   |
| GET STARTED |
|
|||||
|
Copyright (c) 2004 Yale University Yale Journal on Regulation Winter, 2004 21 Yale J. on Reg. 121 LENGTH: 58175 words ARTICLE: Banking the Poor NAME: Michael S. Barr+ BIO: + Assistant Professor of Law, University of Michigan Law School. B.A. Yale College 1987, M. Phil., Magdalen College, Oxford 1989, J.D. Yale Law School 1992. I would like to thank Alan Berube, Stephanie Dorn, and Kevin Smith for their research assistance. This Article builds on work conducted while I served as deputy assistant secretary of the U.S. Treasury from 1997 to 2001. I am indebted to my colleagues at the Department for their work on the unbanked, including Alan Berube, Natasha Bilimoria, Don Graves, Clifton Kellogg, and Melissa Koide in the office of community development policy and Fiscal Assistant Secretary Donald Hammond, Roger Bezdeck, and the professional team at the Financial Management Service. I would also like to thank Michele Kahane and the Ford Foundation for generous support for the research for this Article, and Bruce Katz and the Brookings Institution's Center on Metropolitan and Urban Policy, where I am a nonresident Senior Fellow, for their ongoing assistance. Research for this Article was also supported by the Cook Endowment of the University of Michigan Law School. For comments on earlier drafts, I would like to thank Reuven Aviyonah, Greg Baer, Omri Ben-Shahar, Rebecca Blank, Mike Bylsma, John Caskey, Marsha Courchane, Steve Croley, Cathy Donchatz, Peter Edelman, Richard Friedman, Jonathan Gruber, Don Herzog, Janet Holtzblatt, Jill Horwitz, Elena Kagan, Andrea Kane, Robert Litan, Kyle Logue, Karl Scholz, Ellen Seidman, Dan Sokolov, J.J. White, and the participants in the Law & Economics Workshop at the University of Michigan Law School. SUMMARY: ... Most households can take access to a bank account for granted. ... The unbanked are also largely cut off from mainstream sources of credit, whether for short-term consumer borrowing or home ownership, because, without a bank account, it is more difficult and more costly to establish credit or qualify for a loan. ... To get a sense of the costs to a consumer who conducts most of his financial business through a check-cashing outlet, however, consider a customer that Dollar might describe as among its typical clients - a young, immigrant day laborer with intermittent income and a family at home in Mexico. ... Given the economics of ATM placement and operation, which require high volumes of transactions, a strategy for expanding access to banking for the poor using ATMs or POS, or with advanced functions such as bill payment, will likely require some governmental incentives to be viable in some low-income areas with low penetration of these technologies. ... More widespread use of direct deposit and electronic bill payment would not only lower payment systems costs overall, but also remove some key reasons why low-income people need to frequent high-cost alternative financial service providers. ... TEXT: [*123] Introduction Access to financial services is critical to success in the modern American economy. Most households can take access to a bank account for granted. Yet 22% of low-income families - over 8.4 million families earning under $ 25,000 per year - lack the most basic financial tool, a bank account. 1 These "unbanked" households and other "underbanked" low-and moderate-income individuals face high costs, relative to their income, for basic financial services. For example, a worker earning $ 12,000 a year would pay approximately $ 250 annually just to cash payroll checks at a check cashing outlet. 2 Low-income workers often turn to tax preparation services and costly refund loans to access their government tax refund check under the Earned Income Tax Credit (EITC). The costs of these basic financial transactions can undermine public initiatives that help families move from welfare to work, and can diminish the effectiveness of the EITC in lifting families out of poverty and encouraging workforce participation. Low-income families, particularly those without bank accounts, often lack any regular means to save. These families, often lacking alternative forms of financial resources, need to save, however, as a cushion against short-term crises, such as injury or job loss, as well as for longer-term goals, including buying a home, sending their children to college, or retirement. Of course, a central reason that low-income people find it [*124] difficult to save is that they have low incomes, but evidence suggests that low-income people can save if they have structured mechanisms to do so. 3 The unbanked are also largely cut off from mainstream sources of credit, whether for short-term consumer borrowing or home ownership, because, without a bank account, it is more difficult and more costly to establish credit or qualify for a loan. Even those low-income persons who have an account may, in effect, be "underbanked": They may rely on check cashers to cash their payroll checks; they may lack an institutionalized means to save, such as through payroll deduction plans; or they may not have, or may have tapped out, credit cards, and turn to relatively high cost forms of short-term credit, such as payday loans, to meet their liquidity needs. Alternative financial service (AFS) providers - including check cashers, money transmitters, payday lenders, title lenders, and tax preparation services that provide refund anticipation loans - are providing a wide range of financial services in low-income communities. 4 For example, check cashers provide a means for unbanked employees to convert their paychecks to cash. Payday lenders provide short-term credit to borrowers who cannot access credit cards or are already at credit limits. There are benefits to this market segmentation, and for low-income consumers, it is likely that without such services, they would be even less able to fulfill their financial services needs. Still, such services often come at a high cost to low-and moderate-income borrowers. Some portion of these high costs may be endemic to the nature of the transactions. These are paper-and labor-intensive transactions involving small dollar amounts, conducted on behalf of consumers with low wealth and often uncertain or poor credit history. These transactions are undertaken largely by financial service providers, which, unlike insured depositories, lack direct access to the payments system for check clearance. Moreover, the fixed costs of lending show up in higher prices for loans of short duration and small amounts. Yet some portion of the costs - and consumer problems - can be traced to the patchwork of state and federal law that governs these providers. While the mainstream financial system works extraordinarily well for most Americans, many low-and moderate-income individuals face a number of barriers to bank account ownership. First, regular checking accounts may not make economic sense for many lower-income families. For example, consumers who cannot meet account balance minimums pay high monthly fees, and most banks levy high charges for bounced checks, which families living paycheck to paycheck can ill afford. Second, many [*125] unbanked persons may not qualify for conventional bank accounts because of poor credit history or prior problems with managing a bank account. While some persons undoubtedly pose undue risk for account ownership, many could responsibly use bank accounts structured for their needs. In particular, accounts that do not permit overdrafts would reduce the risk associated with customers despite previous problems with the banking system and would diminish the need to sort out customers who had bounced checks in the past, except for those who were judged to have committed fraud. Third, while many low-income communities contain both banking institutions and alternative financial services providers, in some communities, banks, thrifts, and credit unions are not as readily accessible as in higher-income areas. Fourth, financial institutions may be reluctant, given low expected returns, to invest in research, product development, account administration, bank personnel training, marketing, and financial education necessary to expand financial services to lower-income clientele. That is, banking the poor is unlikely to be seen as sufficiently profitable for many banks to incur the up-front costs of entering this market, particularly because most banks are not institutionally organized to focus on this market segment. Fifth, the technologies that would make it less costly for low-income persons to use banking services are subject to positive network externalities. These externalities may slow the adoption of electronic forms of income receipt and payment. The legal literature on issues regarding access to financial services for the poor focuses largely on debates over usury laws and consumer protections in the alternative financial services sector. 5 In this Article, I hope to shift the debate toward ways in which governmental incentives can harness market and technological forces to expand access to financial services for the poor. The Article begins by systematically exploring the demand for financial services, building on John Caskey's foundational study of the alternative financial services industry 6 and the work of a few other sociologists who in recent years have begun to explore the financial services usage of the unbanked population. 7 I examine the supply of financial services by the alternative financial services and banking sectors. [*126] I next analyze changes in payments and distribution systems that influence the provision of financial services for the poor. On this empirical foundation, I argue that the federal government should act as a catalyst in encouraging the private sector to transform financial services for low-income persons. The case for governmental intervention rests primarily on four arguments. First, bank account ownership contributes to optimal income redistribution policies. Bank account ownership (or a similar means of receiving income) can be thought of as logically prior to the receipt of a governmental transfer of income. The Article takes as a given that our society has a goal, as evidenced by such policies as the 1996 Welfare Reform law 8 and the EITC, to increase workforce participation and reduce poverty among the working poor. 9 Generally, providing government assistance in the form of income enhances social welfare more than providing assistance in kind, both because administration is likely to be less costly and because income assistance provides the recipient with the freedom to spend the income however she desires. 10 Given the high cost of converting income into liquid form, however, promoting bank account ownership for the poor is probably more efficient than simply transferring income. The form in which "income" is transferred changes its value. A governmental check is worth its face value less the cost of converting it to cash. In addition, one unit of in-kind assistance, in the form of sufficient governmental incentives to induce a bank to offer a bank account to a low-income person, would provide the benefit of liquidity to all subsequent income transfers whether from government programs, wages, or other sources. Thus, the concept of income transfers being more efficient than in-kind assistance breaks down when one needs to deliver that income to people in the real world. Moreover, the government saves money by transferring funds electronically, rather than by paper check. Although paternalism forms the basis of much savings policy, and may be justified under some circumstances, paternalism is not the impetus for favoring subsidizing account ownership. The thrust of the argument in [*127] favor of subsidies for bank accounts is to increase the supply of bank accounts tailored to the needs of the poor, not primarily to change demand of the poor for existing types of traditional accounts that many of them currently do not want or for which they may not qualify. 11 Second, high-cost financial services reduce effective take-home pay and thus may undermine employment incentives contained in such measures as the EITC and the Welfare Reform law, although further empirical research is needed to confirm this hypothesis. The costs of changing the delivery system for financial services for the poor are small relative to the likely gains to be had for these programs. More broadly, improved access to bank accounts can reduce the costs of financial services for the poor, expand access to lower-cost forms of credit and increase opportunities for saving - all key to reducing poverty and expanding social mobility. 12 Evidence to date suggests that low-income people can save, but lack the institutional mechanisms available to middle-and upper-income Americans to do so. Providing a better opportunity for the poor to save is likely both to reduce their short-term exposure to liquidity shocks, and to increase their long-term prospects for building their human capital and saving for homeownership or other assets that can help them get out of poverty. 13 The positive network externalities in payments systems that the Article identifies as promising for expanding access to the banking system for the poor - online debit at automatic teller machines (ATMs) and merchant point of sale (POS) terminals, and the automated clearinghouse (ACH) system for direct deposit and bill payment - provide a third justification for governmental intervention. 