these waves, there are observation periods from 2013, 2014, 2015, 2016, and 2017. While there are additional earlier waves, availability of bank account information began with the 2014 wave of data. We build a cross-sectional sample of 16-year-olds to maximize our sample size and not consider the same people at different ages.14 This strategy relies upon changes in Georgia (2017), Kansas (2015), Massachusetts (2014), New Hampshire (2015), and Oregon (2015). 13 See Appendix Table A.2 The only variable that is statistically different from zero at the 0.10 level is state gross product, where an additional $1 billion increases the likelihood of passing the policy by 0.1 percentage points. This marginal effect is quite small. 14 In the SIPP, some observations experience a change in year-of-birth across the panel. To be consistent, we take the first observation (the first time it is asked) and drop observations when year of birth is 2 or more years from the first year reported. 9

(2) 𝑌𝑌 {𝑖𝑖𝑖𝑖𝑖𝑖} = 𝛼𝛼0 + 𝛼𝛼1𝑀𝑀𝑀𝑀𝐿𝐿 {𝑖𝑖𝑖𝑖} +𝛼𝛼2𝑋𝑋𝑖𝑖 + 𝛽𝛽𝑖𝑖 + 𝛾𝛾𝑖𝑖 + 𝜖𝜖 {𝑖𝑖𝑖𝑖𝑖𝑖} In Equation (2), our main coefficient of interest is but we also include birth year fixed effects ( ), state fixed effects ( , and demographic characteristics for whether or not the individual identifies as non-Hispanic white and a dummy for 𝛼𝛼ge 1 n,der. 𝛽𝛽𝑖𝑖 𝛾𝛾𝑖𝑖) 4.2 FDIC Survey Data To determine how MBLs affect downstream account ownership and use of high-cost alternative financial services (AFS), we use data from the FDIC’s Un(der)-Banked Survey (henceforth, FDIC data).15 These data are collected every two years, as a supplement to the Current Population Survey, beginning with 2009. We use the data from 2009, 2011, 2013, 2015, and 2017 in this study. Across all waves, the FDIC survey asks about account ownership (checking and savings accounts) and alternative financial service use for each member of the household. We consider only those observations for whom the primary respondent in the household is 18- to 24-years-old. The survey further asks why individuals are unbanked (if they are). In the 2015 and 2017 data, where the question is asked identically, there are three main causes among those under 25: “I do not have enough money to keep in an account” (35%), “Account fees are too high” (11%), and “I do not trust banks” (10%). The percentages are only slightly different for those of older ages. Again, our main focus for this study is to understand how account ownership at early ages affects downstream financial outcomes. To do this, the FDIC data help us to understand whether or not those in states where account ownership increases see a continued increase in having an account once they reach adulthood. We also investigate whether or not they use alternative financial services. (See Figure 2 for visual tabulations of bank and under-banked status.) We are missing data on AFS use for 28 percent of the sample, who report that they do not know if they did or did not use AFS in the last year. This could be because they either do not want to admit to using it or because they have never used it and are unsure exactly what it means. In either case, we expect this to be uncorrelated with the policy and thus contribute to classical measurement error. 4.3 CCP Data In order to understand how MBLs affect downstream financial outcomes, such as starting one’s financial independence earlier, credit scores, and delinquency rates, we turn to data from the Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP). These data represent a 5 percent random anonymized sample of credit reports across the country, and also include any other household members of those who were selected for the random sample. Since the data are pulled from credit reports, we can observe only outcomes that would appear on a credit report, such as a measure of credit score, credit and debt accounts, and whether or not the individual is behind on an account. This means we cannot observe any demographic characteristics outside of age, such as marital status, gender, race/ethnicity, and education. We also cannot observe any activity on the 15 For more on these data, see https://www.fdic.gov/analysis/household-survey/. 10

market for payday loans or other forms of AFS, commonly referred to as “alternative credit reporting.”16 The data we use are from the fourth quarter of each year spanning from 2009-2017. Since the CCP are administrative data they provide a different perspective than the FDIC data. In addition to reducing measurement error based on self-reports, account ownership and AFS use are in a way, extreme outcomes. Since so many young adults are banked, account ownership may not fully reflect learning from early-life account access. Instead, learning through an independent account as a minor may allow those just starting their financial independence to make smarter financial decisions across the board. This may result in having a credit report relatively earlier, having a higher credit score, and being less likely to be behind on accounts. The CCP data allow us to explore this. 4.4 Additional Data In addition to our three primary datasets, in Appendix B we investigate supply-side responses to the minor account laws using data from the FDIC on the number of state-chartered and federally chartered banks by state and year. Also, data from the Bureau of Labor Statistics on state-level unemployment rate is included as a control in several estimates. 5. RESULTS Do minor account laws affect bank account ownership among minors? Does access to an account impact young adult banking and credit behavior? This section discusses the results of our estimates and provides evidence on the downstream outcomes after minors become young adults, including account ownership, alternative financial services use, credit scores, and loan delinquency. 5.1 Minor account laws increase account ownership In Exhibit 2, data from the SIPP show that among 16-year-olds, state policies allowing minors to own independent bank accounts increase the likelihood that those eligible have accounts. MBL increases the likelihood that 16-year-olds have a checking or savings account by 8.1 percentage points (Column 1), which is about a 29 percent as a marginal effect relative to the mean rate of 28 percent banked. Column (2) further shows that MBL increases the likelihood of having over $100 saved in a checking or savings account by about 8.8 points.17 The $100 threshold suggests that the individual did not simply open the account with a minimum balance and not use it. Column (3) shows that the MBL policy increases the likelihood of having an independently-owned “solo” account. Specifically, the MBL shows a 5.4 percentage point increase in solo-owned accounts, which is 41 percent of the mean rate of 13 percent.18 Column (4) shows that the MBL policy may shift minors away from joint accounts, though the magnitude is small (0.2 percentage points) and not 16 Others have referred to these agencies as “specialty consumer reporting companies.” 17 Mean savings among those with a checking or savings account is $418. 18 The reason that average is not zero before the start of the policy is because nearly all states allow state-chartered credit unions to offer independent accounts to minors. 11

