IRS Should Improve Data Related to Tax-Time Products, GAO Says

TAX REFUND PRODUCTS

Product Mix Has Evolved and IRS Should Improve Data Quality

Report to Congressional Requesters

April 2019

What GAO Found

Trends in the market for tax-time financial products since 2012 include

• the decline of refund anticipation loans (short-term loans subject to finance charges and fees),

• the rise in use of refund transfers (temporary bank accounts in which to receive funds), and

• the introduction of refund advances (loans with no fees or finance charges).

More recent product developments include increased online access to products for self-filers, higher refund advance amounts, the introduction of new products, and for tax year 2019, the reintroduction of fee-based loans.

However, GAO identified some limitations in Internal Revenue Service (IRS) data on product use, including over- or under-counting of certain types of products. IRS has not communicated these data issues to users and has not updated guidance to tax preparers on how to report new product use. As a result, data users (including federal agencies and policymakers) have inaccurate information to inform their findings and decision-making.

Lower-income and some minority taxpayers were more likely to use tax-time financial products, according to GAO analysis of 2017 data from IRS, the Bureau of the Census, and the Federal Deposit Insurance Corporation. Specifically, taxpayers who made less than $40,000 were significantly more likely to use the products than those who made more. African-American households were 36 percent more likely to use the products than white households. Product users tend to have immediate cash needs, according to studies GAO reviewed. For these users, tax-time financial products generally provide easier access to cash and more cash at a lower cost than alternatives such as payday, pawnshop, or car title loans.

GAO's undercover visits with nine tax preparers, its review of selected provider websites, and review of documents obtained from selected banks and tax preparers found disclosures generally followed requirements for disclosing fees. However, disclosure practices by some paid tax preparers may pose challenges for consumers. For example:

• Preparers in GAO's review generally indicated that they present taxpayers with almost all of the documents with fee information after their tax returns have been prepared and the preparers determined the taxpayers qualified for a tax-time financial product. The timing of these disclosures would pose a challenge for taxpayers looking to compare prices for different providers.

• During six of nine undercover visits, GAO investigators explicitly requested literature on product fees but were not provided such information.

• Refund transfer fee information on websites GAO reviewed sometimes was presented only after the tax preparation process started, was in small print, or could be found only after navigating several pages. As a result, taxpayers may face challenges comparing prices.

---

Contents

Letter

Appendix I Objectives, Scope, and Methodology

Appendix II Analysis of Characteristics Associated with Tax-Time Financial Product Use

Appendix III Disclosure of Product and Related Fees and Terms

Appendix IV Comments from the Internal Revenue Service

Appendix V GAO Contact and Staff Acknowledgments

Tables

Figures

Abbreviations

ACTC	Additional Child Tax Credit
CFPB	Consumer Financial Protection Bureau
EITC	Earned Income Tax Credit
ERO	Electronic return originators
FDIC	Federal Deposit Insurance Corporation
Federal Reserve	Board of Governors of the Federal Reserve System
FTC	Federal Trade Commission
IRS	Internal Revenue Service
OCC	Office of the Comptroller of the Currency
Path Act	Protecting Americans from Tax Hikes Act of 2015
SSP	settlement services provider
TILA	Truth in Lending Act

This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

---

April 5, 2019

The Honorable Elizabeth Warren
Ranking Member
Subcommittee on Financial Institutions and Consumer Protection
Committee on Banking, Housing, and Urban Affairs
United States Senate

The Honorable Tammy Duckworth
United States Senate

More than 20 million American taxpayers spent at least an estimated half a billion dollars in 2017 on financial products that are based on their anticipated tax refund, according to the National Consumer Law Center. Tax-time financial products, typically offered by banks and made available by providers of tax preparation services, include refund advances and refund anticipation loans (credit products) and refund transfers (deposit product). In fiscal year 2017, the Internal Revenue Service (IRS) processed more than 150 million individual federal income tax returns, and issued almost 120 million refunds totaling almost $383 billion, according to IRS.

You asked us to review trends in the market for tax-time financial products and the transparency of fees charged for these products. This report (1) describes trends in the market for tax-time financial products and product fees and examines the reliability of IRS data on these trends, (2) describes characteristics of those who use tax-time financial products and factors that influence their decision to obtain the products, and (3) describes regulatory oversight of industry participants and the disclosure of information on product fees and terms.

To address these objectives, we reviewed relevant federal laws, regulations, and guidance documents from the relevant financial regulators — Consumer Financial Protection Bureau (CFPB), Federal Deposit Insurance Corporation (FDIC), Board of Governors of the Federal Reserve System (Federal Reserve), and the Office of the Comptroller of the Currency (OCC) — the Federal Trade Commission (FTC), and IRS. We interviewed officials from the financial regulators, FTC, and IRS. We also interviewed representatives of various industry participants: five tax preparation providers selected because they are national chains, five banks and settlement service providers selected because they work with the major tax preparation providers, and four consumer advocacy groups and two academic researchers selected to provide a range of perspectives.

To examine trends in the use of tax-time financial products and fees, we analyzed available IRS data from 2008 to 2018 compiled from filed tax returns to determine the types and use of these products. We determined these data have some limitations, as discussed later in the report, but were adequate to suggest general trends when supplemented with other information. To supplement these data, we conducted a literature search and reviewed the websites, promotional materials, and other industry literature and public filings of four providers of online tax preparation services, three tax preparers with physical locations, and four banks to help identify trends in product offerings. The tax preparation firms were selected because they are national tax preparation chains, and the four banks because they partnered with the national tax preparation chains. The information collected from providers is not generalizeable to the population of tax preparers and banks offering these products. To examine trends in product fees, because of limited publicly available industry data we collected fee-related information on product fees, ancillary product fees, tax preparation fees, and aggregate fee data. We collected this information from selected preparers' and banks' websites, advertising materials, and public filings. The fee information is not generalizeable to the population of product and related fees.

To identify characteristics of product users, we used a sample of data on demographic and economic variables from the Bureau of the Census and FDIC from 2011, 2013, 2015, and 2017 to conduct a multivariate regression analysis to determine the relationship between individual characteristics and the decision to obtain a product. We statistically controlled for various income, education, tax-filing, and demographic factors. We used a sample of data from IRS from the 2014, 2015, and 2016 tax years to identify other taxpayer characteristics associated with product use. We determined these data to be sufficiently reliable for the purposes of our analysis by reviewing documentation on and conducting testing of the data for errors. We supplemented this information with a review of literature from government and industry reports on the financial needs of taxpayers, particularly those who obtain these products.

To describe the regulatory oversight of industry participants and the disclosure of information to consumers on tax-time financial products, we reviewed relevant laws and regulations. We reviewed reports and guidance documents from IRS, CFPB, FDIC, Federal Reserve, OCC, and FTC on disclosure of financial product fees and terms. To identify existing issues, we interviewed representatives of industry participants and four consumer advocacy groups selected to provide a range of perspectives. To review how product terms and fees are disclosed, in February 2018 GAO undercover investigators visited nine randomly selected tax preparers in Washington, D.C., Maryland, and Virginia to inquire about tax-time financial products. We selected locations based on product use and proximity to lower-income households in each location and to ensure a mixture of state laws governing products and service providers. The undercover visits provide illustrative information that is not generalizeable to the disclosure practices of all tax preparers. We also conducted a content analysis of websites of eight tax preparers and five bank providers that offer the products and reviewed consumer-facing disclosures and product agreements from these firms. We selected the tax preparation firms because they are national tax preparation chains, and the five banks because they partnered with these firms. The results of the website content analysis are not generalizeable to the content of all tax preparation firms' websites. Appendix I provides more detail on our scope and methodology.