14 The social benefits of wide adoption of these systems exceed the private benefits that can be captured by their owners. While many of the network externalities inherent in these systems have already been internalized, further government policies to [*128] reduce these externalities would have wide social benefits and should be adopted. With distributive goals for the poor in mind, I also suggest additional steps. These electronic payment and delivery systems matter in expanding financial services for the poor because they are lower risk and lower cost to consumers, employers, and financial institutions than checks or off-line debit forms of payment systems. Fourth, the patchwork of state laws governing the alternative financial services sector, gaps in federal law, and complicated regulations governing federal programs may also increase the costs and problems associated with providing financial services for the poor. Regardless of one's views about the merits of income redistribution, one ought to favor certain legal reforms. For example, state geographic restrictions on locations of check cashers needlessly inhibit competition among check cashers in low-income communities. In another vein, complicated federal rules governing the EITC and delays in tax refund processing may drive low-income taxpayers to take out expensive refund anticipation loans offered by paid preparers. Thus, the Article recommends changes in several regulatory areas. The Article rejects two common approaches to thinking about financial services for the poor. First, the Article disputes the proposition that the financial services currently provided in low-income communities are necessarily efficient or desirable. That is, while there are benefits to market segmentation, there are also social costs to the current system. Second, the Article disputes the proposition that the remedy for perceived problems in the alternative financial services sector is a return to usury laws. Rather, the Article contends that the financial service system could serve at least some segments of low-and moderate-income households better with modest governmental incentives to the banking sector to spur innovation in serving the poor. The Article makes several key policy recommendations to help transform financial services for the poor. First, governmental incentives should be offered to encourage financial institutions to offer electronically based transaction accounts designed for low-and moderate-income persons. For example, debit-card-based accounts accessed at ATMs and at merchant POS terminals can be offered at much lower cost, and with lower risk to banks and consumers, than checking accounts. Such electronically based accounts may be particularly suited to low-income unbanked persons. Yet the fixed costs of offering these accounts may currently be too high to be borne by banks and their low-income customers, and banks have been largely unwilling to take on the opportunity costs of shifting bank resources toward serving low-and moderate-income persons. Further technological and financial innovation spurred by these incentives can help to drive down costs, reduce risks, and increase competition. [*129] My view, which I defend in Part V, is that tax credits to financial institutions would be the preferable means for delivering the subsidy. Although tax credits could most readily be employed to create a larger scale initiative, other helpful steps short of a new tax credit could be taken. For example, the Treasury Department's "First Accounts" pilot could spur experimentation, but needs to be funded on a multi-year basis. The incentives that I propose could also build on revamped governmental programs designed to move federal and state-run government benefit recipients from receiving certain government income transfers by paper check to electronic benefit transfer (EBT). State EBT programs - focused on welfare recipients and other beneficiaries of state-run benefit programs - should shift away from electronic benefit cards towards provision of debit-card based bank accounts that could be used to receive all forms of income transfer as well as privately earned wages. Treasury should enhance the incentives in the federal Electronic Funds Transfer (EFT) initiative - currently focused on moving Social Security recipients, federal retirees, and other beneficiaries of certain federally run programs to direct deposit - for banks to offer low-cost, electronically based bank accounts to a wider range of low-income households, including those who receive the EITC. Second, taking account of the positive network externalities associated with online debit cards, direct deposit, and bill payment, while recognizing the risk that government intervention in these networks may miss the mark, adjustments may be needed in Federal Reserve Board pricing of both check and ACH services. ACH services are likely still priced too high in relation to checks. In addition, now that VISA and MasterCard have settled the suit brought by Wal-Mart and other merchants alleging antitrust violations arising from the credit card companies' "honor all cards" policies, federal antitrust officials and the courts should pay particular attention to ensuring that the terms of the settlement foster competition among different electronic payment methods. Depending on how the market evolves in the wake of the settlement, there may also be a case for subsidizing the further expansion of online debit infrastructure. Third, regulatory reform could play a secondary but useful role in altering the provision of financial services to the poor. For example, both the elimination of anti-competitive restrictions and the more consistent regulation and enforcement of disclosure could reduce some of the costs and problems associated with alternative financial service providers. Moreover, simplification of state welfare-to-work programs and the federal EITC, and IRS steps to facilitate the timely direct deposit of tax refunds, could help to reduce transaction costs, lower barriers to account ownership, and reduce demand for costly tax preparation and refund anticipation loan services. Furthermore, federal banking regulators should [*130] increase attention on how the Community Reinvestment Act (CRA) could encourage banks and thrifts to provide innovative financial services to the unbanked under the existing regulatory framework. Lastly, studies show that financial education, if coupled with structured opportunities to save, can increase participation in saving plans and increase the level of saving, particularly for low-and moderate-income persons. 15 Financial education is costly, and the benefits of a financially-educated consumer cannot be captured by a financial institution offering education because the consumer may shop for financial services at other institutions. Further, the positive externalities created by the provision of financial education make it unlikely that such education will be offered in a socially optimal quantity by private parties. Although financial education is unlikely to be successful against the backdrop of existing high-cost alternatives, such education focused on new banking products designed to meet the financial needs of the poor may be helpful.This Article proceeds as follows: Part I describes the low-and moderate-income population who use alternative financial service providers as a substitute for, or in addition to, banks, and explores the costs and benefits to consumers of using this sector. Part II analyzes the alternative financial services industry in depth and proposes changes to regulation of that sector. Part III explores barriers to low-and moderate-income persons using financial services at insured depository institutions and proposes strategies to lower these barriers. Part IV analyzes changes in the payments system and electronic financial services networks that could enhance financial services for the poor. Part V focuses on the Article's key policy recommendations for financial services, savings, and financial education. The Article then concludes. I. The Unbanked 16 A. Patterns of Account Ownership Twenty-two percent of low-income families - over 8.4 million families earning under $ 25,000 per year - do not have either a checking or [*131] savings account. 17 Most of the unbanked are low-income: 83% of the unbanked earn under $ 25,000 per year. 18 The unbanked may be especially concentrated in low-income neighborhoods; in low-income areas of Los Angeles the unbanked represent nearly a third of the population, and the unbanked represent over 40% of the population in low-income neighborhoods in New York. 19 Among low-to moderate-income families, households are more likely to be unbanked when they have lower incomes, less wealth, less education, are not working, are younger, have more children, rent their home, and are a racial or ethnic minority. 20 Broadly speaking, the most common reason persons cite for lacking a checking account is not having enough money to be able to afford the costs of account ownership. 21 Other factors cited by [*132] respondents include distrust of banks, not wanting to deal with banks or privacy concerns. 22 Efforts to reach the unbanked also need to pay attention to the racial and ethnic composition, and immigration status, of segments of the unbanked. Recent evidence suggests that, irrespective of the race of the individual, families living in neighborhoods with higher concentrations of Blacks or Hispanics are less likely to own a checking account. 23 Among the banked, low-income minorities are less likely than whites to have a checking account, but more likely than whites (all else being equal) to have savings accounts. 24 By contrast, proximity to a bank branch seems not especially predictive of being banked. 25 Low-income immigrants are more likely than other low-income persons to be unbanked. Moreover, significant immigration in the late 1990s may have contributed to the persistence in the size of the unbanked population. 26 [*133] While important challenges remain, some progress has been made in recent years in expanding access to financial services. The period 1995 to 1998 marked a decline in the percentage of low-income families who are unbanked from 25% to 22%. 27 This decline in the percentage of unbanked may reflect, in part, strong economic growth during the late 1990s that improved the incomes of households at the bottom of the income distribution for the first time in decades (although these gains eroded in the last two years). 28 Increases in the EITC increased the take-home pay of low-income workers and helped to increase labor force participation. Welfare reform, beginning with waivers for states to use welfare-to-work strategies and culminating with the 1996 Welfare Reform law, increased the percentage of welfare recipients entering the workforce. Greater workforce attachment and higher incomes may have increased the benefits of bank account ownership and also may have provided more low-income persons with the wherewithal to meet bank minimums or afford bank fees. Account ownership grew most quickly among groups at or below the poverty threshold and the next largest gains came from those just over the poverty line. 29 In addition to these economic gains, advances in technology, the spread of ATMs and POS terminals, and improvements in payments system efficiencies have lowered the cost and improved the distribution of payments systems that could benefit the poor. Some small-scale governmental and private initiatives may have contributed to this trend. The U.S. Department of the Treasury's efforts to increase electronic payment of federal benefits and a Treasury "First Accounts" pilot project to reach the unbanked have also helped to spur innovation. 30 Recently, partly in response to these government initiatives, some banks, thrifts, and credit unions, as well as community-based organizations, have begun to experiment with products designed to serve the needs of low-income individuals, and to serve the growing Hispanic market in ways that may benefit low-income persons generally. These nascent efforts provide the [*134] context for understanding the challenges that lie ahead. B. The Costs of Being Unbanked The consequences of not having access to mainstream financial services can be severe. High-cost financial services reduce disposable income for those least able to afford it. Such services reduce the value of government transfer programs, including the EITC, and may undermine federal and state initiatives to improve workforce participation and reward work. Lack of access to mainstream financial services also undermines the ability of the poor to save and to access credit, reducing their long-term wealth. Low-income people using check cashers may be more susceptible to robbery because they tend to cash their entire paycheck at regular time periods. Additionally, reducing inefficiencies in the payments system for the poor may have modest positive effects on the economy. First, the "unbanked" face high costs for basic financial services. 31 For example, a 2000 Treasury study found that a worker earning $ 12,000 a year would pay approximately $ 250 annually just to cash payroll checks at a check-cashing outlet, 32 in addition to fees for money orders, wire transfers, bill payments, and other common transactions. 33 Almost all of the checks cashed at check cashers pose relatively low risk: Payroll payments - with low credit risk that could be directly deposited by electronic means, instead of by check, into bank accounts, at significantly lower costs to the payment system - constitute 80% of checks cashed at these check cashing outlets. Another 16% are government benefit checks, which again pose low risk. 34 A large portion of these checks could [*135] presumably be directly deposited into bank accounts at relatively low cost - if low-income people had bank accounts. 35 The costs of these basic financial transactions reduce the effectiveness of federal income transfer programs such as the EITC and may undermine public initiatives to move families from welfare to work. High cost financial services reduce effective take-home pay. Moreover, studies of the EITC suggest that higher take-home pay from the EITC helps to induce labor force participation. 36 One survey found that forty-four percent of a sample of EITC recipients in inner city Chicago used a check cashing service to cash their government refund check. 37 Nationwide, in 1999, nearly half of the $ 32 billion in EITC refunds provided to over 18 million low-income families were distributed through refund anticipation loans, costing EITC recipients $ 1.75 billion for tax preparation services, electronic filing, and loan fees. 38 The high price of converting income checks into liquid form (e.g., cash) may reduce the efficacy of the EITC in encouraging workforce participation because it reduces take-home pay (and reduces it more, the more the person earns), or at the very least these transaction costs significantly increase the taxpayer's costs (the "compliance" costs) of the program. Similarly, although the jury is still out on the long term effects of the 1996 Welfare Reform law, some studies suggest that welfare programs that encourage work, coupled with policies that let families keep more of their earnings before benefits are reduced or eliminated, have helped to increase workforce participation and job retention. 39 The positive effects of welfare reform on workforce participation and income generation, however, may be undermined by high-cost check-cashing services that reduce the effective income of those who are beginning to earn wages. 