statistically different from zero. We take this as evidence that the policy generates an overall increase in account ownership among minors, especially solo minor-owned accounts. The findings from the SIPP data provide evidence that there is a first stage effect of MBL: minor account laws increase account ownership among minors.19 To validate this even further, we investigate the effect of the MBL on having a checking or savings account among 18 and 19-yearolds in the FDIC data, as displayed in Exhibit 3. Restricting the sample to the ages where an individual reaches adulthood (age 18) provides additional evidence of a first stage from the prior estimates. Indeed, we find that the MBL policy increases the likelihood of having an account by 8 percentage points in Column (1) among 18 to 19-year-olds. While this seems on par with the magnitude in the SIPP, the mean account ownership is much higher among 18 and 19-year-olds (90 percent compared to 28 percent). Thus, the effect is economically and statistically significant, though smaller size than for 16-year-olds, at only 9 percent of the mean. All of these results are robust to dropping states that were always treated (as shown in Appendix A, Table A.3). We further show the two-way decomposition of the has account results in Table A.7, where all of the average estimates for each treatment and comparison group are positive and economically significant (Goodman- Bacon 2021). 5.2 Minor account laws and downstream financial behavior We have documented an increase in account ownership after MBL policies allow minors to have independent accounts. We next seek to understand if this early-life access to accounts provides an experiential learning opportunity such that they make stronger financial decisions in the longer-run. One important question is: do those early accounts result in higher rates of account ownership at older ages? Using the FDIC data, when we expand our sample to those 18 through 24 years of age. In Table 2 Column (2), the effect of the policy for 18 to 20-year-olds is one-quarter the size of the effect for 18 and 19-year-olds and is no longer statistically different from zero. Further, Column (3) shows that changing the ages for the sample to 21 through 24 flips the sign of the estimate, although it is not statistically different from zero. These findings suggest that others who did not experience increases in account ownership due to the minor account allowance caught up once they reached adulthood. It is possible that this difference in effect size could be that there is heterogeneity in the effects by state. While the states being used to identify the variation in Columns (1) and (2) are the same, fewer states identify the variation in Column (3). If we restrict the sample to in Column (1) to include only states that identify the effects in Column (3), our effect size actually becomes larger (0.147 with standard error of 0.048). The finding is perhaps not surprising, since account ownership is extremely common even among 18 to 20-year-olds (89 percent have accounts). It could even suggest that those who held accounts jointly with parents started off stronger financially as young adults. A supplemental analysis in Figure 3 plots coefficients and 95% confidence intervals for the effect of minor account laws on whether or not an individual is banked by each age group through age 29. These estimates confirm that the estimated effect falls to zero by age 20 and becomes a more precisely estimated null at later ages. 19 It could be that MBLs affect youth employment by making it easier to have one’s own money. We check this in Table A.4, and though the effects are positive, they are not statistically different from zero. 12

Though we see no change in the likelihood of having an account as individuals age into their early 20s, it could still be that there is some learning from early accounts that shifts them away from making financial decisions that are potentially costly. In Columns (4) and (5) of Table 2, we seek to understand how the MBLs affect use of alternative financial services, which arguably are a substitute for being banked. When focusing on 18 through 20-year-olds in Column (4), we find that the MBLs decrease the use of AFS in the last twelve months by 7.7 percentage points, or 21 percent relative to the mean rate of AFS use of 36 percent. While the coefficient is large, it is not statistically different from zero at the 90% level; we are limited in our sample size because the 2009 survey wave asked the AFS question differently (ever versus in the last year). The magnitude of the coefficient suggests that early in life account access may potentially shift young adults away from high-cost borrowing. However, the final column shows that this effect flips sign and remains statistically indistinguishable from zero when we consider those 21 through 24 years of age. Thus, the relationship changes relatively quickly. We further probe these AFS results by type of AFS used in Figure 4 to determine whether or not the noisy coefficient comes from one type of product. In Panel A, we plot the coefficients and 95% confidence intervals for 18-20-year-olds and in Panel B we do the same from 21-24-year-olds. We see that for the younger sample, the noisy decrease in AFS use comes from a reduction in money orders, though it is still not statistically different from zero. However, there is a clear increase in the use of check cashing services for the older sample. Young adults who accessed accounts early were more likely to be banked at higher rates, which also put them at risk of making early-life account management mistakes, as well as opportunities to learn from these experiences. Supplementary estimates show that these results are robust to dropping states that were always treated (Appendix Table A.5) and adding state-specific linear trends (Appendix Table A.6). The decomposition for the banked outcome (Appendix Table A.7) indicates that the bulk of the average estimates are positive for 18-20 year olds and negative for 21-24 year olds. The substantive comparison comes from treated states versus always treated states and treated states versus never treated states (Goodman-Bacon 2021). To provide further evidence testing whether or not individuals are indeed improving their financial situations as they develop their financial independence, we examine the effects of the MBLs on credit and debt behaviors using the CCP data. In the first column of Table 3, we see if MBLs change the age at which one has their first credit report. For this specification, we create a cross section of over 3.7 million observations that were ages 18-37 in 2017, indicating at what age each had their first credit report. Since having a credit report suggests that the individual had sufficient credit activity to generate a credit file, this measure reflects early experience with credit and debt. This finding suggests that the MBL decreases the age at which one had a credit file, though the magnitude is small (0.19 years) relative to the mean of 20.6 years for the first incidence of credit. If experience with a bank account is a valuable financial experience for minors, we may expect financial decisions to be reflected in higher credit risk scores (where higher scores predict lower risk). In Table 3, we find the effects are very small in magnitude in the case of 18 to 20 year olds on both credit scores and delinquency behaviors (Columns (2) and (4)), and not statistically different from zero. Among 21-24 year olds, the estimates on credit scores are statistically significant and 13

