Economic Analysis Deems Refund Transfers Valuable to the `Unbanked’

Refund transfers (RTs) are bank products that establish a short-term account in which taxpayers may receive their federal income tax refunds.

Demand for RTs spiked in 2011 following the elimination of the IRS debt indicator and increased regulatory limitations placed on refund anticipation loans (RALs).

This surge in sales has led to a renewed interest in the product from consumer groups and regulatory agencies, though few studies to date have focused solely on the economic characteristics of RTs and their impacts on different types of consumers.

This report combines findings from previous literature with new economic analyses of RTs to study the impact of this product on consumers. The following conclusions summarize the main findings of this research:

• RTs Have Substantial Value for “Unbanked” Individuals. For individuals without a bank account, RTs can increase the speed with which taxpayers receive their refund by 11 days or more. Since many of these individuals then use their tax refunds to pay off high-interest debt, an RT can easily pay for itself in reduced interest and other fees.
• The Desire to Postpone Tax Preparation Fees Is Not a Major Driver of RT Use. One feature of an RT is the ability for taxpayers to pay tax preparation fees out of their refunds, if they choose to do so. The available research suggests that less than one-half of individuals who purchase RTs choose to pay their tax preparation fees out of their tax refund. When combined with other evidence, these facts imply that the postponement of fees is a minor convenience for some consumers rather than a major driver of use.
• On Net, RTs Tend to Improve Consumer Welfare. Like any product, RTs provide great value to some consumers, while being not optimally used by others. Unlike most financial products, though, RTs charge only a one-time fee of about $30, so that the most a consumer could possibly lose is the fee. In contrast, some consumers (who are often those in the most dire circumstances) may gain well over $100 from the purchase of an RT. In addition, RTs may play an important role in encouraging low-income taxpayers to take up federal transfer programs such as the Earned Income Tax Credit.
• RTs Are Priced Similarly to Comparable Bank Accounts Aimed at Unbanked Individuals in Government-Sponsored Programs. The Electronic Transfer Account (ETA) program, as well as a new pilot program at Treasury for tax refunds, subsidizes the creation of bank accounts that still charge very similar annual fees for account maintenance as the fee for an RT.

Basic Characteristics of Refund Transfers

Description of the Product

A refund transfer, also known as a refund anticipation check (RAC) or electronic refund check (ERC), is a financial product offered to taxpayers at the time they file their income tax returns. A bank establishes a temporary bank account for the sole purpose of receiving a taxpayer’s refund via electronic direct deposit from IRS. The customer may then receive the funds via a debit card or paper check.

In addition to the temporary bank account, RTs allow taxpayers to pay for the costs of tax preparation directly out of their refund. ¹Some industry stakeholders are considering offerings that would give RT customers the ability to keep the established account open, if the customer chooses to do so.

Many professional tax preparation firms partner with banks to offer RTs to their customers as part of the standard menu of services. As a result, taxpayers purchase RTs directly from financial institutions at tax preparation firms, at the time of tax preparation. The typical fee for an RT is about $30, regardless of the size of refund or nature of the return.

Patterns of Use

IRS has kept records on RT use since 2006, shown in Figure 1 below. In this time, RT use has nearly doubled, with nearly a 50% increase in the 2011 tax filing season, when 18.8 million taxpayers, or nearly one out of five tax refunds, purchased an RT. This sharp increase reflects a massive substitution from refund anticipation loans (RALs) into RTs following regulatory action restricting RALs.

The rate at which taxpayers purchase RTs varies widely across the population. To get a sense of the magnitudes, I use the most recently available data to examine which characteristics of taxpayers predict RT use. ²This analysis uses 2008 IRS returns data available through the Urban Institute.

The single most important characteristic is the Earned Income Tax Credit (EITC) eligibility, and especially those with eligible dependents. These taxpayers are more than three times more likely to purchase an RT than others. Among this population it is also single individuals, who file as “Head of Household,” who are overwhelmingly the most frequent purchasers of RTs. These taxpayers tend to be poorer and younger than others in the population.

Even after correcting for the higher rate with which EITC-eligible filers purchase RTs, there are a number of interesting patterns among RT customers. Individuals under age 35 are twice as likely to purchase an RT as compared with older taxpayers. Active military members are also significantly more likely than civilian taxpayers to purchase an RT. Finally, even after accounting for an individual’s income, those taxpayers living in poorer neighborhoods are also more likely to purchase an RT.