We conducted this performance audit from July 2017 to April 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. We conducted our related investigative work in accordance with standards prescribed by the Council of the Inspectors General on Integrity and Efficiency.

Background

Tax-Time Financial Products

Table 1 provides an overview of tax-time financial products based on information gathered during our review.

Table 1: Types and Features of Tax-Time Financial Products

Product type	Refund advances	Refund anticipation loans	Refund transfers
Availability	Begins when the Internal Revenue Service (IRS) accepts electronic returns, generally in January, and activity typically ends on February 28a	Begins when IRS accepts electronic returns, generally in January, and typically ends in March.	Made available when tax return is filed, generally between January and October
Product terms	Short-term loans Several loan amounts available from bank provider, taxpayer must qualify (approval and loan amounts based on expected refund) Secured by expected refund (refund is collateral) No fees	Short-term, fee-based loans Loan amount is up to the refund amount minus tax preparation fees, finance charges, and additional loan and preparation fees Secured by expected refund (refund is collateral) Taxpayers responsible for a one-time fee and possibly other charges	Fee-based temporary bank account set up to receive tax refunds by direct deposit Taxpayers responsible for one-time flat fee
Method of payment	Taxpayer pays no fees; tax preparer generally pays fees to the bank provider	Fees deducted directly from the loan disbursement	Fee deducted directly from refund proceeds
Timing for receiving funds	Funds disbursed within 24 hours	1-2 days upon loan approval	Tax refunds deposited to temporary bank account within typical IRS time frames (generally within 21 days from filing) Taxpayer can receive remaining funds immediately
Source: GAO analysis of product provider promotional materials. | GAO-19-269 aRefund advances are generally offered only during the first two months of the tax season. According to two national tax preparation chains, this is because people with cash needs typically file early in the season.

Participants in the Tax-Time Financial Products Industry

The tax-time financial products industry consists of four main groups of participants: banks, paid providers of tax preparation services, settlement service providers, and software developers.

• Providers of tax preparation services include paid tax return preparers or electronic return originators (ERO).¹ Not all tax preparers are EROs, but because IRS generally requires returns to be filed electronically for tax preparers filing more than 10 returns, tax preparers generally work with or for an ERO that also may be a tax preparer. Paid preparers and EROs offer their services in-person, on the Internet, or through software sold to taxpayers. They generally offer different refund disbursement options to taxpayers and may partner with banks to offer tax-time financial products.

• Software developers provide software needed to file tax returns electronically and offer tax-time financial products through their software to taxpayers. The largest tax preparation companies have their own software that allows them to prepare returns as well as offer tax-time financial products. Applications for the products generally can be completed through the same software used to file the return.

• Banks provide tax-time financial products. They also may approve and process product applications and perform settlement services (discussed below).

• Settlement service providers serve as intermediaries in transactions to deliver tax-time products. They work with banks to accept and process applications for tax products; allocate payments due to paid preparers, other providers, banks, and taxpayers; and provide distribution instructions to banks. Some banks have affiliates that perform settlement services, and some banks perform these functions themselves.

Figure 1 illustrates the roles of these groups, using the example of a refund transfer transaction.

Regulators

Federal Banking Regulators

The purpose of federal banking supervision is to help ensure that banks throughout the financial system operate in a safe and sound manner and comply with banking laws and regulations in the provision of financial services. At the federal level, banks are supervised by one of the following three prudential regulators and CFPB:

• The Federal Reserve supervises state-chartered banks that opt to be members of the Federal Reserve System, bank holding companies and savings and loan holding companies (and the nondepository institution subsidiaries of those organizations), and nonbank financial companies designated for Federal Reserve supervision by the Financial Stability Oversight Council.

• FDIC supervises all FDIC-insured state-chartered banks that are not members of the Federal Reserve System as well as state savings associations and insures the deposits of all banks and thrifts approved for federal deposit insurance.

• OCC supervises federally chartered national banks, federal savings associations (federal thrifts), and federally chartered branches and agencies of foreign banks.

• CFPB has rulemaking authority to implement provisions of federal consumer financial law and enforces various federal laws and regulations governing consumer financial protection. CFPB also examines banks with more than $10 billion in assets and their affiliates and certain nonbanks for compliance with federal consumer financial laws, accepts consumer complaints on topics such as debt collection and other consumer financial products or services, and educates consumers about their rights under federal consumer financial laws.

FDIC, the Federal Reserve, and OCC are required to conduct a full-scope, on-site risk-management examination of each of their supervised banks at least once during each 12-month period. The regulators may extend the examination interval to 18 months, generally for banks and thrifts that have less than $3 billion in total assets and that meet certain conditions (for example, if they have satisfactory ratings, are well capitalized, and are not subject to a formal enforcement action).²

The prudential regulators generally conduct consumer compliance examinations every 12–36 months and Community Reinvestment Act examinations every 12–72 months. The specific timing depends on a bank's size and its previous consumer compliance and Community Reinvestment Act rating. But the Dodd-Frank Wall Street Reform and Consumer Protection Act transferred consumer protection oversight and other authorities over certain consumer financial protection laws from multiple federal regulators to CFPB. Additionally, for the transferred laws such as Truth in Lending Act (TILA) and Equal Credit Opportunity Act, CFPB has examination and primary enforcement authority for banks with assets of more than $10 billion and any affiliates of such institutions.³

The three prudential regulators also are responsible for supervising for compliance with federal consumer financial laws for insured depository institutions with total assets of $10 billion or less. For example, they examine depository institutions for compliance with consumer financial laws including the Fair Housing Act, the Servicemembers Civil Relief Act, and Section 5 of the Federal Trade Commission Act.

FTC

FTC can enforce Section 5 of the Federal Trade Commission Act, which prohibits unfair or deceptive acts or practices affecting commerce, and TILA, which seeks to promote the informed use of consumer credit. TILA requires disclosures about the terms and cost of credit and standardizes the manner in which costs associated with borrowing are calculated and disclosed.

FTC can enforce a number of additional statutes against certain entities; they include portions of the Gramm-Leach-Bliley Act, which requires financial institutions, including those providing tax-time financial products, to protect consumer data; the Telemarketing and Consumer Fraud and Abuse Prevention Act, which prohibits telemarketers from making misrepresentations in the sale of goods or services, which could include tax-time financial products; and the Military Lending Act, which provides important protections for servicemembers and their dependents seeking and obtaining certain types of consumer credit, including refund anticipation loans.