negative: access to minor accounts reduces credit scores by 2.8 points relative to a mean of 633. MBLs appear to drive financial decisions that reduce credit scores. This is further supported by the fact that we also see increases in delinquency rates due to the MBL for those 21 through 24 years of age in the final column of Table 3. These magnitudes are not small: access to minor accounts increase the likelihood of being behind on account by 0.85 percentage points (5 percent). Taken with our results indicating that MBLs increase the use of non-bank check cashing services in the long-run, our evidence suggests that independent access to accounts for minors can have negative consequences. Early mistakes as teenagers may drive these consumers as young adults to seek out alternative financial services. In comparing our effects on AFS use from the FDIC and on credit scores from the CCP data, we note a few caveats. First, the CCP data do not include information on alternative credit agencies. Thus, AFS use would not be directly incorporated into a credit score. However, high-interest debt is likely to affect one’s participation with the formal financial sector and could indirectly affect credit scores. Second, the FDIC data do not include young adults who still live with their parents, since we only observe cases where the primary respondent in the household is 18- to 24-years-old. Hence, these results only reflect a subset of young adults in the United States. Finally, the FDIC data are self-reports, and as discussed earlier, 28% of the FDIC sample report they are unsure of whether or not they have used AFS in the last year. 6. CONCLUSIONS State laws allowing minors to own their own bank accounts increases account ownership among 16year-olds by 8 percentage points (or 29 percent as a marginal effect). This is consistent with increased financial inclusion for this targeted population. However, this increase fades out by age 24, when compared to others in the same state who did not have early minor-owned account access and across other states where minor-owned account access did not change. We find some evidence of decreased reliance on alternative financial service use early into financial independence. Unfortunately, this reduction in early-adulthood use of AFS is completely negated by detrimental financial behaviors only a few years later. From ages 21 through 24, adults who had access to accounts as minors were more likely to use higher-cost financial services, particularly check-cashing services, were more likely to be behind on accounts, and had lower credit scores than those who were not eligible for such minor accounts. While at first these findings appear to differ from the effects of the CARD Act on young adults (Debbaut, Ghent and Kudlyak 2016), the contexts are quite different. The CARD Act limited to credit cards for people aged 18 through 20 after 2009, largely in response to accusations of creditors engaging in predatory marketing of cards to students with low income. The law resulted in reduced independent credit access and worse credit outcomes for young adults relative to prior age cohorts who came of age with access to independently-owned credit cards before 2009.20 Minor-account policies allow teenagers who are not yet 18 to have a bank account on their own at a state-chartered bank. Teenagers can get bank accounts with a custodian or joint owner in states without minor banking laws, just as 18 to 20-year-olds can get a credit card with a co-signer after the CARD Act. The developmental stage, in terms of education, experience and independence, for a 16-to-17- year- 20 Debbaut, Ghent and Kudlyak note in their paper that the longer-run outcomes of the CARD Act still need to be evaluated; their conclusions are primarily related to reduced access to credit. 14

old is likely to be quite different from that of an 18-to-20 year-old. Bank accounts also serve a very different function from credit cards. Credit cards require an application process and underwriting by a lender, while bank accounts only require basic identification. This results in selection into bank accounts in states with minor bank laws by teenagers who want to have independent accounts—as these people turn 18 they might be the types applicants that credit card issuers would deem as being too risky, denying them access to credit prior to the CARD Act. Bank accounts lack the same kinds of screening mechanisms used for credit cards—since there is not a strong screening mechanism teenage account holders seeking non-custodial accounts may be among those most prone to financial problems. The advent of FinTech and online financial transactions have increased the focus on how much independent access to financial services consumers should have (Welch 2020). Media attention has raised scrutiny about young people’s ability to actively manage investments.21 Nevertheless, teenagers appear to be able to access an ever-expanding set of financial instruments, including access to stock trading.22 Young adults appear to be eager to engage in even riskier financial products and investments (Hasso, Müller, Pelster, and Warkulat 2021). This study suggests that regulators should think twice before expanding independent access to financial products and potentially should repeal some existing expansions, especially for consumers younger than age 18, but perhaps even for young adults. Stricter screening and monitoring of the use of products by young participants in financial products may have benefits for both firms and consumers. 21 See for example: Megan Mccluskey Time June 24, 2021 'Someone's Going to Be Left Holding the Bag.' How Finance TikTok Is Navigating 'Meme Stock' Hype Among Young Investors https://time.com/6073524/meme-stock-tiktok/ 22 Maggie Fitzgerald “Now teenagers can trade stocks with Fidelity’s new youth investing accounts” May 18 2021 https://www.cnbc.com/2021/05/18/now-teenagers-can-trade-stocks-with-fidelitys-new-youth-investing-accounts.html 15