40 [*136] Even in the bulk of states that have moved to EBT for welfare payments, welfare recipients may still face high costs for financial services: First, administrative problems in some state programs make it hard to withdraw sufficient funds for bill payment (e.g., monthly rent). Second, most EBT programs do not link recipients to bank accounts, which means that these recipients need to find other means to convert their work income to cash, to pay bills, to save funds, and to access credit. Third, once welfare recipients are working, payroll checks and EITC refunds to these individuals would push them towards high cost transaction services. In turn, welfare recipients may be dissuaded from opening a bank account because they believe, sometimes mistakenly and sometimes correctly, that their bank account balances will cause them to exceed state welfare program asset limits. 41 Second, low-income families need to save to cushion themselves against personal economic crises, such as injury or loss of a job, and for key life events, such as buying a home, sending their children to college, or entering old age. 42 Low-income households face key barriers to increased saving, 43 and their low income leaves them little opportunity to save. Because they are poor, they face higher opportunity costs for putting their funds toward savings rather than current consumption. In turn, because the poor accumulate little, financial institutions face high costs in collecting their savings relative to the amounts saved, and will thus be reluctant to expend the resources to open accounts for them or will offer them low returns on their savings, further reducing any incentives the poor have to save. Low-income families, particularly those without bank accounts, often lack any regular mechanism to save, such as payroll deduction plans, further reducing the likelihood that they will do so. 44 [*137] In a survey of New York and Los Angeles low-income neighborhoods, 78% of the banked held some form of savings, broadly defined, while only 30% of the unbanked had savings. 45 Obviously, the ability to save is a function of income. But differences remain even after controlling for income. Across income ranges, being banked is highly correlated with saving. 46 Of course, bank account ownership may well be correlated with willingness and ability to save. Thus, one would need to measure differences in propensity to save in order to determine whether and to what extent account ownership itself is a strong factor in increasing savings. 47 Bank accounts can be important entry points for the provision of regular savings plans for low-income workers through payroll deduction. Still, most low-income workers work for firms without savings plans or are themselves not covered by such plans even when their employers have savings plans. Even the tax system, through which the bulk of government savings benefits are provided, disproportionately subsidizes savings for higher-rather than low-income households. The Treasury Department estimates that more than two-thirds of tax expenditures for pensions go to households in the top 20% of the income distribution, while the bottom 40% get only 2% of the tax benefit. 48 Promoting low-income household savings is critical to lowering reliance on high-cost, short-term credit, lowering risk of financial dislocation resulting from job loss or injury, and improving prospects for longer-term asset building through homeownership, skills development, and education. Evidence to date suggests that low-income individuals can save if given the opportunity to do so, at least if offered a significant matching contribution. Some 73% of federal employees earning $ 10,000 to $ 20,000 annually participated in the federal government's Thrift Saving Plan, which provides a government matching contribution, and over half of those earning under $ 10,000 also participated. 49 The 2001 Survey of Consumer Finances found that 30% of families in the bottom income quintile saved in the prior year, and 53.4% of those in the next quintile [*138] saved. 50 Savings account features have an appeal for the unbanked. In a Treasury survey of unbanked federal check recipients, respondents were aware that an ETA savings feature would only pay a nominal rate of interest (explicitly posed in the survey as "$ 2 annually on a $ 100 deposit"), but this feature would account for approximately 25% of the typical respondent's decision on whether to enroll in the ETA. 51 Some lower-income individuals use alternatives to bank accounts to facilitate savings. Anecdotal evidence exists that low-income people purchase money orders with their paychecks at the beginning of the month and hold them for later use. In so doing, they convert their income and benefits into a more illiquid and protected form, either for bill payment later in the month or as "savings" for planned and unplanned expenditures in the future. 52 Researchers have also found that low-income taxpayers over-withhold their income taxes more frequently than higher income taxpayers; some economists suggest that these taxpayers use withholding as an automatic savings mechanism. This may suggest that demand for savings products among the poor is high enough that some will accept a zero or negative interest rate. 53 Third, without a bank account, it is more difficult and more costly to establish credit or qualify for a loan. A bank account is a significant factor - more so, in fact, than household net worth, income, or education level - in predicting whether an individual also holds mortgage loans, automobile loans, and certificates of deposit. 54 After controlling for key factors, one study determined that low-income households with bank accounts were 43% more likely to have other financial assets than households without bank accounts. 55 Low-income persons without bank accounts face higher costs of credit than low-income persons with accounts, 56 and in any event, low-income households generally face higher [*139] costs of credit than households with higher incomes. 57 In effect, low-income individuals must pay more to transform their labor into productive capital and are thus "under-rewarded for their talent." 58 Moreover, as noted above, the savings products, including bank accounts, that low-income persons can access generally provide low levels of return, which reduces their income growth and may lower incentives to save. This lower saving rate is a problem itself and further increases the cost and reduces the availability of credit to these households, which is at least in part a function of their savings. Low-income families find it difficult simply to make ends meet each month and lack access to short-term credit at a reasonable cost to smooth out earnings. The main complaint of low-income families, for example, in Caskey's study, was the "insecurity and stress associated with living from paycheck to paycheck." 59 Most low-and moderate-income households manage to spend all their income each month. 60 Bank account ownership will not suddenly change that, but account ownership may make it easier for low-income households to manage their finances, save even if in modest amounts, and access lower-cost forms of credit. It is difficult to determine causation, but a lack of account ownership is correlated with credit problems. Either unbanked low-income persons have lower propensities to plan financially than other low-income households, or lack of a bank account makes it harder to plan and save. In turn, once credit problems emerge, credit-impaired individuals have a harder time getting access to bank accounts. 61 In Caskey's survey of low-income households, 42% of unbanked households were two months late on bills in the last year, compared with 28% for banked households; 41% of unbanked households were contacted by a debt collection agency in the past year, compared with 25% for banked families. 62 When low-income unbanked families need to borrow, they must turn to expensive forms of credit. Only 14% of unbanked poor families carry credit cards that might help them smooth out payment for short-term increases in consumption or to weather occasional dips in income, while 59% of low-income banked [*140] households carry credit cards. 63 Again, we do not know whether unbanked households without credit cards would be able to handle such cards if they became banked. Account ownership in and of itself is no panacea, however; even low-and moderate-income individuals with bank accounts often lack savings and turn repeatedly during the year to payday lenders (who charge on average 474% APR), and to other forms of high-cost credit. 64 Low-income families often lack health insurance, and those without savings or access to informal networks of family and friends often use payday loans when faced with expenses related to birth or illness. 65 Individuals may also borrow, of course, for less basic reasons, such as entertainment, splurges, gambling, and the like. Policy cannot, even if it were desirable, easily distinguish between the two kinds of borrowing. In my view, it is appropriate to increase opportunities for low-income persons to borrow at lower cost, even if some portion of the borrowing is taken out for reasons some may disfavor. Thus, I argue in Part V that strategies to bring low-income persons into the financial services mainstream need to include initiatives designed to increase savings for short-term financial stability and to improve access to less expensive forms of credit where appropriate - for example, with overdraft protection, account-secured loans, credit cards or loans with automatic withdrawals from pay directly deposited into accounts, but with significantly longer terms than payday loans. Fourth, low-income families who cash their paycheck may face high risk of robbery or theft. 66 By transitioning into bank accounts where they can store a portion of their earnings, withdraw funds in smaller amounts, [*141] pay for goods or services directly using debit, and withdraw funds outside of the concentrated time periods during which benefit checks and paychecks are commonly cashed, these families can decrease their exposure to risk of crime. Fifth, inefficiencies in the payments system impose costs on the national economy. Increasing the efficiency in the payments system for the poor could have modest positive effects on the economy as a whole. A Federal Reserve Board study suggests that the U.S. economy would save over $ 1 for each check that is converted to an electronic payment. 67 The study estimates savings of $ 30 billion per year if one-half of current check volume is converted to electronic payment. 68 Check processing costs between 0.25% and 1% of U.S. GDP, in addition to losses from fraud. 69 While low-income check volume is only a small fraction of the total, electronic payments for the poor could help, and, because of positive network externalities, funds spent converting the poor to electronic payment might speed conversion to electronic payments more generally. Helping low-income households to leapfrog over checking to electronic payments, just as some poor countries have been able to leapfrog over conventional telephone lines to cell or satellite phones, may thus have broader societal benefits. These effects, though positive, are of course likely to be quite small in relation to the overall economy. II. The Alternative Financial Sector The previous Section summarized the costs of being unbanked. This Part explores in depth the growing number of AFS providers, offering a wide range of services, including short-term loans, check cashing, bill payment, tax preparation and rent-to-own consumer goods, in low-income urban neighborhoods. 70 These AFS providers, and others like them, are currently the only means available for many low-income persons to access [*142] basic financial services. Understanding the costs and benefits of such services is critical to assessing the need, if any, for alternatives. A. Check Cashers 1. Industry Overview For many years, check cashers have been used by low-income individuals who seek to conduct basic financial transactions such as cashing checks, paying bills and wiring funds. John Caskey referred to these customers as employing the "cash and carry" method of financial management. Upon receiving a paycheck, they cash the check and pay their bills immediately. While check cashers offer essential services, the fees involved in converting paper checks into cash are high, relative to an alternative world in which low-income households would be able to rely more on direct deposit into bank accounts. The check-cashing industry grew dramatically during the 1980s and 1990s. Today, there are almost 10,000 stores in the U.S. that classify their primary line of business as check cashing, about double the number there were six years ago, and almost five times the number there were fifteen years ago. 71 The industry reports that it processes 180 million checks totaling $ 55 billion annually, generating $ 1.5 billion in fees. 72 Most of these checks are low-risk payroll or government benefit checks: 80% of checks cashed at surveyed check cashers in the 2000 Treasury study were payroll checks, while 16% were government benefit checks. 73 While even payroll checks are not without some credit and fraud risk, average losses from "bad" checks at check cashing firms are low. For example, Ace Cash Express (ACE) reports that 0.5% of the face value of checks bounce, but net losses after collection are 0.2%. 74 By comparison, 0.64% of the face value of interbank checks were returned in 2000. 75 Like the banking industry, the check cashing industry has undergone consolidation in recent years. Larger players are benefiting from greater [*143] economies of scale, shared technology platforms, and their ability to negotiate alliances with service providers (lenders, money transfer services, and ATM networks, among others). 76 ACE, the largest check cashing chain in the country, grew from 452 outlets in 1995 to 1,163 outlets in 2001, largely through acquiring independent stores and smaller chains. While check cashing remains a fragmented market overall, with 80% of the market divided among independent operators and small regional chains, national chains such as ACE capture a disproportionate share of check cashing revenues with their strategic and numerous locations, broader product lines, higher volumes, and generally higher prices for check cashing. 77 The industry's growth has been accompanied by growing diversification of the products and services offered. Nearly all respondents to the 2000 survey provided a core of services: check cashing, money orders, and wire transfer. 78 Many also provided an array of other products including lottery tickets, postage stamps, prepaid telephone cards, payday loans, bill payment, municipal services (such as the paying of parking tickets), and distribution of state benefits. These ancillary services increase revenue per customer while also broadening the industry's customer base beyond the unbanked population. For instance, Dollar Financial claims that 50% of its customers have bank accounts. 