By John N. Friedman, Harvard Kennedy School
Each year, the government transfers more than $300 billion to its citizens in tax refunds at the federal and state level.1
Traditionally these transfers occurred via paper check; more recently, many are transmitted electronically by direct deposit, eliminating processing costs for the government and improving convenience for consumers. But for consumers without bank accounts, or for those unwilling to provide these details to the government, direct deposit is not an option.
In recent years, a number of private companies have offered taxpayers the option to receive their refund via a variety of prepaid debit cards. These services are very popular; in tax season 2012, nearly 20 million taxpayers chose to receive their refunds in this way. Now the federal government and a number of state governments are considering, or have implemented, plans to offer this service directly to taxpayers.
This report studies the economic implications of these government policies; the following summarizes the mains conclusions of this research:
Over the past 15 years, state and federal governments have broadly expanded the use of information technology in the handling of tax returns and refunds. For instance, in 2012, more than 80 percent of federal returns were filed electronically. Many tax refunds are also now paid via direct deposit; in 2012, 75 percent of refunds were issued via direct deposit.2
But for taxpayers who do not have bank accounts to receive such deposits, or who do not wish to provide account information on their tax returns, tax refunds are paid via paper checks. The latest Federal Deposit Insurance Corporation survey estimates that 8 percent of U.S. households (including 17 million adults) are “unbanked,” and another 20 percent are “under-banked.” Tax refunds are a particularly important source of funds for these families, who are far more likely to be poor; for instance, a single mother with two children might receive as much as 40 percent of annual income from the earned income tax credit, which is paid out via the tax refund.
The federal government and some state governments have developed successful electronic payment systems for many other programs, including programs that target low-income families who are often unbanked. For instance, the Supplemental Nutrition Assistance Program (formerly known as food stamps), the Temporary Assistance for Needy Families, and Social Security's Supplemental Security Income program all pay out regular monthly benefits via prepaid debit cards.
The private sector was the first to develop programs by which taxpayers could receive tax refunds via prepaid debit card. These financial products, known as “refund transfers” (RTs) or “refund anticipation checks,” set up a temporary bank account for the receipt of the tax refund. In the most recent tax season, all of the major tax preparation companies have partnered with banks to offer this option to their customers. For instance, H&R Block offers the MasterCard Emerald Card that is issued by H&R Block Bank; Jackson Hewitt offers a Visa Smartcard that is issued by Metabank. These products have proved quite popular; in tax season 2012, nearly 20 million taxpayers used RTs to receive their refunds, accounting for 18 percent of all refunds issued.
Against this background of expanding electronic payments in tax refunds and other federal programs, both the federal government and a number of state governments have introduced programs to offer tax refunds on debit cards. This report details these programs and analyzes their economic implications for taxpayers.
The federal government was the first to introduce the option of tax refunds on debit cards through the MyAccountCard program.3
At the beginning of the 2011 tax season, the government sent letters to 800,000 taxpayers, who were selected as low-income households that were likely to be unbanked. This letter informed them of the option to enroll, which they needed to do before filing their tax returns. Upon enrolling, the taxpayer would receive an account number to include on his or her return and would be mailed a debit card; any tax refunds owed would be deposited into that account.
The prepaid debit cards had a number of costs associated with their use. First, half of the recipients were charged a $4.95 monthly fee for account maintenance; the other half of card recipients faced no monthly fee.4 In addition, the card user faces a range of fees that are standard in such accounts for out-of-network ATM use ($2.50), withdrawals at bank-teller windows ($2.50), card replacement ($4.95), and adding money to the account by means other than direct deposit ($2.50). In addition, the federal government paid $12.95 per account opened to Greendot-Bonneville Bank to defray the costs of account administration.
The MyAccountCard program had very low take-up; just 0.3 percent of those who received an offer applied for the card, and just 33 percent of this group (only 638 people) actually used the debit cards to receive their refunds. Although this low take-up is consistent with other direct-mail financial products such as credit cards, it stands in stark contrast to the popularity of RTs. The program was discontinued in the 2012 tax season, but federal officials have suggested that the program may return.