IRS

The Office of Professional Responsibility within IRS is responsible for ensuring all tax practitioners (defined as certified public accountants, attorneys, enrolled agents, enrolled actuaries, appraisers, and enrolled retirement plan agents) and other individuals authorized to practice before IRS adhere to regulations relating to Circular 230, which governs practice before IRS.⁴

According to IRS, IRS is neither involved in offering, nor responsible for, tax-time financial products. Nonetheless, IRS stated that it addresses these types of products on its website because it is important for taxpayers to understand the terms of the loan products, which constitute an agreement between them and the third-party lender. Although IRS is not statutorily required to collect data on tax-time products, according to IRS officials, the agency retains information on use of the products. Specifically, IRS compiles information from tax returns that indicates whether the taxpayer also applied for a financial product. IRS also issues guidance to EROs on reporting these data through its Handbook for Authorized IRS e-File Providers of Individual Income Tax Returns (Pub. 1345). IRS makes the usage data publicly available on its website, and provides it on a biweekly basis to industry participants that are members of an IRS working group on security issues. In addition to researchers and consumer advocacy groups, federal entities also use these data, including the National Taxpayer Advocate, who leads IRS's Taxpayer Advocate Service — an independent office in IRS whose objectives include mitigating systemic problems that affect large groups of taxpayers. As industry data on product use are generally limited, agencies and researchers rely on IRS for this information.

Tax Credits and Protecting Americans from Tax Hikes Act of 2015

Refundable tax credits include the Earned Income Tax Credit (EITC) and the Additional Child Tax Credit (ACTC). The credits are termed refundable because, in addition to offsetting tax liability, any excess credit over the tax liability is refunded to the taxpayer.⁵ EITC provides tax benefits to eligible workers earning relatively low wages. For tax year 2018, the maximum EITC amount available was $6,431 for taxpayers filing jointly with three or more qualifying children, and $519 for individuals without children. In 2017, EITC provided more than $65 billion to about 27 million taxpayers. ACTC is the refundable portion of the Child Tax Credit and provides tax relief to low-income families with children.

The Protecting Americans from Tax Hikes Act of 2015 (PATH Act) made several changes to the tax law.⁶ One of its provisions stipulates that funds owed taxpayers claiming EITC or ACTC refunds for a tax year cannot be released before February 15 to allow IRS time to review these returns for potential fraudulent activity. This change became effective on January 1, 2017. For the 2018 tax filing season (January through April 2018), refunds for taxpayers who claimed these tax credits were not available in bank accounts or prepaid cards until February 27, 2018.

IRS Data on Use of Tax-Time Financial Products Have Some Limitations, but When Combined with Other Available Data Suggest Product Offerings Have Evolved

IRS data on tax-time financial products for 2016–2018 do not accurately reflect product use and IRS has not updated reporting guidance to tax preparers. IRS data for 2008–2016 and information from industry participants and a consumer advocacy group's reports suggest that trends in the market for tax-time financial products include the decline of refund anticipation loans and that refund transfers became the most used product. Industry data also indicate that product fees for refund transfers increased in 2018; multiple other fees can be associated with tax-time products. New tax-time products and product features continue to be introduced.

IRS Data for 2016–2018 Do Not Accurately Reflect Product Use and IRS Has Not Updated Reporting Guidance to Tax Preparers

Data collected by IRS are the primary source of information on the use of tax-time financial products and are used by federal entities, policymakers, regulators, researchers, and consumer groups. However, we identified some limitations in the IRS data related to use of refund anticipation loans, refund advances, and refund transfers.

Two Products Not Differentiated in 2016 and 2017 IRS Data

First, 2016 and 2017 IRS data may have underreported use of refund advances and overreported refund anticipation loans. IRS officials told us that in 2016 and 2017, IRS made only three indicators available for tax preparers to report tax-time financial products: no bank product, refund anticipation loan, or refund anticipation check (that is, refund transfer). As a result, based on our analysis, it is possible that tax preparers reported refund advances as refund anticipation loans.

According to IRS officials, in 2016, IRS saw a large increase over the prior year in the number of refunds associated with tax-time financial product indicators. The agency performed an internal analysis on these refunds to identify the products being used in the market and found a direct relationship between this increase and new refund advance products. They determined that the original definitions for the indicators did not account for refund advance products.

In 2018, IRS expanded the indicator categories for tax-time financial products to more accurately reflect the products available in the market, including replacing the indicator for refund anticipation loans with two separate indicators for “pre-refund advance products with a fee (RAL)” and “pre-refund advance products with no fees” (most commonly known as refund advances) and adding an open text field to note products not otherwise covered.

Although IRS added another indicator category for refund advances, it has not attached explanatory material to the dataset or otherwise made it known to potential users of the dataset that the 2016 and 2017 data do not distinguish between refund anticipation loans and refund advances. Without explanatory material, users of the data, including the National Taxpayer Advocate and policymakers, could be unaware of the limitations.

Refund Transfers Also Not Always Reported

Second, since 2016 IRS may have misreported the number of refund transfers. A number of industry experts told us that almost all taxpayers who apply for a refund advance also apply for a refund transfer. However, tax preparers can select only one product indicator when reporting a customer's use of tax-time financial products, according to IRS officials. Consequently, tax preparers can report a refund advance or a refund transfer, but not both. Refund advances were introduced in 2016, so data for the 2016, 2017, and 2018 tax seasons currently are affected. We concluded that IRS data on refund transfer use prior to 2016 are not meaningfully affected because our research shows that from 2012 to 2016, as we discuss later, most product users were using one product, a refund transfer.

IRS officials told us that they had submitted a work request for filing season 2018 to allow tax preparers to select more than one type of product per tax return. The officials said that the request was denied due to competing information technology priorities at IRS.

IRS officials told us that tax preparers instead could use an open text field to indicate more than one product was used. IRS officials told us that the open field originally was created to allow for new product lines that do not fit existing descriptions. However, IRS has not provided additional guidance to tax preparers informing them of the potential alternate use for this field. Similarly, IRS has not informed tax preparers about the 2018 addition of a new indicator for refund advances and it has not updated its Handbook for Authorized IRS E-File Providers of Individual Income Tax Returns (Pub. 1345) on how to accurately code tax-time financial products. Without this additional guidance, tax preparers may continue to inaccurately report tax product information, making it challenging to identify trends and potential concerns with taxpayers' use of these products. Furthermore, IRS has not made this issue known to potential users of the dataset.

A strategic goal from IRS's Strategic Plan (for fiscal years 2018–2022) is to advance data and analytics.⁷ Related to data, GAO guidance on assessing the reliability of data states that reliable data can be characterized as being accurate, valid, and complete.⁸ In addition, federal internal control standards state that management should internally and externally communicate the necessary quality information to achieve objectives.⁹

As a result of the data conflation in 2016 and 2017 for refund advances and refund anticipation loans, ongoing issues with reporting use of refund transfers, and outdated guidance to tax preparers, users of the data (including the National Taxpayer Advocate, policymakers, regulators, researchers, and consumer groups) will have inaccurate information to inform their findings and decision-making.¹⁰

Tax-Time Financial Products Have Evolved Since 2012

Despite limitations with IRS data on product use by tax year, our analysis of multiyear trends from these data, supplemented with data collected by the National Consumer Law Center and from Securities and Exchange Commission filings, suggests that use of refund anticipation loans declined, the refund advance was introduced while refund transfers have become the most used tax-time product.

Refund Anticipation Loans

Applications for refund anticipation loans declined sharply from 2010 to 2012, according to IRS data and consumer groups reports. According to a 2010 study, the volume of refund anticipation loans peaked in 2002 with 12.7 million taxpayers.¹¹ Volume began to decline at a faster rate between 2010 and 2011. According to a report by the National Consumer Law Center and the Consumer Federation of America, banks stopped offering the products in 2012 after the loans came under the scrutiny of federal banking regulators.¹² IRS data continued to show use of refund anticipation loans after 2012 but with banks out of the market for refund anticipation loans, it is unclear what types of financial institutions were offering the loans. Consumer advocates with whom we spoke agree that nonbank lenders such as payday lenders likely offered the loans; however, we were not able to identify any. The consumer advocates, researchers, and industry participants with whom we spoke also were not able to provide us with any current information about these lenders.