79 The most notable development in recent years has been the rapid growth in check cashers offering payday loan products. 80 For ACE and Dollar Financial, revenues derived from loan products increased from an average of 4% of total revenue in fiscal year 1997 to an average of 29% of total revenue in 2001. Nearly 45 to 50% of revenue growth at these two firms over that period was attributable to the expansion in payday loan originations. 81 Check cashers typically have high transaction volumes and high profit margins, but often on a relatively small revenue base. The average chain outlet in 2000 processed over 8,000 transactions per month, for estimated [*144] annual revenue of $ 242,500. The estimated pre-tax return on sales for these chain outlets was 28.5%. On average, revenues and returns for independent check cashers were far lower. 82 2. Customers Check-cashing industry clientele are drawn, for the most part, from lower-income urban households. ACE describes its core customer group as having annual family incomes of approximately $ 30,000, and Dollar indicates that its check-cashing customers' median household income is $ 22,500. In most cities, check cashers locate in neighborhoods with below-median incomes and above-average minority population shares. 83 Dollar describes its store base as a "mix of urban sites, which are located in high-traffic shopping areas, and suburban sites, which are located in strip malls near multi-family housing complexes." 84 Check cashers do not provide financial services to only unbanked consumers, nor do all unbanked consumers obtain their financial services primarily through check cashers. The story is much more complex, and local context seems to matter greatly. In lower-income neighborhoods of New York and Los Angeles, for instance, 71% of unbanked individuals who cashed checks primarily used check-cashing outlets, as did 28% of banked individuals. 85 In Atlanta, Oklahoma City, and eastern Pennsylvania, only 17% of unbanked individuals cashed checks at a check casher; almost half used a bank. 86 In Chicago, about two-thirds of all households who accessed some form of financial services at check cashers were found to have bank accounts, but, as one would expect, unbanked households are much more likely to use check cashers than are banked households. 87 Banco Popular reports that 38% of the customers at Popular Cash Express own a checking account, and 33% own a savings account. 88 [*145] Not all of the unbanked use or pay for these services or pay high fees. While there is great regional variation and much further study is needed, a large portion of the unbanked manage to avoid paying high costs for at least some of their financial services. For example, in one study, Caskey found that only 20%-40% of unbanked survey respondents paid fees to cash checks. 89 In another study, in New York and Los Angeles, more than half of the unbanked reported incurring no cost to cash checks, either because they receive their income in cash or have no income (33%) or cash their checks for free at banks and grocery stores (16%). 90 The unbanked mostly use check cashers to cash checks. When low-income households use banks to cash checks, they usually use the bank of issue, rather than their own or some other bank. 91 Most likely, these represent efforts by workers to cash payroll or personal checks "on-us" at the employer's bank. For the portion of such checks that are payroll checks, rather than personal checks, direct deposit could become an alternative to check-cashing, or to the costly use of bank teller and processing time (for which some banks are beginning to charge non-customers). Check cashers appear to capture both some portion of the unbanked population and some portion of the banked population as customers. Why might individuals with bank accounts be drawn to check cashing outlets? Some may be living from paycheck to paycheck, do not have direct deposit offered by their employer(s), and find that they lack sufficient liquidity to wait the two-to-three days for their bank to clear access to funds from a deposited check. Others might regularly wire money to relatives abroad, which, until recent adoption of dual ATM-technology at some banks, was generally less expensive at check cashers than at banks. 92 Still others may not be able to afford the fees or minimum balances associated with checking accounts and thus might own only a savings account without any capacity for transactional services. Some may need to make bill payments by purchasing money orders at a check casher, either because they live in a neighborhood where personal checks are generally not accepted, or because they do not have a checking account. A portion may favor the more convenient hours, locations, culture, or languages spoken at check cashers, and not be willing to use ATMs. One attribute that may distinguish some users of check cashers, both banked and unbanked, from other low-income workers who use banks, may be the nature of their work arrangements. Dollar Financial states that "many of its customers are workers or independent contractors who receive payment on an irregular basis and generally in the form of a [*146] check." 93 Although little independent analysis exists to corroborate Dollar's assertion, it makes sense on its face. 94 These workers do not have the benefit of a steady relationship with an employer that would tend to support a direct-deposit relationship with a bank, nor do they generally have the employment references that are often required to open a bank account. In addition, many undocumented immigrants work within this informal economy and, in many cases, their undocumented status makes it less likely that banks will open bank accounts for them. 95 A Treasury study found a significant correlation between outlet location and the percentage of working adults in the neighborhood who annually worked less than fifty weeks total, although the report also found such a connection between outlet locations and low neighborhood income; the two variables are likely correlated rather than independent. 96 Given that 80% of checks cashed at check cashers are payroll checks, and another sixteen percentage points are government checks, it should be possible to convert a large portion of these checks to direct deposit if low-income workers had access to bank accounts. Even for part-time workers, large employers or temporary employment firms can convert income checks to direct deposits. 3. Costs As with most businesses in the retail sector, regional variations abound in the prices that consumers pay for services offered through check cashers. 97 The Dove survey revealed that, across the four markets studied (Atlanta, Boston, San Antonio, and San Diego), check cashing prices were sensitive to differences in cost-of-living and level of competition. In San Antonio, where wages were lower and AFS competition greater, fees to cash payroll and government checks averaged about 1.5% of the check face value. In Boston, where higher costs of living prevailed and one chain dominated the AFS market, average fees were 2.5% of the check value. Atlanta and San Diego fell somewhere in between these figures. Fees to cash personal checks were much higher, but most outlets refuse to cash such checks given their greater risk. Across the four markets, chain outlets charged a higher percentage fee on average than did independent [*147] operators. 98 ACE's fees for cashing payroll checks average 2.2% of the face amount of the check, and Dollar's fees average 3.3%. 99 Chain check cashers processed an average of 3,400 money orders per month. Fees for money orders were only fifty to sixty cents, and were lower at chains due primarily to the favorable contracts they are able to negotiate with money order suppliers. 100 Money order fees at check cashers are lower than those charged by most banks, whose customers mostly do not use them. In addition, the U.S. Postal Service, for purposes of comparison, charges ninety cents for a domestic money order under $ 500, and $ 1.25 for money orders of $ 500 to $ 1,000. 101 Overall costs for using check cashers vary dramatically by patterns of usage. With respect to bill payment services, only 36% of the unbanked surveyed in New York and Los Angeles overall incur money order or bill payment fees from check cashers. 102 Many of the unbanked in that survey received cash income, had no income, or were able to cash their income checks at banks or stores, often at little or no fee. 103 Of those paying check cashing fees at check cashers, two-thirds incurred costs of less than $ 100 annually. The one-third of those surveyed who incurred more than $ 100 annually comprise only 11% of the total unbanked participants in the survey. Thus, the study suggests that, at least in New York and Los Angeles, the highest cost of using alternative financial service providers may be relatively concentrated among a portion of the unbanked. 104 To get a sense of the costs to a consumer who conducts most of his financial business through a check-cashing outlet, however, consider a customer that Dollar might describe as among its typical clients - a young, immigrant day laborer with intermittent income and a family at home in Mexico. He perhaps earned about $ 18,000 last year, but some of it was in cash, and some of it was paid in checks that he could cash for no fee at a local issuing bank. Altogether, he used a check casher to cash $ 12,000 in checks, at an average fee of 2% of the check face value - $ 240 total. Once a month, he wires $ 500 home to his family, at an average fee of $ 20, or [*148] $ 240 annually. His share of the rent is paid in cash, but he purchases three money orders a month to pay the apartment's electric bill, his cellular phone bill, and his car insurance. At fifty-five cents each, he pays $ 20 for money orders annually. All together, these fees would add up to $ 500 annually for this low-income consumer, nearly 3% of his annual income. 4. Regulation The regulatory structure governing check cashing may have some influence on fees. A number of states cap fees that may be charged by check cashers. However, check cashers also partner with national banks, which are permitted to set non-interest charges according to "sound banking judgment"; the Office of the Comptroller of the Currency (OCC) takes the position that state laws limiting or prohibiting such charges are pre-empted. 105 Other state laws governing check cashers may have an effect on fees. For example, to the extent that they are enforced, state rules limiting the number of check cashers that can operate in a given area may decrease AFS competition and increase check cashing fees. 106 5. Reforms Given the high cost structure of a paper-and labor-intensive industry, it is doubtful that costs of check cashing can be brought down significantly with existing technology. 107 Reduced state regulatory barriers to entry may help enhance competition if they are accompanied by consistent disclosure requirements and enforcement that would make it easier for consumers to shop for financial services. Some have suggested that banks themselves, with cheaper (direct) access to the payments system, might effectively compete for check-cashing services. 108 This Article argues that it would be cheaper, and the services provided more useful, if banks were to compete [*149] with check cashers by offering electronically based banking services, instead of competing with them as check cashers. Advances in direct deposit, debit card infrastructure, and electronic bill payment will also be required to bring down the costs of income conversion. B. Payday Lending Payday lenders provide short-term consumer loans to low-and moderate-income working people who have bank accounts. In the traditional "payday loan" transaction, the borrower writes a postdated (or undated) personal check to a lender. In return, the lender advances the borrower a cash amount equivalent to the face value of the check minus a finance charge. The lender holds the check before either depositing it or, more commonly, receiving cash repayment directly from the borrower, usually on the borrower's payday. In an updated form of the traditional transaction, no check is written; instead, the borrower signs an authorization that permits the lender to debit his bank account on a future date for the amount of the loan plus the finance charge. Loan terms are typically two weeks. Payday lending has become controversial because of concerns that the loans are expensive; that the structure of the product - a short term loan with a balloon payment and high fees - leads to defaults or borrowers falling into a "debt trap" as they repeatedly "roll over" the loan; and that payday lenders use misleading disclosures and aggressive collection. 1. Industry Overview Commercial check-cashing outlets have been in the United States since the 1930s. Payday lenders, on the other hand, did not operate as a formal industry until the early 1990s, although the short-term lending function has long been filled by pawnshops, auto title lenders, retail installment credit, and loan sharks, to name a few. 109 As Caskey notes, most payday lenders prior to the 1990s were check cashers that made payday loans as a casual extension of their core business; he estimates that there were probably fewer than 200 at the time. 110 By 2000, there were more than 10,000 payday lenders doing business in the U.S., with $ 2 [*150] billion in revenue. 111 Major pawn chains recently have entered the payday lending business. 112 Payday lenders annually make about sixty-five million loans to between eight and ten million households, totaling more than $ 10 billion in loan value in recent years. 113 The industry reports gross margins of 30%-45% of revenue, with losses at 1%-1.3% of receivables and return on investment of 24%. 114 As in the check-cashing industry, consolidation has created a few large payday lender chains with an important presence across markets, although the industry overall remains rather fragmented. The three largest "monoline" chains have 2,600 outlets combined. 115 While the payday loan industry nationwide grew significantly in the 1990s, its growth was uneven, with widely varying penetration rates in different states. Florida and Illinois, for instance, are each home to about 500 outlets, while the smaller states of North Carolina and Missouri have 900 and 800 outlets, respectively. 116 The State of California, home to about one in eight persons in the U.S., is home to about one in five of the nation's payday lenders. As Comptroller of the Currency John D. Hawke, Jr., has noted, "California alone has more payday loan offices - nearly 2,000 - than it does McDonalds and Burger Kings ... ." 