Economic Analysis of Refund Transfers

Benefits From Refund Transfers

The primary element of a refund transfer is the creation of a temporary bank account for the purpose of receiving a refund through direct deposit. This can significantly increase both the speed and convenience with which taxpayers receive their refunds.

The creation of such a bank account is of greatest value to “unbanked” taxpayers who do not have access to an existing bank account, since without an RT they would be forced to wait for their refund to arrive via paper check. IRS provides electronic refunds in 11 days on average after a return is filed electronically; in contrast, IRS mails refund checks on average 18 days after filing, and so the taxpayer would not receive the check until 22 or 23 days after filing. ³IRS mails checks on Fridays, and so mailing times are somewhat longer as a result of the weekend. If a taxpayer decides to file using a paper return, a paper check would take six to eight weeks to arrive.

One can quantify the value of bank account creation to unbanked customers by calculating the interest charges they would have had to pay in order to borrow an amount the size of their refund for the 11 or 12 days extra that it takes refunds mailed by paper check to arrive.

To give a rough magnitude of the value, suppose that a taxpayer uses her refund to pay off a credit card bill on which she is paying 15% interest. For the 11 or 12 days that she waits to receive her return, she must continue to pay interest on the debt. Suppose that she receives a refund of $3,000, which is a typical figure for those purchasing an RT. If she paid off her credit card debt 11 or 12 days earlier, as she could after purchasing an RT, she would reduce interest payments by approximately $15.

The interest charges would be significantly higher if the taxpayer instead had to make a cash payment, which would require a cash withdrawal from her credit card. The fees for a $3,000 cash advance, repaid after 11 or 12 days, would include an upfront charge equal to 3% to 4% of the withdrawal plus interest payments for the 11 or 12 days, implying a fee of about $100.

And if her refund were further delayed for some reason, she would face additional interest payments. This figure would be higher still if the taxpayer were repaying higher-interest loans such as payday or vehicle-title loans.

While it is difficult to identify precisely which taxpayers are unbanked, it is easier to identify a subset of these who do have a bank account due to their reporting of dividend or interest income on their tax return. The data suggest that those taxpayers reporting dividend or interest income are substantially less likely to purchase an RT.

Taxpayers who use RTs also benefit more as the size of their refund increases. Data suggest that taxpayers with refunds exceeding $3,000 are more than twice as likely to purchase an RT as compared with those claiming smaller refunds. On the other hand, 7% of RT purchasers claimed a refund of less than $500, a small group for whom the benefits are substantially reduced.

Since the primary value of an RT is to receive a tax refund more quickly than possible through a paper check, economic theory predicts that taxpayers should purchase most RTs in the early period in the tax-filing season. If a taxpayer instead waits to purchase an RT, she could have received the refund sooner simply by filing earlier rather than purchasing an RT.

The latest data available bear out this prediction of theory. More than 70% of RTs are purchased within the first month of tax filing, as compared with fewer than 40% of overall returns. This implies that early filers are twice as likely to purchase an RT.

In addition to the speed with which a taxpayer receives her refund, a bank account can also make the receipt of her refund considerably more convenient. For instance, many taxpayers must pay to cash their refund checks before spending the money. A typical check cashing fee is 2.2% of the value of the check, plus $1. For a refund of $3,000, the check cashing fee would be $67. In contrast, a bank account avoids these fees by permitting taxpayers to either withdraw the cash directly or receive a pre-charged debit card. Combining this with the value of credit implies that unbanked taxpayers can value the account creation at $167 or more.

Consumers who already possess bank accounts may also derive benefits from RTs. For instance, taxpayers may wish to take advantage of direct deposit of their tax refund but worry about submitting their bank account details for fear of identity theft. An RT provides an alternative method for receiving one’s refund.

Another aspect of RTs that some consumers may value is the ability to pay tax preparation fees out of their refunds rather than at the point of service by a tax preparer. For this population, tax preparation fees (exclusive of the cost of the RT) can range from $100 to $300, depending on the complexity of return. For instance, researchers have calculated an average tax preparation fee of $187 at H&R Block in this population in 2009. ⁴Urban Institute, “Characteristics of Users of Refund Anticipation Loans and Refund Anticipation Checks” (2010).