A number of states have also instituted programs that offer tax refunds on prepaid debit cards. In tax season 2012, six states had such a program: Connecticut, Georgia, Louisiana, New York, Oklahoma, and South Carolina. Each of these states continued the program for tax season 2013, and Missouri, Virginia, and Alabama inaugurated debit card programs. In addition, Ohio is considering the use of debit cards for tax refunds in tax season 2014.
In all of the original six states, debit cards were subject to an array of fees for usage. None of the states pays an account setup fee, but two of the states require taxpayers to pay account closure fees in certain conditions. The other fees vary from state to state, although all include fees for out-of-network ATM use. Other fees include account inactivity fees, balance-check fees, and even customer service fees.
One important difference between the state programs is the action required by the taxpayer in order to receive a refund on a debit card. In Oklahoma, Connecticut, and Virginia, paper checks are no longer an option for tax refunds.5 In Georgia, Louisiana, New York, South Carolina, Alabama, and Missouri, taxpayers are offered the option of either paper checks or debit cards (alongside direct deposit).
The use of debit cards differed wildly between those states that continued to offer paper checks as an option and those that did not. For instance, among the roughly 1 million taxpayers in Georgia who did not receive their tax refunds through direct deposit in 2012, just 1.5 percent asked the government to send a debit card. Similarly, South Carolina issued just 5.2 percent of all refunds on government-program debit cards, in comparison with 32 percent that remained as paper checks. In comparison, Oklahoma replaced all of the nearly 500,000 paper checks issued in 2011 with debit cards, accounting for 43 percent of all refunds.
New York presents somewhat of a hybrid case. In New York, taxpayers who do not check one of the boxes are assumed to request a debit card, as compared to the other states still offering paper checks in which debit cards simply appeared as a new option beside paper checks and direct deposit. Yet despite this “nudge” to encourage taxpayers to participate in the debit card program, less than 1 percent of refunds were paid via prepaid debit card.
Combining data across all three states in 2012 that offered both debit cards and paper checks, just 1.4 percent of refunds (155,716) were issued via government-sponsored debit card compared with 34.4 percent via paper check (3,984,790).
There are a number of costs and benefits of debit cards as a means for providing tax refunds for individuals. These benefits include various aspects of both cost and convenience. The findings of this section apply to debit cards offered both through government programs and private companies, and so I draw on literature on both.
One primary benefit is that tax refunds that are electronically issued arrive roughly one to two weeks sooner than those sent by paper check if the taxpayer files the return electronically, but it takes up to two months longer with a paper return. Electronic payment is also a more reliable means for the taxpayer to receive a refund, as it requires much more work to reissue a lost check than to reissue a new debit card.
These advantages are especially important to those low-income taxpayers for whom large tax refunds are used to pay down high-interest debts, such as payday loans or credit cards on which a consumer is behind in payments. For those taxpayers who are unbanked or under-banked, electronic receipt of the tax refund can also save them check-cashing fees, which can be as high as 2 percent of the value of the check cashed.
Although it is nearly costless to transfer tax refunds electronically into an existing account, the primary cost of debit card programs comes when setting up and then managing the account. Banks earn returns from access to deposited funds in proportion to the size of the funds; in contrast, the costs for banks are largely fixed, no matter the size of the account. Therefore, small bank accounts of the size required for most tax refunds are quite costly to administer.
In order to cover the cost of account administration, debit card providers (for both private and governmental programs) charge a number of fees. Some of these fees are unavoidable, such as account setup fees. Other fees are designed so that it is possible to avoid them, but with effort. For instance, there is usually an ATM-service fee for ATMs that are out-of-network. If consumers are not familiar with the network structure, or prefer using the teller window, then costs can mount. As a result, debit cards may be inappropriate for certain financially inexperienced consumers.