References Addo, Fenaba R. 2014. “Debt, Cohabitation, and Marriage in Young Adulthood.” Demography 51 (5): 1677–1701. Agarwal, Sumit, John C. Driscoll, Xavier Gabiax, and David Laibson. 2009. “The Age of Reason: Financial Decisions over the Life Cycle and Implications for Regulation.” Brookings Papers on Economic Activity 2009 (Fall): 51-101. Agarwal, Sumit, Thomas Kigabo, Camelia Minoiu, Andrea Presbitero, and André F. Silva. Serving the Underserved: Microcredit as a Pathway to Commercial Banks. No. 16013. CEPR Discussion Papers, 2021. Alan, Sule, and Seda Ertac. 2018. “Fostering Patience in the Classroom: Results from Randomized Educational Intervention.” Journal of Political Economy 126 (5): 1865–1911. https://doi.org/10.1086/699007. Barr, Michael S. 2004. “Banking the Poor.” Yale J. on Reg 21: 121. Begley, Taylor A., and Amiyatosh Purnanandam. "Color and credit: Race, regulation, and the quality of financial services." Journal of Financial Economics 141, no. 1 (2021): 48-65. Berger, Elizabeth, and Nathan Seegert. "Half banked: the real effects of financial exclusion on firms." Unpublished Working Paper, Cornell University (2020). Bertrand, Marianne, Sendhil Mullainathan, and Eldar Shafir. 2004. “A Behavioral-Economics View of Poverty.” American Economic Review 94 (2): 419–23. Britt, Sonya L. "The intergenerational transference of money attitudes and behaviors." Journal of Consumer Affairs 50, no. 3 (2016): 539-556. Brown, James R., J. Anthony Cookson, and Rawley Z. Heimer. 2019. “Growing up without Finance.” Journal of Financial Economics 134 (3): 591–616. https://doi.org/10.1016/j.jfineco.2019.05.006. Brown, Meta, John Grigsby, Wilbert van der Klaauw, Jaya Wen, and Basit Zafar. 2016. “Financial Education and the Debt Behavior of the Young.” Review of Financial Studies 29 (9). Burgess, Robin, and Rohini Pande. "Do rural banks matter? Evidence from the Indian social banking experiment." American Economic Review 95, no. 3 (2005): 780-795. Célérier, Claire, and Adrien Matray. 2019. “Bank-Branch Supply, Financial Inclusion, and Wealth Accumulation.” Review of Financial Studies 32 (12): 4767–4809. Christelis, Dimitris, Loretti I. Dobrescu, and Alberto Motta. 2020. “Early Life Conditions and Financial Risk-Taking in Older Age.” Journal of the Economics of Ageing 17 (C). https://doi.org/10.1016/j.jeoa.2020.100266. Debbaut, Peter, Andra Ghent, and Marianna Kudlyak. "The CARD act and young borrowers: The effects and the affected." Journal of Money, Credit and Banking 48, no. 7 (2016): 1495-1513. Demirgüç-Kunt, Asli, and Leora Klapper. 2013. “Measuring Financial Inclusion: Explaining Variation in Use of Financial Services across and within Countries.” Brookings Papers on Economic Activity 2013 (1): 279–340. Di Maggio, Marco, Angela T. Ma, and Emily Williams. In the Red: Overdrafts, Payday Lending and the Underbanked. No. w28242. National Bureau of Economic Research, 2020. Dupas, Pascaline, Dean Karlan, Jonathan Robinson, and Diego Ubfal. 2018. “Banking the Unbanked? Evidence from Three Countries.” American Economic Journal: Applied Economics 10 (2): 257–97. Federal Deposit Insurance Corporation. 2014. “FDIC National Survey of Unbanked and Underbanked Households.” Washington, DC. FINRA Investor Education Foundation. 2016. “Financial Capability in the United States-Report of Findings from the 2015 National Financial Capability Study” July. 16