117 This patchwork of growth and concentration may be related to divergent and changing state regulations governing payday lending. 118 Payday loan prices in a number of states would have routinely exceeded the statutory limits on permissible interest rates codified in state usury laws. In the 1990s, however, the industry focused on carving out exceptions from these laws for their loans. 119 Moreover, payday lending has occurred at rates above state usury ceilings through arrangements [*151] between out-of-state banks, thrifts, and payday loan originators. Generally speaking, under the National Bank Act, the payday lender may arrange a loan between the bank and the borrower at terms that are subject to the interest rate ceiling (if any) in the home state of the bank, not to interest rate ceilings of the state in which the payday lender is located. 120 Under many of these arrangements, however, once the loan is booked, the payday lender immediately purchases the entire loan from the bank, with the bank retaining little or no risk. In effect, the lender has "rented" the bank's name for purposes of making a legal loan. 121 State-level efforts to restrict payday lending had been stymied by partnerships between national and state-chartered banks and thrifts, and payday lenders, many of which have now been shut down by banking regulators. 122 The advent of bank-nonbank partnerships in payday loan origination has led to increased technological sophistication to compete with payday lenders that approve loans "on the spot." When a potential borrower completed a loan application at an ACE location, ACE transmitted the borrower's data electronically to its former partner Goleta National Bank. 123 If Goleta approved the loan, it opened a bank account in the name of the borrower, and activated a debit card and PIN connected to that account. ACE delivered the card and PIN to the borrower, who could withdraw the funds at the store or at another retail ATM. The annual report states that this process, from start to finish, took only twenty minutes. The drive towards consolidation in the payday lending industry might be expected to improve industry standards. Smaller independent businesses may be more likely to write out loan forms manually, which may increase the likelihood of error and violation of truth-in-lending laws. Smaller firms also may not have access to the TeleTrack service, a tool that many lenders use to reduce their risk by determining whether the applicant has other outstanding payday loans or credit problems. 124 In theory, partnerships with insured depositories could have improved standards, but evidence to date suggests the opposite. For example, a [*152] central basis for the OCC's termination of one bank's payday lending partnership was that the bank was rolling over payday loans multiple times without any assessment of the borrower's ability to repay. 125 What explains the incredible boom in payday lending in the 1990s? A combination of factors was at work, probably not dissimilar from the factors that Caskey found were responsible for the rise in check cashing in the 1980s and early 1990s. 126 Deregulation in the banking industry increased competition and decreased the availability of less-profitable products, such as short-term, small loans. 127 Retailers have largely replaced sales installment contracts with sales by credit cards, limiting financing options for those without credit cards. 128 Finance companies, while once essential providers of small loans, over the last decade have focused on home equity financing, 129 which can be an important source of liquidity for consumer purchases or debt consolidation, but only for those who are home owners and have equity in their homes. A number of studies in the 1980s and early 1990s found that nearly 20% of U.S. households were credit-constrained. 130 Moreover, a growing number of individuals have little to no liquid savings. In a financial emergency, they have no "backup" funds to meet their immediate needs, and may see a payday loan as the only viable solution. 131 Moreover, the number of borrowers with adverse credit histories is on the rise. 132 Some payday borrowers may be bad risks. Despite the rapid growth of credit card availability, even among low-income families, many of these credit-impaired borrowers are not able to take advantage of credit alternatives such as credit card advances or overdraft lines on their checking accounts. For some credit-impaired individuals, establishing a history of managing a bank account, coupled with financial education may permit them over time to gain access to credit. For others, such access is unlikely ever to occur. On the supply side, payday lending is a highly profitable enterprise, [*153] with a return on sales of 30%. 133 Check cashers may see payday loans not only as a profitable product, but also as a way to diversify their customer base, especially in light of a long-term relative decline in the market for checks (as compared to electronic payments). 134 2. Customers The first thing that distinguishes payday loan customers from many check casher customers is that the former must, by definition, have bank accounts. Lenders are generally unwilling to advance funds to individuals who cannot provide them with proof of account ownership. While payday loan consumers are not unbanked, they could well be referred to as "underbanked": They may lack the savings, credit history, or financial know-how to avoid purchasing a high-cost credit instrument. Customers that use payday lenders tend to be low-or moderate-income, younger than the average age of the population, and otherwise credit constrained. One study found that half the customers surveyed had household incomes between $ 25,000 and $ 50,000. 135 Average annual income for customers in studies done by several states was consistently lower, around $ 25,000 in each case. 136 Most customers were below the age of forty-five. 137 While younger than the U.S. population as a whole, customers were well into their working life. With respect to ownership of assets and access to alternative forms of credit, the CFSA study reported that 42% of respondents indicated that they owned their home, 138 consistent with the Illinois finding. In Wisconsin, 26% of respondents were homeowners. Some payday borrowers could tap into their home equity for emergency credit, but bad credit records may preclude that option for others, and the sub-prime home equity loans held by some of these homeowners have their own high costs [*154] and risks. 139 In the Illinois study, only 11% of customers had a bank-issued revolving credit card, 140 much lower than overall rates reported for the general population even at the lowest income levels. 141 In one study, 56% of payday customers had credit cards, but payday borrowers were three times as likely to have debt payment-to-income ratios of 30% or higher, and four times as likely to have declared bankruptcy, as compared to the adult population at large. 142 Another study, controlling for socioeconomic variables, indicated that African-American families, families who had bounced one or more checks in the past five years, and families in neighborhoods where new check cashers and payday lenders had opened were significantly more likely to borrow from a payday lender. 143 Some anecdotal evidence suggests that the decision to use payday loans may be influenced more by past credit problems than by income. 144 In summary, payday loans are not products for the poorest of the poor. But they seem to be an increasingly popular credit tool among a growing moderate-income working population that has credit problems, often has little savings, and may view payday loans as a convenient, or perhaps only, option for accessing cash in a financial crunch. 3. Costs Payday loans carry high implicit annual interest rates. A 2001 survey of payday lenders revealed that nearly all charged APRs in excess of 300%. The most common APR quoted by lenders was 390% - the equivalent of a $ 15 fee on a two-week $ 100 loan. 145 Nearly a third, however, quoted fees that amounted to APRs of at least 500%; the average APR was 470%. At an average loan size of about $ 300, the average fee for [*155] the average loan is about $ 54. 146 These APRs are, of course, high compared to more traditional credit options like credit cards. 147 This is due, in part, to the higher transaction costs associated with underwriting and servicing payday loans compared to other forms of credit. Payday loans, unlike credit cards, require lenders to interact face-to-face with borrowers each time they originate a new payday loan. They need to conduct more follow up with borrowers than other lenders, and must charge enough to cover loan losses. 148 Notwithstanding these differences, high store profitability indicates that prices may be higher than one would expect in more efficient segments of the financial services market. Further research is warranted on possible barriers to further price reductions, including the possibility that variations in state laws raise the costs to national chains seeking to pursue payday lending on a national basis; the possibility that disclosures are not adequately policed so that consumers are not fully informed of prices; and the possibility that these price structures are inherent in the labor-intensive nature of the transaction. Payday lenders argue that their prices are comparable to one possible alternative for a cash-strapped consumer - bouncing a check. 149 Bounced check fees, according to a Federal Reserve study, averaged $ 20.73 at banks in 2001, 150 so depending on the face value of the bounced check, a payday loan could be a more or less expensive short-term option. Moreover, bouncing a check is not the only response other than a payday loan to the problem of credit constraints. 151 Payday lenders also argue that an annualized percentage rate is not a fair tool for assessing the price of short-term credit. 152 APRs are widely used for other short-term credit, [*156] however, such as the monthly balance on bank-type revolving credit cards, even though nearly 60% of card holders pay off their balances at the end of each month. 153 Many borrowers, moreover, take out payday loans repeatedly throughout the year. The high incidence of loan renewals, or "rollovers," in the payday loan industry is one of its most salient features. Upon maturity of the loan, many borrowers find themselves unable to repay the loan principal in full. As a result, the lender allows them to pay the finance charge on the loan, and to roll the remaining principal - plus a new finance charge - over into a new loan. A "same-day advance" is a functional equivalent of the rollover. The borrower pays the loan in full, but that same day takes out another payday loan in an amount equivalent to the balance paid earlier. Still other borrowers pay off the loan with proceeds from another payday lender. Evidence from multiple states points to the fact that significant proportions of payday loan consumers roll their loans over on a frequent, if not habitual, basis. 154 A study of payday borrowers in Illinois found that the median borrower had more than ten loan contracts over a two-year period, and that one-fifth of borrowers had twenty or more contracts in that time. 155 In Wisconsin, 56% of payday borrowers took out at least eleven loans in one twelve-month period. 156 In Indiana, 77% of all payday transactions were rollovers, and the average annual number of loan renewals was ten. 157 In North Carolina, the typical payday loan customer took out seven loans in one year from one lender. 158 The CFSA study found that three-quarters of payday borrowers rolled over their loan at least once, and that 30% had seven or more rollovers. 159 Using the [*157] Wisconsin statistic as an example, the typical payday loan consumer, who takes out eleven two-week payday loans per year, for the average loan amount of $ 300, at the average 470% APR from the Consumer Federation of America (CFA) survey, spends nearly $ 600 annually in fees. Frequent-use customers are the revenue drivers for payday loan businesses. In North Carolina in 2000, 40% of all payday loan revenues were generated by the 18% of customers who took out an average of at least one loan per month. 160 Each 1% increase in the share of customers who borrow at least monthly from the company increased the outlet's bottom line by $ 790. 161 The frequent use of payday loans should perhaps not come as a surprise. On its face, the typical transaction appears to be the product of underwriting that assumes that the borrower will not be able to repay the loan within two weeks, but will have to rollover the loan. Most lenders would be rightly skeptical that a moderate-income borrower who turns to a payday lender for $ 300 would be able to afford an additional $ 50 out of her next paycheck to cover the finance charge, beyond the $ 300 balloon payment she must make to repay the principal, a mere two weeks hence. If she is in fact able to repay the loan principal and the fee on time, the amount she pays may be enough to send her back to a payday lender when cash runs short before her next pay day. Over time, one would expect a market with returns on sales exceeding 30% to attract new entrants who charge lower prices, or to convince existing participants to lower prices to attract new customers. Currently, competition takes place for location, convenience, and, perhaps, size of loan. Price competition would lead to downward pressure on prices marketwide, presumably making these loan products more affordable for credit-constrained families. At the same time, however, the available evidence indicates that frequent users of payday loans account for a disproportionate share of industry revenues, and that stores in search of greater profits would market repeated use to more of their clients. If product prices were lowered by an appreciable amount, more customers would be able to repay their loans, and the number using the products frequently to repay prior loans would decrease. However, the number of customers using payday loans for other uses would presumably rise with a decline in loan costs. Consolidation in the payday loan industry suggests that the large chains would increasingly use proprietary technology to [*158] deliver loans faster and reduce losses. Even with greater competition among payday lenders, however, loan prices might remain high, given that borrowers usually are not able to develop positive credit histories. Payday lenders usually do not report these histories to the credit bureaus. Positive credit histories could otherwise be used to lower their cost of borrowing by seeking better rates from competing lenders based on their solid credit history. 