Data from the federal MyAccountCard pilot program give some idea of the average variable fees incurred by users who chose to participate. The Urban Institute analysis of that program reports that individuals paid an average of $5.11 per month in variable fees during the time when the card was in use. Individuals who used the debit card for more months incurred fewer monthly fees; for instance, among those taxpayers who used their card for at least six months (through July 2011), monthly charges were $3.16 on average and thus represent an average of $19 paid in fees. Most of these fees resulted from ATM usage fees.
In all likelihood, this average level of fees masks substantial differences across individuals; unfortunately, the report does not provide the numbers to investigate these differences.6 In addition, there is no data available for fees accrued through use of debit cards issued by private companies for comparison.
Therefore, use of the debit card to receive the tax refund is not costless to individuals. There are clearly some individuals who will benefit from receiving their tax rebate on a debit card—for instance, those individuals who would pay high check-cashing fees, as well as those individuals who will use the card to minimize fees—but there are others who may wish to receive a check.
The previous section detailed the costs and benefits of tax refunds via debit cards for individuals; this section now turns to the problem from the perspective of the government, and hence focuses again on government-sponsored debit card programs.
The primary focus of the debate has been the apparent increase in efficiency from the use of electronic payments. For instance, a number of studies have reported that it is far cheaper for the government to issue refunds electronically. One 2011 study estimated the total cost of mailing a paper check at $1.02, in comparison with only $0.10 per electronic payment.7 The New York Department of Taxation and Finance estimates that it costs $0.58 per check mailed, in comparison with $0.007 for an electronic payment. The federal government has estimated that it would save $125 million per year if all federal payments (including tax refunds, Social Security payments, and other benefit payments) were processed electronically.
However, these savings must be weighed against the additional costs incurred through the use of debit cards to remit tax refunds. One direct cost is the account setup fee, paid by the government to partnering banks. For instance, the federal government paid $12.95 for each account created to Greendot-Bonneville Bank in the pilot of the MyAccountCard program in 2011. This is similar to the $12.60 account setup fee paid in the Electronic Transfer Account (ETA) program used to disburse Social Security payments.
In order to compare the extra processing costs of paper checks to the account setup fees for prepaid debit cards, it is important to determine the number of times a given debit card account will be used for government payments. Given the number above, it appears that the government must conduct at least 12 transfers to a single account before it is cost-efficient to use debit cards.
The ETA program meets this hurdle easily. Social Security payments are monthly, so a participant must only use his or her card for a year before costs are recouped. In addition, Social Security payments are regular. This regularity suggests that users will be more likely to maintain the accounts in between payments without incurring account inactivity fees.
In contrast, tax refunds do not pass the cost-efficiency test. Most accounts are used for only a single tax refund, after which taxpayers deactivate the account or may face inactivity fees. This requires a new account setup fee each year, which clearly outweighs the extra $1 in processing costs.
The federal MyAccountCard pilot program explicitly acknowledged this economic reality; for instance, the Urban Institute evaluation bluntly stated that “[t]he nature and frequency of tax payments require different thinking [than the ETA program] …. Simply converting a once-a-year check payment to a once-a-year plastic card would not provide savings for the government.”
In an attempt to boost the economic feasibility of debit cards, the MyAccountCard pilot encouraged users to maintain their debit card accounts so that they could be used again for future tax refunds, as well as to encourage entry into the financial system for unbanked households. However, the data suggest that after six months only 37 percent of taxpayers who received their refunds via debit card were still using the card, and only 16 percent were still depositing new money into the account at that point. The program also offered a linked saving account, but only 12.6 percent of active card users even activated the account and just 1.4 percent deposited money into that account.
In sum, the evidence suggests that it is quite difficult to generate recurrent use of these debit cards, and so it remains a highly inefficient option for tax refunds.