Gerardi, Kristopher S., Harvey S. Rosen, and Paul S. Willen. 2010. “The Impact of Deregulation and Financial Innovation on Consumers: The Case of the Mortgage Market.” Journal of Finance 65 (1): 333–60. https://doi.org/10.1111/j.1540-6261.2009.01531.x. Goodman-Bacon, Andrew. "Difference-in-differences with variation in treatment timing." Journal of Econometrics (2021). Hacamo, Isaac. 2021. “The Babies of Mortgage Market Deregulation.” The Review of Financial Studies 34 (2): 907–48. https://doi.org/10.1093/rfs/hhaa073. Harvey, Melody. 2019. “Impact of Financial Education Mandates on Younger Consumers? Use of Alternative Financial Services.” Journal of Consumer Affairs 53 (3): 731–69. Hasso, Tim, Daniel Müller, Matthias Pelster, and Sonja Warkulat. "Who participated in the GameStop frenzy? Evidence from brokerage accounts." Finance Research Letters (2021): 102140. Hayashi, Fumiko, Sabrina Minhas, and others. 2019. “Who Are the Unbanked? Characteristics Beyond Income.” Economic Review 014 (4): 23–40. Horn, Samantha, Julian Jamison, Dean Karlan, and Jonathan Zinman. Does lasting behavior change require knowledge change? Evidence from savings interventions for young adults. No. w28011. National Bureau of Economic Research, 2020. Jappelli, Tullio. 2010. “Economic Literacy: An International Comparison.” The Economic Journal 120 (548): F429--F451. Lewis, Barbara R, and Graham H Bingham. 1991. “The Youth Market for Financial Services.” International Journal of Bank Marketing 9 (2): 3–11. Lewis, Barbara R. 1982. “Student Accounts---A Profitable Segment?” European Journal of Marketing 16 (3): 63–72. Lusardi, Annamaria, and Olivia S Mitchell. 2014. “The Economic Importance of Financial Literacy: Theory and Evidence.” Journal of Economic Literature 52 (1): 5–44. Lusardi, Annamaria, Olivia S Mitchell, and Vilsa Curto. 2010. “Financial Literacy among the Young.” Journal of Consumer Affairs 44 (2): 358–80. Malmendier, Ulrike, Geoffrey Tate, and Jon Yan. 2011. “Overconfidence and Early-Life Experiences: The Effect of Managerial Traits on Corporate Financial Policies.” Journal of Finance 66: 1687–733. https://doi.org/10.1111/j.1540-6261.2011.01685.x. Melzer, and Donald P. Morgan. "Who Pays the Price? Overdraft Fee Ceilings and the Unbanked." Overdraft Fee Ceilings and the Unbanked (June 2021). FRB of New York Staff Report 973 (2021). Nguyen, Hoai-Luu Q. 2019. “Are Credit Markets Still Local? Evidence from Bank Branch Closings.” American Economic Journal: Applied Economics 11 (1): 1–32. https://doi.org/10.1257/app.20170543. Office of the Office of the Comptroller of the Currency. 2017. “School-Based Bank Savings Programs: Bringing Financial Education to Students.” Community Developments, March, 1–10. Popov, Alexander, and Sonia Zaharia. 2019. “Credit Market Competition and the Gender Gap in Labor Force Participation: Evidence from Local Markets.” European Economic Review. https://doi.org/10.1016/j.euroecorev.2019.02.009. Porteous, David. 2015. “Low-Balance Bank Accounts---Part 1: Is There a Pathway towards Profitability?” Journal of Payments Strategy & Systems 9 (3): 305–10. Postmus, Judy L, Gretchen L Hoge, Jan Breckenridge, Nicola Sharp-Jeffs, and Donna Chung. 2020. “Economic Abuse as an Invisible Form of Domestic Violence: A Multicountry Review.” Trauma, Violence, & Abuse 21 (2): 261–83. Rhine, Sherrie L W, William H Greene, and Maude Toussaint-Comeau. 2006. “The Importance of Check-Cashing Businesses to the Unbanked: Racial/Ethnic Differences.” Review of Economics and Statistics 88 (1): 146–57. 17

Shah, Anuj K, Sendhil Mullainathan, and Eldar Shafir. 2012. “Some Consequences of Having Too Little.” Science 338 (6107): 682–85. Sherraden, Margaret, Clark Peters, Kristen Wagner, Baorong Guo, and Margaret Clancy. 2013. “Contributions of Qualitative Research to Understanding Savings for Children and Youth.” Economics of Education Review 32: 66–77. Stoddard, Christiana, and Carly Urban. 2020. “The Effects of State Mandated Financial Education on College Financing Behaviors.” Journal of Money, Credit, and Banking 52 (4): 747–76. Sun, Stephen Teng, and Constantine Yannelis. 2016. “Credit Const‐raints and Demand for Higher Education: Evidence from Financial Deregulation.” Review of Economics and Statistics 98 (12–24). https://doi.org/10.1162/REST_a_00558. Tank, Janki, and Katherine Tyler. 2005. “UK Student Banking Revisited: Influences and the Decision-Making Process.” Journal of Financial Services Marketing 10 (2): 152–64. Traweek, Virginia, and Malcolm Wardlaw. "Societal Trust and Financial Market Participation: Evidence from the Freedman's Savings Bank." Available at SSRN 3164418 (2020). Urban, Carly, Maximilian D Schmeiser, J Michael Collins, and Alexandra Brown. 2020. “State Mandated Financial Education and the Credit Behavior of Young Adults.” Economics of Education Review 78 (October): 101786. VanHoose, David. 2017. “Regulation and the Structure of the Banking Industry.” In The Industrial Organization of Banking, 243–80. Springer. Washington, Ebonya. 2006. “The Impact of Banking and Fringe Banking Regulation on the Number of Unbanked Americans.” Journal of Human Resources 41 (1): 106–37. Welch, Ivo. "Retail raw: Wisdom of the Robinhood crowd and the Covid Crisis." Available at SSRN 3696066 (2020). Welch, Ivo. The Wisdom of the Robinhood crowd. No. w27866. National Bureau of Economic Research, 2020. 18

Tables and Figures Figure 1: Maps of State Policies and Presence of State-chartered Bank Panel A: State policies Panel B: ZIP Codes that have state-chartered banks Notes: Panel A maps the year in which states allowed minors to have independent accounts in state-chartered banks. There are five states with no allowances. These data are also in Appendix Table A.1. Panel B depicts the ZIP codes in which at least one state-chartered bank operates. With the exceptions of Alaska, Arizona, Hawaii, Nevada, and Utah, nearly all ZIP codes have at least one state-chartered bank. 19