162 Moreover, lower prices alone would not address the basic problem created by these short term loans - the debt trap most borrowers find themselves in as they repeatedly rollover payday loans during the course of the year. 4. Regulation The regulatory landscape for payday lenders is evolving, as states react in divergent ways to the growth of payday lending, and as the OCC and other federal bank and thrift regulators respond to partnerships between AFS providers and insured depository institutions. The evidence on the prevalence of rollovers has led many states to adopt limits on the number of consecutive times a payday lender may renew a loan, and has led the industry's trade association to adopt a four-rollover limit in its "best practices." But these efforts have been to little effect. These rules leave open a big loophole: They do not bar lenders from accepting cash or a check from a borrower to "pay off" the existing loan and then immediately providing a "new" payday loan. Same-day advances do not appear to be covered under state laws, and the industry "best practices" are silent on the matter. Moreover, the rules do not prevent another firm from providing a payday loan to pay off the first firm's loans. State-enacted rollover limits, perhaps as a consequence, do not appear to affect the percentage of payday borrowers renewing loans or the average number of loans taken out. 163 State regulation of price has not fared any better. There has been a great deal of state legislative action over the last few years regarding payday lending, but no clear trend in state laws is emerging; some states tightened restrictions while others loosened them to permit greater flexibility for payday lenders. 164 Seventeen states, Puerto Rico, and the Virgin Islands have small loan interest rate caps or other usury limits that effectively prohibit payday loans; 165 five states have no small loan or usury [*159] cap but require licensing of lenders; 166 and twenty-eight states and the District of Columbia have specific laws or regulations authorizing payday lending. 167 Of the twenty states and the District of Columbia in CFA's survey, six states have usury laws governing small loans; two states had no laws governing payday lending; and twelve states (and the District of Columbia) have implemented payday lending laws or regulations. 168 Table 1. Average Payday Loan APRs by State Regulatory Environment, 2001 [SEE TABLE IN ORIGINAL] In the states surveyed in which usury ceilings were low enough to effectively prohibit payday lending, rates on payday loans to residents of that state were, paradoxically, the highest. Payday lenders in these states were either violating state usury laws or had partnered with insured depository institutions that "exported" the regime of the bank's home state to originate payday loans, effectively operating outside state usury laws but with added costs incurred to form partnerships with out of state lenders. In the states surveyed that had no usury ceilings for small loans - and therefore effectively lacked any price regulation for payday loans - payday lenders generally charged higher-than-average interest rates on loans. In states in which safe harbor laws or regulations permitted payday lending but capped fees, interest rates were somewhat below average. State usury laws seem to have perverse effects on payday loan pricing and [*160] operations, while states with licensing laws fare somewhat better than average in terms of surveyed prices. State usury and payday lending laws collided with federal bank and thrift regulation as a result of partnerships between payday lenders and a small number of depository institutions. Initially, a handful of national banks were exporting high interest rates to payday lending companies in states with usury laws, 169 thus circumventing those states' implicit restrictions on payday lending. 170 Some small, state-chartered banks had also become involved in payday lending, 171 as had some thrifts. At first, the regulators issued guidance explaining the risks involved in payday lending. 172 After about a year of experience supervising these partnerships, the regulators became increasingly concerned with the risks involved. The OCC, 173 and then the OTS 174 and Federal Reserve Board, 175 in turn, all effectively ended these partnerships. Today, only FDIC-regulated depositories are still engaged in this market. 176 One state-chartered bank, after being ordered by the Federal Reserve to end payday lending, [*161] withdrew from the Federal Reserve System and was approved by the FDIC for federal insurance as a state non-member bank in order to continue to operate its payday lending. 177 What is the basis for shutting down these partnerships? Regulators have cited three basic reasons. First, the OCC has explained that banks should not "rent" their names to payday lenders to evade state usury and consumer protection laws, 178 but should take responsibility as the lender for ensuring proper underwriting and disclosure, as well as appropriate consumer protections. 179 Second, the OCC argued that firms engaged in such partnerships are exposed to "significant reputation, strategic, transaction, and compliance risk" when "nonbank vendors may target national banks ... in order to avoid state law standards that would otherwise apply to their activities." 180 Third, the OCC warned that payday lending could be unsafe and unsound. The OCC put national banks on notice that it reserved the right to examine and regulate both the bank and the third party service provider and to assess special fees for such supervision. 181 Using its safety and soundness supervisory authority - rather than any consumer protection rationale - the OCC has now shut down all known national bank-payday lending operations, and the OTS and Federal Reserve have followed suit. It remains to be seen whether the FDIC will find similar safety and soundness concerns after it has experience in supervising banks engaged in these partnerships. In the meanwhile, questions regarding federal pre-emption are unlikely to go away. Before it had ended national bank-payday [*162] partnerships, the OCC had taken the position that the National Bank Act did not pre-empt state usury claims against the payday lending partner of national banks. 182 Plaintiffs had sued the payday lender under state law. The District Court remanded to state court for lack of federal question jurisdiction 183 and ACE settled, agreed to pay $ 1.3 million in restitution, cease its relationship with Goleta National Bank and comply with Colorado's licensing and usury laws. 184 By contrast, in another case, the same parties successfully argued that the National Bank Act pre-empted state law claims against both the national bank and the payday lender. 185 Furthermore, in Anderson v. H&R Block, 186 the court held that plaintiffs' claim that H&R Block and Beneficial National Bank's refund anticipation loans were usurious was governed by the National Bank Act, but since there was no complete pre-emption, the claims could be heard in state court. The Supreme Court overturned the Eleventh Circuit in Beneficial National Bank v. Anderson, 187 holding that the National Banking Act fully pre-empted state law claims for usury pleaded against national banks and thus provided a basis for removal of the case - including supplemental state claims against other defendants - to federal court, where presumably the defendant believes the claims will bear less weight. 5. Reforms With respect to any remaining bank or thrift partnerships with payday lenders, 188 given the bank and thrift regulators' strong assertions of pre-emption of state payday lending laws affecting insured depository institutions, it is incumbent on the regulators to use their authority under the Federal Trade Commission Act to take action against banks and thrifts that are engaged in "unfair and deceptive trade practices" in the course of [*163] payday lending activities. 189 Regulators should pay particular attention to the problem of short-term balloon payments, repeated refinancing, and inadequate or misleading disclosures under the Truth in Lending Act (TILA). 190 In addition, greater attention to the CRA service test could help to shed light on bank practices. 191 For example, if repeated rollovers indicate that the lender failed to underwrite the payday loan by determining a borrower's ability to repay, contrary to safety and soundness guidelines, 192 then the bank may have engaged in an "illegal credit practice" 193 for purposes of the CRA. Such an illegal practice should adversely affect the bank's CRA rating. Congress should also consider legislation mandating that payday lenders report borrowers' performance to the credit bureaus, so that responsible borrowers have the opportunity to pursue alternative credit products based on their credit history. With respect to state regulation, the picture is mixed. Although a number of states have sought to invalidate bank-payday lender partnerships, 194 given the Court's strong interpretation of the pre-emptive effect of the National Bank Act on state usury laws, 195 these laws are unlikely to withstand legal challenge. State regulators are likely to be more successful in directly acting against the non-bank partner in such arrangements. 196 Given the ineffectiveness of state rollover laws, some states are now focusing on legislation that would provide for longer minimum terms for payday lending to prevent short-term balloon loans that are repeatedly refinanced from becoming a "debt trap." In my view, over the long term, banks could compete with payday lenders by offering alternative, lower cost and lower risk products. In principle, one such alternative might be bank overdraft protection. Although there is currently much controversy surrounding the adequacy of [*164] disclosures and the cost of current bank overdraft policies, 197 in theory overdraft policies could be provided at lower cost than payday loans because, since there is no need for face-to-face interaction, the transactions can take place electronically and automatically at low risk and cost to banks. Moreover, repayment of the overdraft could be scheduled so that regular minimum payments (through automatic debiting of the customer's account) repay the overdraft over a reasonably long time period, rather than the current payday loan of two weeks or bank overdraft practice of thirty days. Overdraft protection should also be disclosed as an extension of credit using APRs consistent with the requirements of TILA. C. Title Lenders Title lenders represent a variation on payday lenders. 198 Instead of holding a check or a debit authorization until payday, title lenders hold collateral - in most cases, an automobile title (and/or the keys to the car, or in some cases a device permitting the title lender to disable the car) - for a typical term of one month. 199 Some title lenders loan money collateralized by other household assets, such as appliances. 200 Title loans range from $ 250 to $ 1,000, and are generally over-secured. If the borrower fails to repay the loan at maturity, the lender will often extend the loan for another fee, in the same way that a payday loan is rolled over. Should the borrower be unable to make payment, or should the lender decide to stop renewing [*165] the loan, the lender repossesses the collateral, sometimes "retaining the proceeds of the sale, even if the value of the automobile exceeds the loan amount." 201 Some auto title lenders sell repossessed cars in the retail market. 202 The title industry grew out of pawnbrokers' efforts to lend larger amounts than televisions or jewelry could collateralize. 203 In a traditional pawn transaction, the pawnbroker makes a fixed-term loan to a consumer who leaves collateral in the hands of the broker. If the customer does not repay the loan at maturity, the collateral becomes the property of the broker. 204 Given the similarity between pawn transactions and title loans, title lenders have been able to claim the advantage of pawnbrokers' exemptions from, or special limits under, many states' usury laws, 205 while using quite different underwriting processes. 206 The title loan industry originated in the southeastern United States, and burgeoned most rapidly in Florida. Between 1995 - when legislation was adopted to legalize the industry - and 1999, 600 title loan outlets opened in Florida. In 2000, however, the state passed a new law limiting allowable interest to 30% annually, and the practice has all but disappeared since then. 207 Yet it still thrives in Georgia and Tennessee, the two states that legalized the practice early. 208 Title lenders have sprouted up in other states - for example, in Missouri, Illinois, and Oregon. While no title lenders are publicly held, Title Loans of America is the largest lender. In 1999, the firm had 300 outlets. 209 Prices for title loans appear to be similar to those for payday loans. [*166] The OCC has found that rates often exceed 25% per month, 210 for an APR of 300%. In Illinois, for instance, a 1999 survey revealed an average APR of 290% on title loans. 211 In Florida, the typical fee in 1998 for a one-month $ 400 loan was $ 88, or 264% APR. 212 Some title lenders have developed partnerships with national banks, raising concerns similar to those raised by partnerships with payday lenders. 213 Several lawsuits and articles document the fact that problems with rollovers are just as prevalent as in the payday lending industry. 214 As with payday lending, title lending is often undertaken without an assessment of the borrower's ability to repay (other than by seizure of the collateral). 215 With title lending, however, the borrower risks losing her car, which may be her regular way to get to work, and to transport children to and from school or child care. Alternative credit products or savings might help low-income families to reduce reliance on title lenders. D. Tax Preparers and Refund Anticipation Lenders The EITC provided a critical supplement to income for twenty million low-income households this year. While tax preparation firms provide important services to low-and moderate-income persons, tax refund anticipation loan (RAL) fees lower take home pay from the EITC, cutting against the distributive goals of the program, and may somewhat reduce its effectiveness as a work incentive for RAL borrowers, although further empirical research is needed to explore this question. 