The IRS Taxpayer Advocate Office, the Financial Crimes Enforcement Network, and consumer advocates have long raised concerns about refund anticipation loans.¹³ For example, in 2007 the National Taxpayer Advocate expressed concerns about how the loans were offered to consumers and whether consumers adequately understood the product. Consumer advocates questioned the high interest rates the loans could carry, how loan fees reduced EITC benefits taxpayers received, and the ramifications of borrower default. In a 2008 advance notice of proposed rulemaking, IRS and the Department of the Treasury also shared concerns that refund anticipation loans offered tax preparers an incentive to fraudulently inflate refund claims and to market the loans to taxpayers who might not understand the full cost of the product.¹⁴

Banking regulators raised concerns as well. OCC and FDIC noted consumer protection and safety and soundness risks to banks that offered refund anticipation loans. FDIC encouraged consumers to have tax refunds directly deposited into their own bank accounts and raised concerns about other options that claimed to speed up a refund for a sizable cost, according to FDIC officials. The Office of Thrift Supervision, which had supervisory authority over federal thrifts at the time, ordered a medium-sized thrift to cease making refund anticipation loans in 2010.¹⁵ In part due to concerns expressed by OCC, national banks stopped offering the loans by 2010 and FDIC-supervised banks stopped offering them by 2012.

An IRS decision also contributed to FDIC enforcement actions on refund anticipation loans. Before 2011, IRS used a tool called the debt indicator that acknowledged whether any of a taxpayer's refund could be used to pay certain outstanding debts.¹⁶ IRS provided the debt indicator to tax preparers at the time the taxpayer's return was filed electronically. Banks used the debt indicator in their underwriting tools to help determine a borrower's likelihood of loan repayment. FDIC determined that without the debt indicator, a bank would have to develop and adopt a more robust underwriting process to make these loans in a safe and sound manner. According to FDIC, IRS's elimination of the debt indicator created a safety and soundness concern because it removed a key data element used for determining a borrower's ability to repay. Losing this information increased the risk of loss for lenders and at that time helped inform FDIC’s consent orders with two banks under its supervision to stop offering refund anticipation loans. In 2011 (the first tax season without the debt indicator), the number of returns with a refund anticipation loan indicator reported by IRS decreased to 1.17 million from 6.9 million in the prior year.

IRS data continue to show use of refund anticipation loans after 2012, albeit at a much lower volume. For example, in 2016, IRS data show about 468,500 returns with a refund anticipation loan indicator and in 2017 the number appeared to spike to about 1.7 million.¹⁷ However, as discussed earlier, the data for these two years may be misleading because they likely conflate refund anticipation loans with refund advances. In 2018, IRS created a separate reporting category for refund advances and the 2018 data show about 356,000 returns with a refund anticipation loan indicator as of October 2018.

Refund Transfers

Use of refund transfers — which allow for direct deposit of refund checks through temporary accounts that banks open for taxpayers — far exceeded use of refund anticipation loans and refund advances since 2008, according to IRS data. The number of taxpayers who used a refund transfer more than doubled from 2008 through October 2018 to exceed 21 million. As banks stopped offering refund anticipation loans in 2012, refund transfers (also known as refund anticipation checks) began to increase. Unlike other tax-time financial products generally only available early in the tax season (which generally runs through mid-April), refund transfers are usually available after April.

However, IRS data on refund transfers since 2016 have limitations. Although a refund transfer is not required to get a refund advance, a number of industry experts told us that almost all taxpayers who apply for a refund advance also apply for a refund transfer. But because tax preparers could select only one product indicator when reporting use of tax-time financial products, they could report a refund advance or a refund transfer, but not both. As discussed previously, IRS made changes in 2018 to allow preparers to add information about other product use but has not issued explanatory material about the changes

Refund Advances

In 2016, a few banks began offering refund advances to taxpayers. Refund advances are no-fee, nonrecourse loans.¹⁸

It is difficult to determine usage trends for this product, although available data indicate an increase in use from 2016 to 2017.

• First, accurate IRS data on refund advances are not available for 2016 and 2017 because IRS did not provide an option for tax preparers to report refund advance products. As previously discussed, IRS added a separate reporting category for refund advances in 2018. As of October 17, 2018, IRS data show about 1.65 million returns with a refund advance indicator.

• Second, publicly available data from industry and other sources (consumer advocacy and research organizations) are limited. According to data reported by the National Consumer Law Center, major tax preparation companies facilitated the sale of about 365,000 refund advances in 2016. According to industry sources, use increased to about 1.63 million in 2017, when one of the largest tax preparation companies began offering refund advances. Industry data for 2018 were not yet publicly available at the time of this report.

• Third, taxpayers often obtain refund advances and refund transfers in tandem. But as discussed previously, IRS reporting indicators did not include an option for reporting use of multiple products until 2018.

Use of refund advances also may have increased in 2017 because tax preparers increased the size of the advances. One lender that offers refund advances to tax preparers told us that the driving factor in demand for refund advances was the available loan amount. The maximum advance amount that tax preparers offered taxpayers in 2016 was $750. In 2017, the maximum increased to $1,300.

Most industry participants and consumer groups told us that they believe that provisions of the PATH Act requiring IRS to delay issuance of EITC or ACTC returns and associated refunds until after February 15 led to an increase in demand for refund advances. They said that the delay puts pressure on taxpayers eligible for EITC or ACTC who depend on getting their refund early in the tax season (a refund advance can help mitigate the impact of this delay). Others stated that an increase in demand due to the PATH Act is possible, but the correlation between the two cannot be determined. One industry provider suggested that increased demand for refund advances also could be the result of marketing by tax preparation companies.

Limited Public Data Suggest Refund Transfer Fees Generally Increased in 2018

Our analysis of publicly available data about product fees for refund transfers showed that fees increased in 2018. In particular, our analysis of fee data collected by the National Consumer Law Center shows that in 2014–2017 refund transfer fees charged by paid tax preparers remained generally unchanged at between $32.95 and $34.95.¹⁹ According to fee information we were given during our undercover visits, paid tax preparers generally charged their customers $39.95 or $49.95 during the 2018 tax filing season for a refund transfer that sometimes included both federal and state tax refunds. In one case the fee was $65, which included a paper check disbursement. Also in 2018, we found that online providers of tax filing services and software charged online filers who prepared their own returns between $12 and $39.99 for a refund transfer.²⁰

According to our analysis, factors that can affect the fee a taxpayer pays for a refund transfer include the following:

• Filing method. Our review of providers’ websites shows that taxpayers who filed their own returns online using preparer software paid an average fee of $31.13 in 2018, which was lower than the $39.95 or $49.95 that paid preparers charged their customers.

• Disbursement method. The manner in which the taxpayer chooses to receive a tax refund may affect the fee. For example, our review of industry literature indicates that one bank set the fee at $29.95 if the refund was disbursed to a prepaid card offered by an affiliate vendor or at $39.95 if the refund was directly deposited or disbursed as a check. Another bank gave tax preparers the option to offer a free refund transfer for disbursement onto a prepaid card, $15 for a direct deposit, or $20 for a paper check.