Figure 2: Summary Statistics of Minor Accounts Panel A: Average Account Ownership by Account Type (Age 16) Panel B: Average Account Ownership and AFS Use by Age (18-24) Notes: The data in Panel A are drawn from the SIPP 2014 and 2018 waves for a sample of 16 year-olds. Any Account equals one if the individual reported having a checking or savings account; Own Account equals one if the individual has an independent account; Joint Account equals one if the individual has an account with another person; Over $100 in Account equals one if the individual has over $100 in a checking or savings account. The data in Panel B are drawn from the FDIC (2009, 2011, 2013, 2015, 2017). Has Bank Account equals one for individuals if they have a checking or savings account from a bank or credit union. Used AFS equals one if the individual reported using alternative financial services in the last 12 months and zero otherwise. 20

Table 1: Overall Effects of Minor Account Laws on Account Use for 16-Year-Olds Has Over $100 Has Solo Has Joint Account Saved Account Account (1) (2) (3) (4) Minor Banking Law 0.0811* 0.0883** 0.0536* -0.0019 (0.0480) (0.0384) (0.0317) (0.0420) Mean DV 0.2779 0.0852 0.1329 0.1548 N 3,417 3,417 3,417 3,417 *Statistically significant at 10% level; ** at 5% level; *** at 1% level. Notes: Robust standard errors clustered at the state level in parentheses. Linear probability models estimated. All models estimate Equation (1). The data include 16-year-olds only. Data from the 2014 and 2018 waves of the SIPP. Table 2: Downstream Effects of Minor Account Laws on Account Ownership and AFS use Has Account Use AFS (1) (2) (3) (4) (5) Minor Banking Law 0.0811** 0.0233 -0.0214 -0.0773 0.0706 (0.0325) (0.0236) (0.0144) (0.0814) (0.0644) Mean DV 0.896 0.892 0.903 0.359 0.319 N 1,202 2,463 8,896 1,226 5,138 Ages 18-19 18-20 21-24 18-20 21-24 Notes: Robust standard errors clustered at the state level in parentheses. Linear probability models estimated. All models estimate Equation (1). Data from the FDIC Un(der) Banked Survey. “Has Account” equals one if the individual has a checking or savings account and zero otherwise. “Use AFS” equals one if the individual has used alternative financial services in the last 12 months and zero otherwise. 21

Figure 3: Effects of Minor Account Laws on Account Ownership by Age Notes: The data are drawn from the FDIC (2009, 2011, 2013, 2015, 2017). Each point is an estimate in Equation 1, restricting the sample to the age on the x-axis for the regression. We pool 18 and 19 year olds due to a smaller sample of 18 year olds. Banked equals one for individuals if they have a checking or savings account from a bank or credit union. 22

Figure 4: Effects of Minor Account Laws on AFS Use by Type Panel A: 18 through 20-year-olds Panel B: 21 through 24-year-olds Notes: Coefficients and 95% confidence intervals reported. Robust standard errors are clustered at the state level. Linear probability models estimated. All models estimate Equation (1). Data from the FDIC Un(der) Banked Survey. All dependent variables are measured as use in the last 12 months. Payday is payday loan use; Pawnshop is the use of a loan from a pawnshop; MO is a money order; Check casher is a check-cashing service; RTO is a rent-to-own service; Remittance is a remittance sent through a non-bank institution; Refund is a refund advance loan, such as an advance on a tax refund. Panel A is for 18 through 20-year-olds and Panel B is for 21 through 24-year-olds. 23

Figure 5: Event Studies Panel A: Has Account Ages 18-20 Ages 21-24 Panel B: Use AFS Ages 18-20 Ages 21-24 Notes: Robust standard errors clustered at the state level in parentheses. Linear probability models estimated. All models estimate Equation (1). Data from the FDIC Un(der) Banked Survey. Has account equals one if the individual has a checking or savings account and zero otherwise. Use AFS equals one if the individual has used alternative financial services in the last 12 months and zero otherwise. Panel A is for the dependent variable “Has Account,” and Panel B changes the dependent variable to “Use AFS.” 24

Table 3: Table of Downstream Effects of Minor Account Laws on Credit Outcomes Early Credit Equifax Risk Score Past Due (1) (2) (3) (4) (5) Minor Banking Law -0.188* -0.470 -2.801*** -0.00011 0.0085** (0.112) (0.917) (0.953) (0.0013) (0.0035) Mean DV 20.64 642.29 633.30 0.068 0.169 N 3,756,628 2,017,077 5,568,291 2,284,659 6,146,305 Ages 18-37 18-20 21-24 18-20 21-24 Years 2017 2009-2017 2009-2017 2009-2017 2009-2017 *Statistically significant at 10% level; ** at 5% level; *** at 1% level. Notes: Robust standard errors clustered at the state level in parentheses. All models estimate Equation (1) with data from the Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP). Early Credit is the first year the individual shows up in the credit file. For this specification, we use a large age range in one point of time (2017). Risk Score is the CCP’s measure for credit scores and ranges from 280 to 850. Past due equals one if the individual is behind on any account and zero otherwise. We estimate linear probability models for that variable. 25