1. Industry Overview Tax preparation services can be distinguished from other AFS services in a few important ways. First, they are the only type of provider examined in this Part whose core functions are not usually thought to be providing services for income receipt, conversion of income into cash, bill payment, saving, or credit. Tax preparers do, however, play important roles in each of these financial services. Tax preparers facilitate the taxpayer's receipt of income tax refunds; they help to convert tax refunds owed to taxpayers into liquid form. They transmit payments to Treasury [*167] for sums taxpayers owe on their tax returns. They arrange for credit to be provided to many taxpayers in the form of refund anticipation loans. Second, tax preparers also cater to middle or upper-income clientele, as well as the lower-income population that is the focus of this paper. Third, they generally cost less on an average annual basis for lower-income clients than check cashers or payday lenders cost for the typical client they serve. In large part, this is because most customers seek their services only once a year, at tax time. Fourth, low-income client use of tax preparation services and refund anticipation loans offered by such preparers appears to be higher among recipients of the federal EITC than other taxpayers, although almost all taxpayers earning under $ 30,000 per year who file tax returns would need to file a return even absent the EITC. The federal EITC is a wage subsidy provided to families who earn under about $ 35,000. In tax year 2001, the credit provided over $ 30 billion to over eighteen million families through refundable credits against federal income tax. The average family with children that year earned a credit of nearly $ 1,800. Unfortunately, the credit and its rules can be difficult to understand for families who have complicated living arrangements, such as children who spend time living with a parent and another relative, who have low levels of education, or who do not speak English as their first language. 216 Conflicting and complex rules governing different tax provision rules for determining household status, dependents, and the like make tax preparation services attractive. Additionally, low-income families may not understand the refund process or timing, or may worry about increasing IRS audits of EITC claimants. For the low-income population, tax preparers provide two major products and services. The first is return preparation and filing of what has become for some a complicated task of filing a tax return, including an EITC claim. Typically, preparers will fill out a client's federal return, the accompanying Earned Income Credit (EIC) schedule, and a state return, if the client is required to file or is eligible for a refund from the state. In most cases, preparers file low-income clients' returns electronically, so as to expedite the processing of the refund by the IRS. As many as 67 to 68% of EITC recipients hired a commercial tax preparer to prepare their returns, and more than half of all EITC recipients filed their returns electronically. 217 [*168] Second, many EITC recipients also use refund anticipation loans (RALs) and similar products marketed by many tax preparation services, including the two large national chains, H&R Block and Jackson Hewitt. RALs are quite similar to payday loans in that they provide advances on a borrower's anticipated income - in this case, a tax refund. In the case of the RAL, the loan is repaid when the IRS issues the borrower's expected refund. RALs serve three main purposes: First, customers are usually able to receive cash proceeds from their loans within two days or less of electronically filing their tax returns, 218 which is eight to ten days sooner than if they had requested direct deposit of their refund to their bank account, if they had one. They may need the funds for daily needs, to catch up on recurring payments on which they have fallen behind, or to repay other short-term loans. Families who receive the EITC may live from paycheck to paycheck and may be unable to save for large purchases such as a car, and thus many seek assistance in getting quick access to the relatively large refund dollars. 219 Second, RALs also permit taxpayers without bank accounts - and, consequently, without direct deposit capabilities - to obtain their refunds without waiting approximately four to six weeks for a paper check from the IRS. 220 Third, taxpayers who do not [*169] have the funds to pay for tax preparation services up-front, but believe that they need help filing for the EITC, find RALs and similar products necessary simply to pay preparers to file for their refund. Tax preparation fees are deducted from the proceeds of the RAL, encouraging tax preparers to work with low-income customers, who in turn can more easily pay for the services provided. 221 Thus, the complexity of the EITC and the desire to have forms professionally prepared may itself drive some decisions to take out RALs independent of a desire to obtain a quicker refund. The need for commercial preparation could be reduced for some EITC recipients, who could request that the IRS calculate the credit based on their EIC schedule. 222 This is an option that few currently pursue, however, and EIC schedules are quite complicated. 223 Tax preparation services and refund loans can consume a nontrivial portion of an EITC recipient's refund. A survey of providers in Washington, D.C. in 2002 found that the preparation and electronic filing of federal and state returns, and associated schedules, cost low-income taxpayers about $ 100 on average. 224 The purchase of a RAL for an anticipated $ 1,500 refund added roughly $ 90 to this amount. Thus, for EITC recipients filing electronically and choosing to take out a RAL, total fees would consume an average of 13% of the EITC or nearly 8% of the total refund from the EITC and other credits. 225 Annual percentage rates on RALs are generally in the 150% to 300% range, depending on how quickly the IRS processes the refund, and thus how quickly the loan is repaid. 226 In addition, for the estimated 22% of EITC recipients who lack a bank account, or four million households, the additional fee to cash a $ 1,500 RAL check issued by the bank partner of the tax preparer would be [*170] at least $ 30 on average at a check casher, despite the low-risk nature of the checks. 227 Another study found EITC recipient use of check cashers to be double that rate, at 44.5%, 228 and CFA reported double that cost for cashing refund checks, at $ 67 on average. 229 The refund anticipation and electronic tax preparation and filing industry is growing. 230 The commercial tax preparation firms earned $ 357 million in fiscal year 2001 from refund anticipation loans, more than double the amount they earned in fiscal year 1998. 231 Given the high demand for tax preparation services by EITC claimants, electronic tax filing and preparation services are disproportionately represented in neighborhoods with concentrations of EITC recipients. 232 Tax preparers that emphasize refund loans are concentrated in low-income neighborhoods. In zip codes with relatively high concentrations of EITC recipients, there are 50% more electronic tax preparation services per filer than in low-EITC zip codes. 233 RALs are used by a significant portion of EITC recipients. In 1999, 38% of EITC recipients received a refund loan, compared to only 4% of other taxpayers. Seven and one-half million EITC recipients took out RALs. Nearly half (47%) of all EITC dollars were received through a [*171] RAL. 234 All told, an estimated $ 1.75 billion in EITC refunds in 1999 was spent for tax preparation, electronic filing and tax refund loans, by those EITC recipients who use these services. 235 Check cashing fees would add a further $ 120 million to the total reduction in EITC benefits reaching low-income families without bank accounts. Commercial tax preparers do provide an essential service to EITC recipients who do not understand how to file for the credit, or who want to have a professional fill out their forms. 236 As well, many EITC recipients who would otherwise be unaware of the credit may learn of it through marketing of tax preparation services in their neighborhoods. Thus, tax preparers may contribute to the high take-up rate for the EITC among eligible persons. Nonetheless, a significant part of the federal government's EITC expenditure is used by low-income families to pay commercial tax intermediaries for filing and refund loans. Viewed as a cost of compliance with the EITC program, EITC claimant expenses for tax preparation and RALs would swamp governmental costs of administering the EITC, although the total operating costs of the EITC program - including both taxpayer compliance and governmental administration - would still be only about half the costs for governmental administration alone of other major low-income programs, such as food stamps and welfare. 237 EITC compliance costs could be reduced. [*172] Simplification of the EITC could reduce the need for tax preparation services and should remain on the congressional agenda. 238 With respect to refund loans, some portion of EITC recipients may need their refund cash immediately to meet an emergency expenditure. Many recipients, however, are likely unaware that they could receive their refund within ten to 14 days if they were banked and filed electronically through IRS Direct Deposit. The IRS could diminish demand for RALs by speeding up refunds and advertising refund times. In addition, many EITC recipients are probably unaware that they are receiving a loan against their refund as opposed to an expedited refund itself. 239 For example, H&R Block was found to have used misleading advertising and to have trained its tax preparers to focus on "rapid refunds" rather than explaining that the products were loans. 240 While H&R Block changed its practices to clarify the nature of the refund loan transaction, some observers believe that its advertising remained somewhat unclear, 241 and most independent agents and smaller firms may be able to escape scrutiny altogether. Enhanced disclosure may help avoid some consumer mistakes. In considering the role of refund loans in the marketplace, it is useful to contrast them with payday loans. Both are an expensive short-term source of credit, used primarily by low-and moderate-income families with no savings. Payday loans arguably represent a product that is absent elsewhere in the marketplace - small unsecured loans for people with blemished credit histories. The main problem is in their structure - a short-term balloon with high fees that often leaves consumers in a cycle of perpetual debt. Refund loans, on the other hand, rarely turn into long-term problems for taxpayers or credit problems for the lender; the loan is fully collateralized by the payment due from the IRS, a reliable payor. 242 By aggressively marketing RALs, however, and - at least in some well documented instances - disguising their nature, 243 some tax preparers seem to have capitalized on a lack of information among low-income families. As with payday lending, RALs are offered by tax preparation firms in partnerships with banks. IRS rules bar tax preparers from directly [*173] providing RALs, 244 and using the bank charter, among other things, permits tax preparers to offer RALS without needing to comply with local usury laws. 245 2. Regulation The IRS role in this area could be critical. The IRS ended its practice of providing notice of anticipated refunds, and an indicator of any offsets for child support or other federal debts, to tax preparers in 1995 in order to reduce fraud and other problems associated with RALs. 246 But the IRS responded in 1999 to congressional electronic filing mandates by again providing tax refund and offset information to preparers as an inducement to expand electronic filing, and to provide the IRS with information on their detection systems. 247 The IRS also delayed EITC refunds in order to conduct basic anti-fraud and error detection. 248 These changes may have increased the supply of tax preparers willing to prepare low-income returns because they could be paid up front from the proceeds of a RAL. The changes also may have increased the demand for RALs as well because delaying refunds exacerbated low-income persons' need for cash. 249 The IRS regulates aspects of RAL transactions through its oversight of electronic return originators (EROs). EROs are authorized by the IRS to file returns electronically on behalf of taxpayers, and receive a number of benefits from the IRS, including promotional material, permission to use the IRS e-file brand name, and indirect benefit from public service [*174] announcements advertising electronic filing. 250 The IRS requires preparers to disclose that "RALs are interest bearing loans and not substitutes for or a faster way of receiving a refund." 251 The IRS permits RALs to be repaid by direct deposit into special RAL accounts held by the bank partnering with the tax preparer, rather than the taxpayer. 252 The tax preparer (or related parties) cannot make the RAL directly, nor can the tax preparer directly cash a refund check issued to a taxpayer whose return the filer prepared. (A preparer that is also a financial institution may cash a refund check, but not if the preparer has made a RAL.) 253 The IRS also regulates RAL fees. Under IRS rules, authorized providers: may not base their fees on a percentage of the refund... . Separate fees may not be charged for Direct Deposits. An Authorized IRS e-file Provider may assist a taxpayer in applying for a RAL and may charge a flat fee for that assistance. However, the fee must not be related to the amount of the refund or a RAL. The Provider must not accept a fee from a financial institution for any service connected with a RAL that is contingent upon the amount of the refund or a RAL. 254 The IRS also regulates the advertising of RALs. The IRS "prohibits the use or participation in the use of any form of public communication containing a false, fraudulent, misleading, deceptive, unduly influencing, coercive, or unfair statement of claim." In addition, "a Provider must adhere to all relevant federal, state and local consumer protection laws that relate to advertising and soliciting." Moreover, with respect to RALs, the Provider and financial institution must clearly refer to or describe the funds being advanced as a loan, not a refund. The advertisement on a RAL must be easy to identify and in readable print. That is, it must be made clear in the advertising that the taxpayer is borrowing against the anticipated refund and not obtaining the refund itself from the financial institution. 