• Incentives offered to tax preparers by banks. Incentives from banks for tax preparers can increase fees for taxpayers. Our review of banks’ promotional materials for tax preparers also indicates that some bank providers offer tax preparers different fee structures for a product — that is, the preparers can charge a higher fee to earn a rebate. For example, one bank offered a tax preparer the option to provide a refund transfer to clients for $39 (which includes an $8 incentive paid to the tax preparer) or for $29 (no incentive payment). On their websites, two banks marketed the no-incentive option to tax preparers as a way to be competitive (by offering low-cost options to their customers).

• Using a refund advance. According to a report by the National Consumer Law Center, one bank set a higher fee for a refund transfer if taxpayers also applied for a refund advance. When taxpayers used only a refund transfer, the fee was $29.95 for the federal refund and an additional $9.95 for the state refund, for a total of $39.90. If the taxpayer also applied for a refund advance (a no-fee product), the refund transfer fee was $44.95. Thus, taxpayers paid $5.05 more for a refund transfer if they also received a refund advance.

Our analysis found that, in addition to the product fee, taxpayers may be charged other fees when they use a refund transfer.

• State refund transfer. In some cases, the refund transfer fee covered the deposit of a federal and a state refund. In other cases, the fee only covered the federal refund. In these cases, if the taxpayer received a state refund, the tax preparer charged an additional fee of $10 or $12.

• Disbursement services. According to documentation we reviewed, a tax preparer may charge an additional fee of $25 if taxpayers choose to get their refund as a paper check or $7 for a cash transfer to a third party.

• Prepaid card use. The long-term use of prepaid cards used to disburse a refund may add to the overall cost of getting a tax product. We reviewed cardholder agreements and fee schedules for several prepaid cards commonly used to disburse funds from a tax refund and found they generally carry monthly fees of about $5. The issuer of the prepaid cards also may charge consumers a fee every time they access cash at automated teller machines, deposit more money onto the card, or do not use the card for a certain period of time.

• Software fees. Companies that design tax preparation software may charge a fee or fees associated with the tax product. Taxpayers may pay one or more of these fees when they use a refund transfer to receive their tax refund. The bank deducts these fees from the taxpayer’s refund after receiving funds from IRS or the state taxing authority. The fee categories are technology fee (up to $18 in our review), a transmission fee that may be a fixed amount (such as $2) or a variable amount, and a processing fee of $6.

Comparative Fee Scenarios

To determine how the fees associated with a refund transfer can affect the total tax preparation fees a provider may charge a taxpayer, we reviewed fee data we collected. We then identified the types and totals of fees generally associated with tax products and created four possible scenarios based on this analysis (see fig. 2). We designed two scenarios with online self-filers (taxpayer uses a refund transfer and taxpayer does not use a refund transfer) and two scenarios with paid preparers performing the filing (taxpayer uses a refund transfer and taxpayer does not use a refund transfer).

• Software fees. Companies that design tax preparation software may charge a fee or fees associated with the tax product. Taxpayers may pay one or more of these fees when they use a refund transfer to receive their tax refund. The bank deducts these fees from the taxpayer's refund after receiving funds from IRS or the state taxing authority. The fee categories are technology fee (up to $18 in our review), a transmission fee that may be a fixed amount (such as $2) or a variable amount, and a processing fee of $6.

Tax-Time Financial Products Have Continued to Evolve Since 2016

Recent and emerging developments in the market for tax-time financial products include higher loan amounts and new products, according to our analysis of selected tax preparers' websites and marketing materials, and information we were given during our undercover visits. For example, in 2018 refund advances became available to online filers. They previously were offered only to taxpayers who obtained paid tax preparation services in person (at a “storefront”).²¹

The maximum amount for a refund advance has continued to increase. In 2016, the maximum loan amount available to a taxpayer was $750. In 2018, the maximum loan amount available was $3,250 and for 2019, one preparer has offered an advance of up to $3,500. One industry participant told us that the industry in general is in a race to increase borrowing limits to remain competitive and attract more customers.

In 2018, banks offered a new product that combines the features of a refund anticipation loan and a refund advance. The product allows the taxpayer to apply for a refund advance (up to a fixed amount) with no fee or finance charges, the option to apply for an additional loan with a fee (similar to a refund anticipation loan), or a combination of the two products known as a hybrid. For 2018, two banks offered this additional loan (not to exceed $1,000) at an annual percentage rate of 29.9 percent. For 2019, one bank offered taxpayers the option of a no-fee advance of up to $1,000, or an interest-bearing loan of $2,000, $3,000, or $5,000 based on the expected refund. The interest-bearing loans would carry an annual percentage rate of 26.07 percent in addition to a fee of $30–$75, depending on the loan amount. Also for 2019, one national tax preparation company has offered the option of a no-fee advance of up to $3,500 or a fee-based advance of up to $7,000, which would carry an annual percentage rate of 35.9 percent.

In addition, demand for refund transfers has increased among online self-filers. As more people file their own tax returns by using web-based software, the number of refund transfers used by self-filers may continue to increase.²² Because few tax preparers offer refund advances to online self-filers, taxpayers are still more likely to get a refund advance from a paid tax preparer.²³

Finally, issues relating to the applicability of TILA disclosure requirements to refund transfers could affect the market for tax-time products. According to representatives of two consumer advocacy organizations, deferment of tax preparation fees until the refund is received constitutes an extension of credit; therefore, refund transfers should be treated as loan products. Tax preparers and a policy research and education organization with whom we met do not believe that refund transfer fees meet the definition of a loan.

Should regulators decide that a refund transfer constitutes an extension of credit, and would therefore be a credit transaction with a finance charge, refund transfers would become subject to provisions of TILA. These changes could affect taxpayers' access to this product as well as product pricing. According to Securities and Exchange Commission filings of some tax preparers, if refund transfers were successfully characterized as such, the additional requirements and costs could limit their ability to offer these products to clients.

Refund advances were promoted by providers as a fee-free, interest-free credit product, and thus TILA disclosure requirements are generally not considered applicable for them. However, new interest-bearing credit products announced for 2019 may be subject to consumer protection regulations.

Lower-Income and Some Minority Taxpayers Were More Likely to Use Tax-Time Financial Products for Various Reasons

Our Analysis Found That Lower-Income, African-American, and Single Taxpayers Were More Likely to Use Tax-Time Financial Products

Using FDIC data, we conducted a multivariate regression analysis to examine the relationship between economic and demographic variables and tax-time financial product use. This approach allowed us to test the significance of the relationships between each variable and the likelihood of using tax-time financial products, while controlling for other factors.²⁴

Income-Related Characteristics

Lower-income households were more likely to use tax-time financial products than higher-income households, particularly when they used paid tax preparers to file their taxes, according to our analysis of 2017 FDIC data.²⁵ More specifically, we estimated that households with incomes between $20,000 and $39,999 were more likely to use tax-time financial products to receive their tax refunds more quickly through paid tax preparers than households with incomes of $60,000 or more.²⁶ For example, we estimated that

• households with incomes between $20,000 and $29,999 were 34 percent more likely to use tax-time financial products than households with incomes of $60,000 or more;²⁷ and

• households with incomes between $30,000 and $39,999 were 61 percent more likely to use the products than households with income of $60,000 or more.²⁸

Moreover, our analysis of FDIC data suggests that households that received EITC were more likely to use tax-time financial products, compared to households that did not receive EITC.²⁹

Our results also suggest that wealth, as measured by homeownership, was associated with the household decision whether to use tax-time financial products. Homeowners were 34 percent less likely to use tax-time financial products than non-homeowners, controlling for other factors.