Many of the debit card programs established by state governments do not include any account setup fees paid by the government to partnering banks. As a result, governments may be able to reduce the government budget by eliminating the processing costs of paper checks. For instance, the commissioner of the Connecticut Department of Revenue reported that the state saved nearly $300,000 in tax season 2012 by issuing debit cards rather than paper checks for all refunds under $5,000; New York projected that it would save more than $500,000 by allowing debit cards to be optional for taxpayers.
These figures are highly misleading as to the actual costs and benefit of these programs. It is true that state governments can cut their own costs by switching to debit cards from paper checks, but this simply pushes costs off the government budget and onto taxpayers directly in the form of fees. The roughly $1 of government savings is very small in comparison with the average fees incurred by individuals who received their tax refunds via debit card. Recall that the taxpayers who chose to apply for the MyAccountCard paid an average of $19 in fees over the first six months of using the card. Users of state programs, especially those that mandate debit cards for those not receiving direct deposit, might be even less financially savvy than the average taxpayer who chose to enter the MyAccountCard program. Those taxpayers exceeded the $1 government savings in fees paid just in the first month.
The inefficiency of government-sponsored debit card programs most likely reflects the structure of incentives for state revenue office administrators. Especially in recent years, as tax receipts have fallen, these officials have faced pressure to cut costs. By converting tax refunds to debit cards, the revenue office shifts costs off-budget and onto taxpayers. In addition, government revenue offices face no direct pressure from taxpayers akin to the consumer pressure faced by private firms. While consumers can always choose a different private firm, taxpayers cannot avoid filing taxes with the state revenue office.
Debit cards reduce some costs and increase others, relative to paper checks, but neither method of delivery is clearly dominant. Some consumers may benefit substantially from the option of debit cards, while others will strongly prefer paper checks. As a result, the government should continue to allow taxpayers the option of receiving refunds via paper check.
In addition, the argument that debit cards save money for governments holds only because the government imposes far greater costs directly on taxpayers. While the $1 reduction in processing costs should not be ignored, it is an order of magnitude less than the other costs and benefits that debit cards may entail.
The previous section has argued that some but not all consumers may benefit from the receipt of tax refunds via debit cards, but that arguments for mandatory government use of debit cards (alongside direct deposit) based on reduction in processing costs are unlikely to hold. But if taxpayers are to have the option of receiving their refunds via debit card, is it more efficient for government or the private sector to provide this service?
Government versus private provision is a classic question in public economics. For most services, private companies will provide higher quality service at a lower price than government. The basic force driving this conclusion is that private companies face continual incentives from consumers to provide value; those companies that do not will lose customers and profit, and may eventually go out of business. This market feedback provides a direct incentive to companies to price goods efficiently. More importantly, it also creates an incentive to innovate in response to consumer demands.
In contrast, government faces no similar competitive pressures for two reasons. First, once government offers a service, it usually monopolizes the market and eliminates other competition. Unless private companies stay active in the market or threaten to re-enter, the government faces no market incentive to improve service quality or to innovate. Second, even in the presence of market competition, government employees do not benefit personally from any innovations they discover.
Despite the basic difference in incentives, there are conditions under which private provision of goods fails and government may be able to improve service. The literature has identified a number of classic categories of such situations; I now evaluate whether the use of debit cards for tax refunds qualifies as a special situation demanding of government intervention, or whether private companies will provide a superior product given cost.
When private companies compete in a healthy and well-informed market for consumers, governments can rarely provide higher quality service than private companies. However, if there is no competition in private markets, then governments may be able to improve upon quality and price. This may occur through a number of different channels.
If a private company exercises monopoly power over the market, then its prices may be too high, resulting in inefficiently low take-up from consumers. This does not appear to be the case in the market for providing tax refunds. A number of large professional tax preparation companies (e.g., Jackson Hewitt and H&R Block) compete along with their banking partners in the debit-card-refund with tax-preparation market. These companies also compete with Intuit (the maker of TurboTax, and in partnership with Santa Barbara Tax Products Group, a financial services contractor) to provide tax refunds for tax returns prepared online or with computer software.