Appendix A: Table A.1: State Policies State MBL Policy State MBL Policy Year Year AK 1993 MT 1977 AL 1980 NC 2012 AR 1997 ND 2009 AZ 1973 NE No policy CA 2012 NH 2015 CO 2003 NJ 1948 CT 1995 NM 1963 DC No policy NV 1999 DE 1953 NY 2002 FL 1997 OH 1997 GA 2017 OK 2000 HI 1993 OR 2015 IA 2002 PA No policy ID No policy RI 1995 IL 1965 SC 1985 IN No policy SD 1969 KS 2015 TN 1969 KY 2006 TX 1997 LA 1985 UT 1996 MA 2014 VA 2010 MD 1980 VT 2001 ME 2007 WA 1981 MI 1909 WI No policy MN 1985 WV 1969 MO 1967 WY 1977 MS 1942 Notes: This table shows the first year in which minors (15, 16, and 17 year olds) were allowed to have their own independent bank accounts through state chartered banks in the given state. 26

Table A.2: Predicting State Implementation of Laws Allowing Minor-owned Bank Accounts Coefficient (Standard Error) Governor is Democrat -0.0339 (0.032) Unemployment Rate 0.0038 (0.020) State-chartered credit union members per capita 0.5043 (0.687) Federally-chartered credit union members per capita 0.3776 (0.647) Number of banks per 100,000 people -0.0002 (0.000) Medicaid beneficiaries per 100,000 people -0.0119 (0.007) SSI beneficiaries per 100,000 people -0.225 (0.174) Gross state product (100 millions) 0.00011* (0.0005) Food Stamp beneficiaries per 100,000 people 0.0012 (0.011) Poverty Rate 0.00146 (0.006) Population (in millions) 0.0296 (0.067) Financial education high school requirement -0.0032 (0.059) N 1,030 Notes: Robust standard errors clustered at the state-level in parentheses. Coefficients are estimated from two-way fixed effects linear probability models that account for state and year. The dependent variable equals one if the state passed a MBL in the given year and state and zero otherwise. Data come from the National Credit Union Foundation, the FDIC, and the University of Kentucky’s Poverty Center. Financial education graduation requirements can be found out www.carlyurban.com/home/financial-education. All data span 1980-2015. 27

Table A.3. Effects of Minor Account Laws on Account use for 16 Year Olds, Omitting Always Treated States Has Over $100 Has Solo Has Joint Account Saved Account Account (1) (2) (3) (4) Minor Banking Law 0.110* 0.0725** 0.0529 0.0264 (0.0535) (0.0400) (0.0317) (0.0497) Mean DV 0.304 0.105 0.154 0.165 N 976 976 976 976 *Statistically significant at 10% level; ** at 5% level; *** at 1% level. Notes: Robust standard errors clustered at the state level in parentheses. Linear probability models estimated. All models estimate Equation (1), dropping states with policies before 2012. The data include 16-year-olds only. Data from the 2014 and 2018 waves of the SIPP. Table A.4. Effects of Minor Account Laws on Youth Employment for 16 Year Olds DV=1 if respondent worked at all (1) (2) (3) Minor Banking Law 0.0802 0.0796 0.0736 (0.0619) (0.0632) (0.0594) Mean DV 0.147 0.147 0.151 N 3,417 3,417 976 *Statistically significant at 10% level; ** at 5% level; *** at 1% level. Notes: Robust standard errors clustered at the state level in parentheses. Linear probability models estimated. All models estimate Equation (1), dropping states with policies before 2012. The data include 16-year-olds only. Data from the 2014 and 2018 waves of the SIPP. 28

Table A.5. Effects of Minor Account Laws on Account Ownership and AFS use, Omitting Always Treated States Has Account Use AFS (1) (2) (3) (4) (5) Minor Banking Law 0.109** 0.0524* -0.0759 -0.119 0.0458 (0.0319) (0.0295) (0.0267) (0.108) (0.0496) Mean DV 0.917 0.919 0.910 0.357 0.310 N 391 810 2,985 406 1,648 Ages 18-19 18-20 21-24 18-20 21-24 *Statistically significant at 10% level; ** at 5% level; *** at 1% level. Notes: Robust standard errors clustered at the state level in parentheses. Linear probability models estimated. All models estimate Equation (1), but they drop states where policies came into effect 2007 and earlier. Data from the FDIC Un(der) Banked Survey (2009, 2011, 2013, 2015, 2017). Has account equals one if the individual has a checking or savings account and zero otherwise. Use AFS equals one if the individual has used alternative financial services in the last 12 months and zero otherwise. Table A.6. Effects of Minor Account Laws on Account Ownership and AFS use, with Statespecific Linear Trends Has Account Use AFS (1) (2) (3) (4) (5) Minor Banking Law 0.110 0.0520 -0.0300 -0.0540 0.0961 (0.0875) (0.0355) (0.0271) (0.105) (0.0809) Mean DV 0.896 0.892 0.904 0.359 0.319 N 1,202 2,463 8,896 1,226 4,932 Ages 18-19 18-20 21-24 18-20 21-24 *Statistically significant at 10% level; ** at 5% level; *** at 1% level. Notes: Robust standard errors clustered at the state level in parentheses. Linear probability models estimated. All models estimate Equation (1), but they include state-specific linear trends. Data from the FDIC Un(der) Banked Survey (2009, 2011, 2013, 2015, 2017). Has account equals one if the individual has a checking or savings account and zero otherwise. Use AFS equals one if the individual has used alternative financial services in the last 12 months and zero otherwise. 29