Other Characteristics, Including Race, Age, and Household Head

Households of some minority groups were more likely to use tax-time financial products when filing tax returns than white households. For example, using FDIC data, we estimated that African-American households were 36 percent more likely to use tax-time financial products than white households after controlling for other factors. Other research (a 2013 study) found that African Americans were more likely to use refund anticipation loans than white individuals.³⁰

According to our analysis of 2016 IRS data, which included information about tax-time financial product use and locality, use of tax-time financial products was more concentrated in some areas of the South and the West (see fig. 3).³¹

Our analysis of FDIC data further suggests that other characteristics associated with use of tax-time financial products include age and household type. For example, households headed by younger persons (15–39 years old) were more than twice as likely to use the products as households headed by older persons (60 or older), controlling for other factors.

Households headed by single adults with families were more likely to use tax-time financial products than households headed by married couples.³² For example, according to our analysis of FDIC data, we estimated that households headed by unmarried females with families were 76 percent more likely to use tax-time financial products than households headed by married couples, controlling for other factors. Using IRS data from 2016, we found that a higher proportion of product users filed as unmarried heads of household, compared to the general tax filing population. Among those who used tax-time financial products, about 39 percent filed as single, 22 percent filed as married, and 37 percent as unmarried heads of household.³³

Reasons for Using Refund Products Include Obtaining Cash Faster and Not Paying Tax Preparation Fees Up Front

Reasons to use tax-time financial products include more quickly obtaining cash from the expected tax refund, not having to pay tax preparation fees out of pocket, and obtaining cash more cheaply than with alternative short-term funding options, according to our review of federal and industry reports.³⁴

Quick Access

Taxpayers generally might have to wait weeks for refunds from IRS:

• Taxpayers who file paper returns can expect to receive their refund about 6–8 weeks after the date on which IRS receives their return, according to IRS guidance.

• Taxpayers who file electronically generally can expect to receive their refunds within 21 days, or faster if they opt to have refunds deposited directly into their bank accounts.

• As previously discussed, IRS must delay payments of refunds on which EITC, ACTC, or both are claimed until at least February 15 of each year. Effectively, the refunds might not be disbursed to bank accounts (or prepaid cards) of tax filers until the end of the month.

In contrast, users of tax-time products can obtain cash very quickly. For example, refund advance recipients generally receive loan funds within 24 hours of applying, and in some instances within the same hour they apply, according to selected tax preparer documents and websites that we reviewed. Refund transfer products also allow those who do not have the option of directly depositing refunds into a temporary account instead of waiting longer to receive a paper check. According to our analysis of IRS data from 2016, tax-time financial product users were more likely than other taxpayers to receive their tax refunds by direct deposit.

Taxpayers may use tax-time financial products because they need cash quickly. Studies we reviewed found that product recipients tend to have pressing financial obligations. One study's review of available literature from 2010 found that product recipients tend to live paycheck-to-paycheck or lack sufficient savings to cover prior, current, or future spending.³⁵ Another study published in 2010 found that recipients use the products to pay for pressing financial obligations, both expected and unexpected, and for their tax preparation. According to the study, many users of tax-time products become delinquent on rent, utilities, and other expenses during the winter with the expectation that they will be able to pay obligations after receiving tax refunds.³⁶ As one study found, the annual tax refund represents the largest single cash infusion received all year by about 40 percent of checking account holders.³⁷

Tax Preparation Fees Not Paid Out of Pocket

Lower-income taxpayers also use tax-time financial products to defer payment of fees related to tax return preparation, according to federal government and industry reports that we reviewed. Tax preparation fees vary greatly based on the tax forms used, including the EITC worksheet. One of the largest national tax preparation chains reported that its average tax preparation fee was between $205 and $240 in 2017.

Consumers may perceive any costs associated with tax-time financial products and tax return preparation as lower than they actually may be because the costs are not paid out of pocket. Fees for the products and tax return preparation are deducted from the refund before it reaches the consumer. In general, studies have found that the transparency of a payment method affected the payer's willingness to spend.³⁸ One consumer advocacy organization representative posited that paying for tax-time financial products and tax preparation from a refund makes consumers less sensitive to the real cost of tax-time products and preparation services.

Instead of using tax-time financial products to defer payment of tax preparation fees, lower-income taxpayers can access free filing services through several IRS programs (see sidebar). However, these options do not allow taxpayers to use tax-time financial products to access refunds faster.

IRS estimates that about 70 percent of taxpayers are eligible to access its free filing software, and we estimated about 3 percent of taxpayers use this service. According to IRS officials, while IRS does not have a marketing budget to promote the free file programs, the predominant reason so few taxpayers use them is because there are many free tax preparation options on the market, such as tax preparation software.

Higher Refunds and Tax Preparation Assistance

Taxpayers also may use paid tax preparers because they do not think they can fill out tax returns on their own, believe that preparers will help them receive higher refunds, or both, according to federal government and industry reports we reviewed. For taxpayers who did not use tax-time financial products, we did not find a clear association between paid tax preparation and higher average refunds. On the other hand, for taxpayers who used tax-time financial products, we found that average tax refunds were higher for taxpayers who filed through paid tax preparers than for taxpayers who self-filed online (see table 2). According to IRS data, nearly all taxpayers who used refund loan products filed their taxes through paid tax preparers, as refund advances were not available online until the 2018 tax filing season. There may be various reasons for the association between higher refunds, paid tax preparation, and product use. Those who use tax-time financial products tend to be eligible for tax credits such as EITC, which can increase the size of tax refunds. Fifty-four percent of EITC claimants used a paid preparer. However, a 2017 study found that the combination of paid tax preparation and tax-time financial product use was associated with relatively high incorrect tax payments (specifically, overpayments of EITC compared to online self-filing and product use or no product use).³⁹

Table 2: Average Refund Amounts, by Tax Filing Methods and Tax-Time Product Usage, 2014–2016

	2014		2015		2016
	Self-prepared online	Practitioner-prepared	Self-prepared online	Practitioner-prepared	Self-prepared online	Practitioner-prepared
No tax-time financial product	$2,103	$2,028	$2,029	$2,108	$2,548	$2,168
Tax-time financial product	$3,359	$3,954	$3,340	$4,044	$3,255	$4,064
Source: GAO analysis of IRS data. | GAO-19-269 Notes: The table presents average refund amounts by tax filing method and tax-time financial product use. Differences in average refund amounts across tax-time financial product use are statistically significant with the exception of self-prepared taxes online in 2016. Differences in average refund amounts across tax filing methods are statistically significant with the exception of tax refunds with no tax-time financial product in 2016. Statistical significance is measured at the 5 percent level, meaning the difference in estimates is significant at the 95 percent confidence level.

Furthermore, our analysis of IRS data found that taxpayers who used tax-time financial products received higher refunds on average than those who did not use tax-time financial products, regardless of tax filing method — although other factors might explain this association. For example, taxpayers who have high refunds have a greater incentive to use the products than taxpayers who have relatively small refunds or owe taxes.