Another force that may justify government provision of services is a clear efficiency advantage due to economies of scale. For instance, government health services such as Medicare pay a small fraction of the administrative costs per patient compared with private insurers. But while the federal government could clearly generate a larger customer base, and therefore greater bargaining power, than private companies, it is less clear that individual states can do the same. For instance, H&R Block issues far more tax refunds via debit cards in a given tax season than the state of Connecticut.
In addition, a straight comparison of fees may be misleading on the total costs of each approach to taxpayers, however, since government programs cater to a very different population than private companies. In particular, taxpayers must specifically choose to purchase an RT from a private company, while debit cards issued by the government are spread to a much wider population.
To see why this can matter for the efficiency of government provision, suppose that taxpayers differ in their financial savvy: some are easily able to avoid fees, while others are not. Because many government programs encourage all taxpayers to receive refunds by debit card, they include many individuals who will end up paying a significant amount in fees.
Many of the most vocal complaints about the debit card programs run by states in tax season 2012 centered on the use of debit cards by those with little or no experience in the financial system. Elderly taxpayers, who have never used a debit card or ATM, were forced to locate ATMs and distinguish between those which were in- and out-of-network. (In an ironic twist, Oklahoma's debit card program charged taxpayers who called the consumer-help line in order to figure out how to use the card without incurring fees). In South Carolina, some taxpayers did not realize that they had “chosen” to receive their tax refund via debit card.
An efficient solution to this problem is for governments to allow taxpayers to choose between receiving their refunds via paper check or debit card (as did Georgia, New York, and South Carolina), and to make taxpayers highly aware of this choice. In practice, however, this limits the ability of the government to provide cost reductions. Many fewer taxpayers will choose debit cards under these conditions, reducing the ability of the government to achieve a more efficient scale or use increased bargaining power to achieve lower prices. For instance, just 14,664 taxpayers in the entire state of Georgia chose to receive their refund through the government debit card program in 2012, eliminating any ability to achieve a more efficient scale than the private sector.
In addition, the set of taxpayers who choose debit cards will be savvier than average, reducing the usage fees collected. Since the partnering banks must collect enough revenue to offset costs, they will respond by raising other fees or insisting on an account setup fee in future years.
These problems may also magnify over the long run as more consumers become aware of the fees associated with government debit card programs. Partner banks may also attempt to use low initial prices in order to win government contracts, only later increasing fees in order to cover costs. Recall that consumers may not fully understand all of the costs associated with a product during their first opportunity to purchase the product but over time become better attuned to the fees that they will pay. As taxpayers become more familiar with government-sponsored debit cards, the average fees paid by each participating taxpayer will fall, driving partnering banks to increase either variable or account setup fees. In contrast, private companies already serve mostly repeat customers, and so pricing is unlikely to increase significantly because of changing mix of consumers.
John N. Friedman is an assistant professor of public policy at the Harvard Kennedy School and a faculty research fellow at the National Bureau of Economic Research.
This research was supported by funding from the American Coalition for Taxpayer Rights.
©2014 The Bureau of National Affairs, Inc. All rights reserved. Bloomberg Law Reports ® is a registered trademark and service mark of The Bureau of National Affairs, Inc.
This document and any discussions set forth herein are for informational purposes only, and should not be construed as legal advice, which has to be addressed to particular facts and circumstances involved in any given situation. Review or use of the document and any discussions does not create an attorney-client relationship with the author or publisher. To the extent that this document may contain suggested provisions, they will require modification to suit a particular transaction, jurisdiction or situation. Please consult with an attorney with the appropriate level of experience if you have any questions. Any tax information contained in the document or discussions is not intended to be used, and cannot be used, for purposes of avoiding penalties imposed under the United States Internal Revenue Code. Any opinions expressed are those of the author. The Bureau of National Affairs, Inc. and its affiliated entities do not take responsibility for the content in this document or discussions and do not make any representation or warranty as to their completeness or accuracy.
To view additional stories from Bloomberg Law® request a demo now