The RT delays actual payment by approximately 11 days, but more importantly it defers payment from a period just before the receipt of the refund to the time just after. For many low-income households, tax refunds reflect a large share of annual income, and so these households are particularly liquid just after its receipt. Correspondingly, many taxpayers may be particularly starved of funds just before, which is when they would file their taxes.

It is important to note that more than half of all RT consumers do not use the product to delay payment of tax preparation fees. For instance, recent research estimates that nearly half of RT purchasers did not even use a paid tax preparer to file their returns, instead accessing RTs through home or online tax prep software. ⁵Urban Institute, “Who Needs Credit at Tax Time?” (2010). Other individuals who purchase an RT do use a tax professional but pay no preparation fees, either because they are exempted from them (usually on account of their low income) or because they use coupons and special discounts to effectively reduce the price to zero. Finally, data suggests that a few consumers choose to pay all fees up front. ⁶Data from one of the largest facilitators of RTs in the country show that 2.97% of RTs from independent electronic refund originators do not include tax preparation fees.

The remaining consumers who do choose to delay payment face the same fee as those who do not take up this option. Considering all of these facts, the deferral of payment should be considered a minor add-on feature of an RT for most consumers.

There is a current legal debate as to whether the 11-day delay in payment of the tax preparation fees constitutes a “loan.” ⁷Some have pointed out that the tax preparation service is not complete until the taxpayer’s refund is deposited in the account established by the RT, and thus that there is no deferral of the tax preparer’s fees at all since the transaction is not complete when the tax refund is filed with IRS but only when the refund is deposited. Historically, most consumer groups viewed the RT as a non-loan alternative to a refund anticipation loan (which clearly is a loan). ⁸For instance, see documents from the National Consumer Law Center, such as the Model RAL Act http://www.nclc.org/images/pdf/high_cost_small_loans/ral/model-refund-anticipation-loan-act.pdf. More recently, the Federal Deposit Insurance Corporation has argued that the delay in payment is a loan and should therefore trigger various disclosure and regulatory requirements. ⁹See Republic Bank, Form 8K, filed with the Securities and Exchange Commission May 3, 2011, Exhibit 99.2.

From an economic perspective, it is not clear what distinguishes this situation from a multitude of others, such as legal services, in which payment often occurs slightly after the actual performance of services (such as a monthly billing cycle) and yet the legal requirements of a “loan” do not apply. Nevertheless, the focus of this report is the economic characteristics of RTs, and therefore it takes no stance on the resolution to this essentially legal question.

But whether or not the delay in payment legally constitutes a loan, there is no question that some RT consumers may value it. IRS recently considered expanding the use of “split refunds” to allow taxpayers to cover tax preparation fees out of their refund; however, IRS decided not to expand the program. ¹⁰See http://banktalk.org/2011/06/30/irs-will-not-expand-the-split-refund/. This means that RTs are currently the only way for consumers to take advantage of this option.

The value of this convenience depends critically on the nature of credit to which the taxpayer has access at the time when she files her return. For instance, if a taxpayer has access to a credit card on which she is current (even if she carries a balance), then the 11-day delay in payment is not worth much. ¹¹For instance, the national average credit card interest rate on purchases is 12.83% (taken from http://www.creditcards.com on 10/11/2011); borrowing the average tax prep fee of $187 for 12 days would cost less than $1 in interest. In contrast, if an individual purchased an RT for the sole purpose of delaying payment of tax preparation fees, the cost of this credit would be slightly lower than for borrowing the same amount using a payday loan. ¹²Payday loans usually charge 18% of the face value of a loan, often lasting about the same 11 days as the deferral of payment. Eighteen percent of a typical tax prep fee of $187 is $34. See http://bpp.wharton.upenn.edu/tobacman/papers/payday.pdf.

Charges for Refund Transfers

The financial institutions offering RTs almost always charge a fixed fee for an RT, regardless of the size or any other characteristic of a taxpayer’s return. In the 2011 tax filing season, most taxpayers paid about $30 for an RT.

Perhaps more important than the actual fee structure are two features that are not present in the pricing of RTs. In contrast to many financial products, especially those targeted toward low-income individuals, an RT is a one-time purchase with no continuing relationship or responsibility once the refund is delivered. This arrangement rules out recurring fees that either increase over time or continue long beyond the point when the consumer has finished using the product.