Tax-Time Financial Products Cheaper Than Alternatives

For lower-income taxpayers, tax-time products generally provide more cash at a lower cost than other small-dollar loan alternatives such as payday loans, auto title loans, and pawnshop loans, according to our review of federal government and industry reports.⁴⁰ The amounts of alternative loan products are based on the value of the collateral the consumer provides. Average loan amounts are $150 for pawnshops, about $500 for payday loans, and under $1,000 for automobile title loans, according to industry statistics and CFPB and other studies. In contrast, refund advances were offered for up to $3,250 for the 2018 tax filing season.

Furthermore, the alternative products generally include fees, unlike refund advances. For example, fees for payday loans generally range from $10 to $30 per $100 borrowed. Automobile title lenders generally charge a fixed price per $100 borrowed, with a common fee limit of 25 percent of the loan per month. In contrast, refund advances are offered at no cost to the consumer.

Tax-time financial products also may be easier to access because, unlike alternative loans, they generally can be obtained without regard to credit history. However, tax-time financial products generally are only available during tax season.

Loans provided by nonfinancial companies (often called fintech firms) are another source of short-term financing. However, fintech firms generally provide much larger loan amounts than tax-time financial products, and include fees, unlike refund advances.

Providers We Reviewed Generally Disclosed Required Information but Some Disclosure Practices May Hinder Consumer Decision-Making

The federal banking regulators oversee banks that offer tax-time financial products and IRS sets standards of practice for certain service providers (including some tax preparers). While our nongeneralizeable review found that selected banks and tax preparers generally followed existing OCC and IRS disclosure requirements, some tax preparers' disclosure practices may present challenges for consumers trying to compare product options.

Industry Participants Are Subject to Varying Levels of Oversight

Banks and Settlement Service Providers

FDIC, the Federal Reserve, or OCC are responsible for the safety and soundness supervision of banks within their authority (which offer tax-time financial products) and may have supervisory authority over third-party service providers (which provide settlement services). We identified five banks that partnered with several national tax preparation chains in recent years to offer tax-time financial products (refund transfers and refund advances). Of the five banks, FDIC supervised one medium-sized and one small bank, OCC supervised two medium-sized banks, and Federal Reserve supervised one medium-sized bank.⁴¹

As previously discussed, FDIC, the Federal Reserve, and OCC are to conduct full-scope, on-site risk-management examinations of each of their supervised banks at least once in each 12–18 month period. FDIC officials told us that its regular safety and soundness examinations may include an examination of the bank's tax-time financial product offerings. OCC officials told us that they examine tax-time financial products in every annual examination of the banks they supervise that offer these products.

Because the five banks each has total assets of less than $10 billion, the three regulators also are responsible for enforcing compliance with federal consumer financial laws (such as TILA and the Electronic Fund Transfer Act) that govern disclosure requirements for certain tax-time financial products. Officials from the regulators told us that they received few complaints about tax-time financial products offered by their supervised banks. We discuss the disclosure requirements and compliance with the requirements in more detail later in this section.

The regulators' consumer compliance examiners also may review a bank's tax-time financial products — if, for example, a bank offers a new product or there are a number of consumer complaints about a current product. Examiners employ a risk-focused approach with a focus on consumer harm in selecting products to evaluate for compliance with applicable consumer laws and regulations. Furthermore, compliance examiners may decide, based on the potential for consumer harm and a bank's compliance management system, that there is enough residual risk to scope the product into the examination. FDIC officials said that a bank with a lot of activity in the market for tax-time financial products would have to assure examiners that it had performed appropriate due diligence.

Regulators also can take other oversight actions, ranging from enforcement to raising awareness among consumers. In 2015, CFPB took an enforcement action, along with the Navajo Nation, to ban an owner of four tax preparation franchises from the market and levy civil penalties for understating refund anticipation loan rates and deceiving customers about the status of their tax refunds.⁴² Our search of CFPB's complaint database did not identify any consumer complaints on tax-time financial products. CFPB published a blog post in February 2018 that describes the different tax-time financial product options and the process for obtaining them, and cautions consumers to consider all fees, charges, and timing associated with the products.⁴³

FTC staff we interviewed told us that supervision authority over many financial services providers has been given to CFPB, but that FTC still has the authority to enforce many financial statutes and rules, including rules administered by CFPB.⁴⁴ FTC brought an enforcement action in 2017 against an online tax preparation provider alleging that it failed to secure consumer accounts. FTC officials also told us that, while they received numerous complaints on tax-related issues, FTC's complaint database does not separately classify complaints based exclusively on tax-time financial products.⁴⁵

FTC also has issued guidance to educate consumers regarding tax-related scams and other consumer protection issues that arise during tax time, and to businesses, including tax professionals, to help them detect cyber threats. FTC also co-sponsors a series of educational events for consumers and businesses surrounding tax identity theft awareness week.

Software Developers

Software companies we interviewed stated that they are subject to IRS regulations relating to electronic filing of tax returns. Software developers provide tax software to tax preparers so that they may file tax returns electronically and assist taxpayers in obtaining tax-time financial products. One software company told us that this involves working with IRS to ensure that returns can be electronically submitted, IRS can receive data, and the software is in compliance with IRS's required data schemas.

Tax Return Preparers

IRS officials said that IRS does not monitor or have direct oversight authority over tax-time financial products, but requires some paid tax preparers to meet standards of practice or other requirements. The extent to which IRS has oversight over paid preparers depends partly on whether the preparer is a tax practitioner or unenrolled preparer.

Tax practitioners are subject to regulations (Circular 230) that establish standards of practice.⁴⁶ For example, practitioners must return tax records to clients, exercise due diligence in preparing tax returns, and submit records and requested information to IRS in a timely manner. IRS officials told us that they monitor the suitability of these practitioners and their adherence to the rules. Additionally, certain tax practitioners known as enrolled agents generally are required to pass a three-part examination and complete annual continuing education, while attorneys and certified public accountants are licensed by states but are still subject to Circular 230 standards of practice if they represent taxpayers before IRS.

Alternatively, unenrolled preparers — the remainder of the paid preparer population and the majority of paid preparers — generally are not subject to these requirements. In 2011, IRS issued final regulations to establish a new class of registered tax return preparers to support tax professionals, increase confidence in the tax system, and increase taxpayer compliance. However, the U.S. District Court for the District of Columbia ruled in 2013 and the U.S. Court of Appeals for the District of Columbia Circuit affirmed in 2014 that IRS lacked sufficient authority to regulate all tax preparers.⁴⁷ IRS officials also told us that all authorized IRS e-file providers have to follow certain requirements to be able to file tax returns electronically.

Banks and Tax Preparers in Our Review Generally Followed Guidance for Disclosing Product Fees, but All Related Fees Were Not Always Disclosed Clearly or Early in Process

We found selected authorized IRS e-file providers generally followed the requirements established by IRS on the disclosure of product fees, and banks generally followed the disclosure guidance relating to tax-time financial products issued by OCC.⁴⁸ (We conducted nongeneralizeable reviews of website content, industry documents, and disclosures made during our undercover visits.) Two of the five banks we reviewed are regulated by OCC. One of the two FDIC-supervised bank and the Federal Reserve-supervised bank told us that they voluntarily follow OCC guidance.

More specifically, IRS established the following disclosure requirements for authorized IRS e-file providers, generally known as EROs, that relate to tax-time financial products:

• EROs must obtain taxpayers' written consent before disclosing any tax return information to other parties in relation to an application for a tax product.