Table A.7. Bacon Decompositions for Banked Status Outcomes Average Weight Estimate DD Comparison SIPP Earlier Treatment vs. Later Control 0.013 0.074 Later Treatment vs. Earlier Control 0.008 0.263 Treatment vs. Never Treated 0.107 0.023 Treatment vs. Already Treated 0.871 0.094 FDIC (18-20) Earlier Treatment vs. Later Control 0.099 -0.018 Later Treatment vs. Earlier Control 0.066 0.005 Treatment vs. Never Treated 0.208 0.001 Treatment vs. Already Treated 0.627 0.009 FDIC (21-24) Earlier Treatment vs. Later Control 0.034 0.005 Later Treatment vs. Earlier Control 0.044 -0.039 Treatment vs. Never Treated 0.320 -0.030 Treatment vs. Already Treated 0.603 -0.012 Notes: This table shows the decomposition (Goodman-Bacon 2018) for the models that correspond to the “has account” variable in the SIPP and FDIC. The models do not include controls. Average estimate is the average DD estimate for the DD comparisons labeled. 30

Appendix B: Supply Side Effects of MBLs While our focus is on how young people use bank accounts, we can also rule out that state minor account laws are associated with shifts in the structure of financial services markets. Since only statechartered firms can offer minor accounts, we use the same difference-in-difference model in an event study specification to determine if the passage of minor account laws change the presence of state-chartered banks and nationally-chartered banks per capita at the zip code level. There is no clear response by banks to the minor account laws. We use data from the FDIC on state and nationally-chartered bank locations by state from 1994-2018 to determine how minor account laws affect the presence of state and federal banks. We then use these data and an event-study style difference-in-difference specification to determine whether or not the policies affect entry or exit of state-chartered banks per capita in the market. Specifically, we estimate Equation (2), omitting the year just before the policy takes effect (t-1) and clustering standard errors at the state level. We include state- and year-level fixed effects. is the number of state (and alternatively federal) banks in a given zip code z in year y within state s. We continue to control for the presence of a state credit union minor account law. and Y {s c ,t o } ntain all preceding periods and periods post, respectively. For reference, the average state and federal banks per capita are 17.6 and 16.4 banks per 100,000 people. As in Equation (2)𝛼𝛼, − w 5 e omi𝛼𝛼t 5 st ates that were always treated the entire period (those with laws passing in 1994 or earlier). (2) 5 YT { h s, e t } tw=o𝛼𝛼 p 0 an+e ls ∑ o {𝑖𝑖 f = F − i 5 g ,𝑖𝑖 u ≠ r − e 1 B ,− . 2 1 } d𝛼𝛼i 𝑖𝑖 s𝑀𝑀pla𝑀𝑀y𝐿𝐿 t { h 𝑖𝑖 e ,𝑖𝑖 } e v+en𝛾𝛾t 1 s𝐶𝐶tu𝐶𝐶d𝐿𝐿y𝐶𝐶 f𝑤𝑤ig { u 𝑖𝑖 r ,𝑖𝑖 e } s+ fo r𝛽𝛽 s 𝑖𝑖 t+ate δa s nd+ fϵe { d s e ,t r } al banks per 100,000 people with 95% confidence intervals. There are no visible changes in state-chartered or nationallychartered banks in the long-run, there are no statistical differences at the 95% level pre- or postperiod estimates. There is no clear supply-side response by firms to changes in state laws. Our findings so far suggest that early access to independent accounts increases the likelihood of being fully banked through most of young adults' 20s but is not sustained in the long-run. In addition, there does not appear to be an increase in state chartered banks due to the policy. While we suspect that youth induced into owning bank accounts in their own name due to the policy change are unlikely to have substantial assets to save in their teenage years, it could still be in banks' best interests to offer minor accounts if the policies increase assets in state chartered banks in the longrun or if the policies have spillover effects for other individuals with more assets (e.g., siblings or parents). We test whether or not state and federal bank assets change due to the policy in the same event study style specification as in Equation (2). These asset data from the FDIC are at the state level from 1994-2018. Figure B.2 shows that there are no changes in state-chartered bank assets after the policy in the short- or long- run. For reference, the means are 0.248 and 4.32, interpreted as $2,480 and $43,200 per person in assets at state and federally-chartered banks, respectively. We find no evidence to suggest that the policy substantively changes the supply of state or federal banks in an area. Minor-owned bank accounts are a small portion of the overall market, with low account balances. States with minor bank account laws are not engaging in a broader pattern of promoting banking services that could be an alternative explanation for these findings. 31

Figure B.1. Event study of minor banking laws on the number of state and federal banks State-chartered banks Federally-chartered banks Notes: Each point represents an estimate and 95% confidence interval of from Equation (2), where i represents each period pre and post policy change. The excluded group is the year before passage i=-1. 5+ years prior includes years over five years before the passage of MBL, and 4+ years post includes ove𝛼𝛼r 𝑖𝑖4 years after the passage of MBL. The outcome variables are the number of state and federal banks per 100,000 residents, respectively. Data are from the FDIC from 1994-2018. Figure B.2. Event study of minor banking laws on the assets in state and federal banks State-chartered banks Federally-chartered banks Notes: Each point represents an estimate and 95% confidence interval of from Equation (2), where i represents each period pre and post policy change. The excluded group is the year before passage i=-1. 5+ years prior includes years over five years before the passage of MBL, and 4+ years post includes ove𝛼𝛼r 𝑖𝑖4 years after the passage of MBL. The outcome variables are the number of state and federal bank assets per capita in ten thousands, respectively. Data are from the FDIC from 1994-2018. 32