Because tax preparation firms charge all fees in a single, upfront statement, consumers are likely well aware of the fees they are paying. A series of recent papers in economics has identified many products as having “shrouded attributes,” which are aspects of products that increase long-term charges or costs in ways that consumers do not account for when making a purchase decision upfront. ¹³Gabaix and Laibson, “Shrouded Attributes, Consumer Myopia, and Information Suppression in Markets” (2006). For instance, many consumers may purchase a computer printer without adequate consideration for the continuing costs of toner or ink, which may amount to far more than the initial cost of the printer itself. A more dangerous and policy-relevant equivalent is an adjustable rate mortgage that resets to far higher interest rates after a few years. Since interest rates do not increase in this example until several years in the future, consumers may not be fully aware of the charges that they are committing to when initially signing up for the loan. As these examples should make clear, there are no such “shrouded attributes” present in RTs.

Balancing Costs and Benefits
From Refund Transfers

As the preceding discussion makes clear, the benefits of RTs for consumers vary considerably depending on the circumstance. This section puts all of the evidence together to consider what is the likely aggregate impact of RTs on consumer welfare.

The primary beneficiaries from RTs are those without regular access to bank accounts. This includes both “unbanked” consumers, who have no access to a bank account, as well as “under-banked” consumers who have limited access to a bank account. Since neither tax preparation companies nor IRS maintains data on whether a taxpayer has access to a bank account, it is difficult to calculate the exact fraction of RT consumers who are unbanked. The literature suggests that 10% to 20% of the households in the United States have no regular bank account access, but among poorer households the fraction is estimated at 35% to 45%. ¹⁴“Poorer households” refers to households with income below twice the federal poverty line. In 2011, the FPL was $22,350 for a family of four.

Not all taxpayers who purchase RTs come from this lower income range, but those without bank accounts at any income level derive the most benefits and are therefore likely to constitute a disproportionate fraction of customers. I therefore take the midpoint of this range—40%—as an estimate of the fraction of RT customers who are unbanked.

The facts suggest that RTs benefit consumers, on average. If we assume that the average unbanked taxpayer uses her refund to pay off credit card debt, the estimates above imply that unbanked consumers may value an RT at $82. ¹⁵The taxpayer avoids $15 in interest payments and $67 in check cashing fees. If 40% of consumers are unbanked, then consumers gain at least $33 on average from the purchase of an RT. ¹⁶$33 = $82 × 0.4 But this figure is likely an underestimate, since it assumes that the 60% of consumers who do have access to a bank account derive absolutely no value from the product. More likely, some of the banked customers have other reasons for purchasing an RT, and therefore derive some positive benefit.

In addition to the average consumer benefit, it is important to consider the distribution of benefits across those consumers who purchase RTs. As with any product, there are some consumers who benefit more than others, as well as some consumers who may not benefit at all. But there is a key asymmetry between the very large potential gains for those who benefit most from RTs and the limited costs for those who do not.

The benefits from RTs, especially for unbanked taxpayers with debt who are owed large tax refunds, can be very large, amounting to many times the cost. In contrast, a consumer can never lose more than $30 from an RT, even in the rare case that she derives no value from the product whatsoever. This stands in contrast to many other financial products that can, in the worst case, cost the consumer a great deal.

What is more, the largest benefits from RTs accrue to those who are in the worst financial situations. Taxpayers without bank accounts gain the most from RTs, and especially those with high-priced debt. These individuals, often minorities, tend to be quite poor and living in very poor neighborhoods. RTs seem to contribute to the broader array of effects, both public and private, to help such individuals.

Broader Impacts on Consumer Welfare

This report has thus far concentrated on the direct benefits and costs of RTs for consumers. There are a number of broader effects that may be important in evaluating the overall impact of RTs.

Increasing EITC Take-Up

The EITC is the primary income transfer program for low-income individuals in the United States. The program pays a refundable credit equal to a fraction of a taxpayer’s earnings up to a cap. For instance, single mothers with two children receive a tax credit equal to 40% of earnings up to the cap of $5,751.

This credit is reduced once individuals earn more; the EITC for our mother with two children would reduce by $0.21 per dollar earned beyond $16,690, disappearing once she earned more than $40,964. The subsidy rate, the cap, and the phase-out threshold all vary with the number of children and tax filing status of the household.

From this example, two facts about the EITC should be clear. First, the credit can be a very large fraction of annual income. Second, the formula is complicated enough that many taxpayers have difficulty claiming it in the proper way on their own. IRS statistics suggest that 25% of those eligible for the EITC do not claim the credit, with most of these non-claimants declining to file any tax return at all.