255 3. Reforms The IRS, in responding to congressional pressure to increase e-filing and decrease EITC errors, has helped to create the market for RALs. [*175] It now bears a special responsibility to help end it. There are several steps that the Treasury Department and the IRS could take to improve the manner in which EITC recipients receive financial services. First, and most importantly, Treasury and the Congress need to continue efforts to simplify the EITC, for example, by altering the definition of qualifying children. Simplification should help to drive down error rates, which are costly in their own right and more expensive to EITC recipients who take out RALs incorrectly anticipating a refund, and diminish the need for expensive tax preparation services. 256 Second, the IRS should expand free tax preparation and electronic filing availability; 257 greater availability of these services would, of course, diminish the need to take out RALs in order to pay for preparation services. The biggest barriers to an expansion of free tax preparation services are lack of funds, lack of sites that provide for electronic filing, and, more critically, lack of effective ways to assure the quality of these tax preparation services. Moreover, if Congress wants the IRS to expand e-filing availability, it should pay for expanding the private sector infrastructure necessary to implement it, rather than relying on RAL fees paid by low-and moderate-income tax payers to cover the tax preparers' costs of implementing e-filing. Congress could appropriate funds for the purpose, or use an e-filing tax credit to offset the costs. Third, since the IRS now has the technical capacity to split refunds, the IRS should permit refunds to be direct deposited into more than one bank account. If refunds are permitted to be split into more than one account, tax preparers could compete by offering tax preparation services that are paid not out of the proceeds of RALs, but paid directly to them electronically out of tax refunds through direct deposit to them of a portion of the refund, diminishing the risk to the preparer and eliminating one reason to take out a RAL. If this reform is combined with public and private sector efforts to bring EITC recipients into the banking system, the remaining portion of the refund could be direct deposited into the client's own bank account. Given the large average refund size, the portion deposited into the client's account might serve not only for short-term needs, but also, for some clients, as a base for savings. [*176] Fourth, coupled with better error and fraud detection and prevention efforts, 258 the IRS can speed up EITC refunds, 259 and do more to encourage direct deposit of refunds into bank accounts, both directly 260 and through employers, commercial preparers, and Volunteer Income Tax Assistance (VITA) sites. Again, if Congress believes that more efficient tax processing, in the form of e-filing and direct deposit of refunds, is in the government's interest, Congress should appropriate funds or provide a credit to pay for these improvements rather than letting them be cross-subsidized by fees from RALs and tax preparation. Fifth, as explained in more detail below, 261 EITC recipients can become a central focus of efforts to bank the unbanked. The following steps should be taken to accomplish this goal: Treasury should expand its Electronic Transfer Account (ETA) program to permit use of ETAs for EITC receipt. Congress should appropriate more funds for Treasury's First Accounts program to support innovative efforts to reach EITC recipients without bank accounts. The IRS should establish partnerships with large employers to encourage employees to open bank accounts and establish direct deposit of paychecks and tax refunds. Moreover, the tax preparation firms themselves should partner with banks to develop and offer individual, low-cost, electronically based bank accounts for their clients. Their clients could use the accounts to receive direct deposit of their income tax refunds, to withdraw funds at ATMs and POS using debit cards, to save, and for their other financial services needs throughout the year. The tax preparers would gain a new marketing tool and might see higher rates of client retention. Sixth, the IRS can use its oversight of e-file preparers to improve the market for EITC recipients. Towards this end, the IRS should make enforcement of existing rules, especially regarding advertising, a priority; provide more detailed rules regarding non-deceptive advertising, including disclosures of how the offered product compares with the IRS's current anticipated refund times; and force greater transparency in pricing, including by requiring that RAL funds be provided to EITC recipients in a [*177] form that does not require an additional cost to convert to cash. For example, funds could be transferred to debit or stored value cards that can be used at ATMs or POS without cost. III. The Banking Sector A. Barriers to Banking the Poor While changes in the AFS sector could improve the delivery of financial services for the poor, fundamental change would be accelerated by enhanced competition from banks, thrifts, and credit unions using electronically based accounts to serve low-and moderate-income customers. Mainstream providers can offer a range of services to meet the needs of low-income communities. Today, however, while the banking system works extraordinarily well for most Americans, many low-and moderate-income individuals face a number of barriers to account ownership. There are five key barriers: the structure and price of existing accounts, prior credit problems of the unbanked, low perceived profitability of serving the poor, lack of bank distribution systems in low-income areas, and the need for financial education. First, regular checking accounts may not make economic sense for many lower-income families. 262 Three main problems are high minimum balances, monthly fees, and the risk and cost of bouncing checks. Consumers who cannot meet account balance minimums for a checking account at a bank often pay high monthly fees. Thirty percent of banks offered an account requiring minimum balances for checking, with an average minimum monthly balance of $ 527, and a monthly fee of $ 7.12 for falling below the minimum. 263 Another 38% of banks offered fee-only accounts, charging an average of $ 4.74 per month for checking accounts without monthly minimum balances. 264 The number of banks offering free checking accounts appears to have jumped to 32% of banks by 2001; 265 yet, most of these banks have high minimum balances that low-income [*178] persons cannot meet. Free accounts may also be available only to particular segments of the population, such as full-time students and the elderly. Nearly all banks, including those offering otherwise free accounts, levy high charges for bounced checks or overdrafts that low-income families with little or no savings face a high risk of paying and can ill-afford. Avoidance of these bounced check fees may be an important determinant of the decision to become or remain unbanked. The average fee for checks that "bounce" and are rejected for not-sufficient-funds (NSF) was $ 20.75 in 2001, and the average fee for overdrafts was $ 20.50. 266 In fact, depository institutions target the fee for NSF as a "core fee driver" in generating revenues from checking accounts. 267 Consumer organizations have become increasingly concerned about "bounce protection" plans offered by some banks without adequate disclosure of fees. 268 In addition, a customer depositing someone else's check endorsed to her faces risks if the check turns out to be bad. Customers are charged an average fee of $ 7.11 by 74.1% of banks for checks that the bank customer deposits that are, in turn, returned for insufficient funds in the check writer's account. 269 Unbanked customers may fear that a large fraction of their income comes from parties whom they have little reason to trust will make payment on their checks. 270 Checking accounts are costly for depository institutions to offer 271 and [*179] need to be offset with sufficient revenue (from float and fees) that may not be present in accounts that low-income customers could afford to use. In addition to direct fees for servicing the checking account, financial institutions may also charge high fees for money orders or other products that their typical customers do not often use, but that lower-income consumers use frequently. Moreover, banks hold checks that are not "on us" for a matter of days before depositing funds, unlike check cashing outlets; for low-income customers, the few days wait may not be practical. In addition to fees from checking accounts, banks also usually charge fees for savings accounts. No-fee passbook accounts are available at only 15% of institutions; no-fee statement accounts are available at 17% of banks and savings associations. For other accounts, minimum balances to avoid fees ranged from $ 157.86 for passbook accounts to $ 184.42 for statement accounts. Monthly fees of $ 2.15 for passbook accounts and $ 2.50 for statement accounts were charged for accounts falling below the minimum required balances. 272 There is wide variation in the structure and pricing of accounts and fees across metropolitan areas and states. 273 Fees for checking accounts also vary by the nature and size of the institution. Multistate banks tend to charge higher fees than single-state banks, including about $ 3 more for NSF and overdrafts. 274 Multistate banks are more than fifteen percentage points more likely to offer free checking accounts 275 but at least ten percentage points less likely to offer free savings passbook or statement accounts. 276 Large institutions are now more likely to offer free checking accounts, and more likely to impose higher fees than do medium and smaller institutions, but less likely to offer free savings accounts. 277 The number of large banks reporting that they offer "free" checking jumped twenty-two percentage points from 2000 to 2001, following significant increases in the percentage of small and medium-sized banks offering such accounts in the previous year. 278 These free checking accounts, however, may still be ill-suited to low-income consumers, given minimum balances and high fees for NSF, overdraft, and deposit returns. [*180] Table 2. Fee Structure by Size of Institution [SEE TABLE IN ORIGINAL] The structure of these accounts is a key driver in keeping the unbanked out of the banking system. As discussed above, surveys consistently show that the price of account products and minimum account balance requirements are important determinants of being unbanked. 279 Studies have confirmed that many of the unbanked would become "banked" if they had access to a relatively low-cost electronic account of the type that this Article proposes. These accounts could plausibly be offered by financial institutions with modest governmental incentives to cover start-up costs. 280 In fact, the unbanked have responded to account products tailored to their needs. For example, in Puerto Rico, Banco Popular introduced Acceso 24, an electronic account, with no minimum monthly balance, free direct deposit, unlimited ATM access, and a low monthly fee. The bank has enrolled tens of thousands of low-income customers in the product since 1995. 281 Cultural issues and reluctance to use banks may matter, 282 but many of the unbanked already use, or have used, the banking system. Nearly half the unbanked, according to one study, use banks, thrifts, or credit unions to [*181] cash checks at least some of the time, 283 although far fewer unbanked households in inner city communities use such institutions. 284 Between 48% and 70% of the unbanked have had an account at a financial institution at some time in the past. 285 Existing products, however, are too costly, too risky, or not well-suited to their needs. Moreover, because bank hours and locations may be less convenient for low-income workers than those offered by AFS providers, and branch expansion or longer hours are costly, I focus my proposals on expanding access to financial services through ATMs and other electronic delivery mechanisms that can be expanded to more locations and with longer hours less expensively than bank branches. As discussed below, 286 I propose ways in which such electronically based accounts should be tailored to the needs of low-and moderate-income households if they are to be brought back into the financial services mainstream, or enter for the first time. While high fees pose one barrier to access for the unbanked, a second barrier comes from difficulties that many unbanked persons may have in qualifying for conventional bank accounts because of past problems with the banking system. The ChexSystem, a private clearinghouse used by most banks to decide whether to open bank accounts for potential customers, records that nearly seven million individuals have had their accounts closed for prior problems, such as writing checks with insufficient funds or failing to pay overdraft charges. 287 Records of prior problems are kept in the system for five years, during which time these individuals will likely be unable to open a conventional bank account at most banks, thrifts, and credit unions. While some individuals undoubtedly pose undue risk for account ownership, many potential customers could responsibly use bank accounts. Banks could obviate this concern by working with the private clearinghouses to better distinguish among types of past problems, by offering accounts contingent on completion of financial counseling, 288 and by offering electronically based accounts with [*182] online bill payment or automatic money orders, and without check-writing privileges, that pose little risk of overdraft. 289 To date, banks have been reluctant to take these steps because the expected returns from such accounts are low. Third, while many urban communities contain adequate numbers of both banking institutions and AFS providers, in some low-income urban and rural communities, banks, thrifts, and credit unions are not as readily accessible to potential customers as such institutions are in higher-income areas. A 1997 Federal Reserve Board study found that low-income central city neighborhoods have far fewer bank offices per capita than higher-income areas and those outside the central city. 290 In Chicago, 40% of low-to moderate-income neighborhoods had only check cashers, 32% of low-to moderate-income areas had both check cashers and banks in about equal proportions, and 28% of low-to moderate-income areas had only banks. Areas in Chicago with only check cashers have a greater percentage of minority households than other low-to moderate-income areas in Chicago. 291 Similar patterns may persist in the distribution of ATMs. In New York and Los Angeles, there are nearly twice as many ATMs per resident in middle-income zip codes as there are in low-income | |||||