IRS goes to great lengths to encourage as many eligible households as possible to claim the credit by sending out letters to seemingly eligible non-claimants but the response rates are rarely higher than 20%. Recent research suggests that the fundamental reason why these households do not claim the EITC is that they simply do not understand its provisions. ¹⁷Bhargava and Manoli, “Why Are Benefits Left on the Table? Assessing Incomplete Take-Up with an IRS Field Experiment,” UCLA Working Paper (2011).

Professional tax preparation firms are an important intermediary between EITC claimants and the tax system. Tax firms ensure that eligible individuals file and claim the EITC. These firms also help taxpayers navigate the rather complicated provisions to claim the correct size of credit. This is a very important step, as incorrectly filed credits can significantly delay payment. One in four EITC claims contain some mistake, and so the Treasury Department recognizes the service performed by tax preparation firms in helping the EITC-eligible population. ¹⁸EITC Due Diligence Program description, http://www.treasury.gov/tigta/auditreports/2011reports/201140023fr.pdf.

Getting individuals to file via tax professionals can contribute to the effective administration of the EITC. For instance, researchers have found that the increase in state electronic tax filing around the year 2000 significantly increased the fraction of low-income taxpayers filing for the EITC, primarily by increasing the involvement of tax preparers. ¹⁹Kopczuk and Pop-Eleches, “Electronic Filing, Tax Preparers and Participation in the EITC” (2007). Note that there is no inherent link between electronic filing and the EITC; but by discouraging self-filers, the policy increased EITC take-up.

RTs perform a similar public function to electronic filing by further encouraging low-income taxpayers to use professional preparation services. RTs increase the share of EITC-eligible individuals who properly file for the credit, thereby helping to implement the policy. Although there is less direct research, it is likely that tax preparers help these same individuals claim other credits to which they are entitled, such as those for education or savings.

Banking the Unbanked

Another broader impact of RTs may be to encourage low-income individuals to become “banked.” RTs offer an important short-run option for unbanked taxpayers to receive tax refunds. Some of these individuals may rightly fear that holding a permanent bank account would cost too much in fees or charges. But for many of these unbanked individuals, it would be cheaper to acquire low-fee permanent bank accounts.

RTs offer individuals the ability to learn about the benefits of debit cards without risk of unexpected banking fees. In other accounts, low-income individuals may be particularly at risk for banking fees due to their inability to maintain minimum balance requirements.

In this way, the RT product is very similar to an electronic transfer account (ETA), an account designed by the U.S. Treasury beginning in 1999 in order to facilitate the receipt of Social Security payments and certain other federal benefits (though not EITC benefits). Individuals hold these accounts at private banks, which may charge a maximum of $3 per month in fees to cover costs. The annual fees from ETAs amount to more than the $30 fee for a bank account through an RT.

This program eases the administrative burden of paying benefits, as well as encourages low-income individuals to gain accounts. After nearly 12 years in operation, the program has been very successful; the Treasury has determined that by 2013 enough recipients will have gained bank accounts that paper checks will be phased out. ²⁰See the ETA program website, https://www.eta-find.gov/.

The success of ETAs suggests that RTs might be even more valuable to the low-income community if they were converted into permanent accounts. ETA providers receive a one-time payment of $12.60 for opening the account to offset administrative costs of the bank or credit union that provides the account. This suggests that a similar plan for tax refunds would require government support as well.

In order to be successful, however, a program must combine federal subsidies with deep cooperation with private-sector partners in the tax refund and bank-product industries. During the tax filing season of 2011, the U.S. Treasury piloted a new program (without such cooperation) offering a low fee debit card on which households could have their tax refunds deposited. ²¹See http://www.treasury.gov/press-center/press-releases/Pages/tg1021.aspx. Much like ETAs, these cards offered a permanent account and carried fees of up to $4.95 per month. But despite the subsidies, fewer than 1 percent of individuals offered an account decided to enroll in the program.

The popularity of RTs suggests that consumers do demand such accounts but prefer products such as RTs already offered in the market. The Treasury might have greater success in the future executing a similar program through a partnership with tax preparation firms or banks offering RTs to take advantage of their familiarity with the market and direct access to consumers. Further research on the behavior of the unbanked may also point to other, similar products that firms could offer in the context of tax preparation that would encourage the unbanked to become banked.