• EROs must ensure taxpayers understand that if they use a tax product, the refund will be sent to the bank and not to them.

• If taxpayers choose to use a fee-based loan, EROs must advise that the product is an interest-bearing loan and not an expedited refund.

• EROs must advise taxpayers that the bank may charge them interest, fees, or both, in the case of any shortages on the refund.

• EROs also must disclose all deductions to be made from the expected refund and the net amount of the refund.⁴⁹

In 2015, OCC issued risk-management guidance for national banks that offer tax refund-related products. This guidance advises that banks should specify to customers, as applicable,

• the total cost of the tax product, separately from the tax preparation cost;

• that total costs will be deducted from and reduce the refund amount;

• that tax refunds can be sent directly to the taxpayer without the additional costs of a tax product;

• that customers with deposit accounts can receive their refund without incurring fees through direct deposit in about the same time as it would take to receive a tax refund-related product;⁵⁰ and

• the ongoing periodic maintenance and transaction fees related to any product intended for long-term use.

In addition, OCC's guidance establishes that banks should clearly disclose all material aspects of the product in writing before the consumer applies or pays any fees for a tax-time financial product.

Also, representatives of the American Coalition for Taxpayer Rights, a group representing the leading tax preparation, tax software, and bank providers, told us that its members signed a joint statement with attorneys general from six states on disclosure practices for refund transfers.⁵¹ The member providers agreed to explain to taxpayers the different options for filing and receiving a tax refund, including no-cost options, and the associated costs and features of each option. The providers also agreed to disclose the optional nature of the products, the timing of the refund, and to present the disclosures in a clear and conspicuous manner understandable by a reasonable consumer.

Our nongeneralizeable review of documents received from selected banks and tax preparers found disclosures generally followed OCC guidance or IRS requirements, respectively. However, our review of these documents and selected tax preparer websites also found — and our undercover visits of selected tax preparers suggested — that the level of transparency on product fees varied and product fees and information were not always clearly disclosed.

• Bank documents were more likely than information provided by paid preparers (in person or online) to include more disclosures about the fees and terms of tax-time financial products. For example, of the 12 bank documents we reviewed, all disclosed that funds would be sent to the bank if the taxpayer used a tax product. Almost all the bank documents disclosed the fees associated with the product and all disclosed that the fees would be deducted from the refund. In contrast, while written disclosure is not required, less than one third of ERO documents disclosed that the taxpayer using a tax-time financial product would receive funds from the bank instead of IRS.

• However, almost all the documents are presented to taxpayers after returns have been prepared and preparers have determined that taxpayers qualified for a product. The timing of when a tax preparer makes these disclosures would pose a challenge for taxpayers looking to compare prices for different providers. That is, they would not learn of the total fees — partly because the paid preparer could not determine the amount of some tax preparation fees until well into the preparation of the tax return.

• A taxpayer trying to determine the cost of using a tax refund to pay for online tax preparation services only would be able to compare the prices of two of the eight online providers we reviewed. The remaining six did not disclose this fee in a prominent way — with some disclosures made in small print or requiring navigation through several pages after the product page — or at all.

• A taxpayer choosing to file taxes using the services of a paid tax preparer in a brick-and mortar-location, and opting to use the refund to pay for tax preparation fees, would be unlikely to be able to compare prices among different providers.⁵² For example, during six of our undercover visits, our investigators explicitly requested literature on product fees. However, the preparers stated that they did not have the literature available or only provided us with business cards and other promotional material.

• Our analysis shows that providers do not consistently explain products or disclose fees to taxpayers. For example, providers told us, and industry documents show, that a refund transfer is not required to get a refund advance. However, during our site visits, tax preparers tied the use of a refund transfer to a refund advance four out of five times. In two of these cases, the tax preparer included the fee for a refund transfer as part of processing an advance product, while in another two cases the tax preparer said that a refund transfer was required with the advance. Also, during our site visits, three of the nine tax preparers did not disclose the cost of a refund transfer.

Appendix III provides more information on our analysis of bank and tax preparer disclosure practices.

According to industry participants, only taxpayers expecting a refund can qualify for a tax product; consequently, the tax preparer generally cannot determine whether the taxpayer qualifies until after the tax return is completed. Once this is determined, the tax preparer must request the taxpayer's consent to offer a tax product.⁵³ EROs with whom we met told us they may disclose fee information at various points throughout the process of tax preparation, and do so verbally or through their in-store computer interface. Bank disclosures are provided to the taxpayer before the product application has been submitted.⁵⁴

Some researchers and representatives from consumer advocacy organizations with whom we met were concerned about the timing of disclosures of tax-time financial product fees. Consumer advocates said disclosures given to taxpayers were inadequate, unhelpful, or timed in such a way as to prevent meaningful comparison shopping. Specifically, one consumer advocacy organization said that taxpayers they serve do not understand the fees associated with filing through preparers. Representatives from another consumer advocacy organization said that taxpayers do not know the total cost for tax-related financial products and services until they already have taken steps to file their returns. In its 2017 Report to Congress, the National Taxpayer Advocate recommended that IRS require all e-file participants offering tax-refund financial products to provide a standard “truth-in-lending” statement to help taxpayers better understand the terms of the refund anticipation loan product.⁵⁵ IRS did not adopt the National Taxpayer Advocate's recommendation but agreed that e-file providers should be transparent about the costs associated with the loan products offered to taxpayers as part of the return preparation process.

As previously discussed, courts have determined that IRS does not have sufficient authority to regulate individuals who are solely tax preparers and not licensed by IRS — in effect, the majority of the paid preparer population.⁵⁶ Previously, we asked Congress to consider legislation granting IRS the authority to regulate paid tax preparers, if it agreed that significant paid preparer errors existed.⁵⁷ As of March 2019, this Congressional action we have recommended remains open. The lack of consistency about the timing of fee disclosures for tax-time financial products may add to the rationale for Congress to consider regulating preparers. Such statutory authority could allow IRS to require that tax preparers make tax-time financial product disclosures or ensure meaningful transparency in the sale of the products.

Conclusions

For lower-income taxpayers with pressing financial obligations, tax-time financial products can offer an alternative to higher-cost short-term products such as payday loans. Taxpayers can purchase tax-time financial products from many tax preparers; however, according to our review of selected tax preparers and banks, the price and associated fees of these products can vary. And disclosure practices by some paid tax preparers may pose challenges for consumers looking to compare prices for different providers.

IRS is an essential source for data on tax-time financial products, but to date IRS has offered limited options to tax preparers for accurately reporting usage of all available tax-time products. Furthermore, IRS has not informed tax preparers about changes made in reporting options and has not informed users of IRS's product data about known issues with the data. Consequently, data on product usage are not reliable. Improving the quality of data collected on these products would help ensure that federal agencies, policymakers, regulators, consumer advocacy groups, and researchers have quality information to report on tax policy and consumer protection issues and inform their decision-making.

Recommendations for Executive Action

We are making a total of two recommendations to IRS.

The Commissioner of Internal Revenue Service should communicate data issues regarding the refund anticipation loan indicators for tax years 2016 and 2017 and the refund transfer indicators since tax year 2016 — for example, by attaching explanatory material to the dataset. (Recommendation 1)

The Commissioner of Internal Revenue Service should improve the quality of tax-time financial product data collected; for example, by allowing authorized e-file providers to indicate more than one type of tax-time financial product for each return or by informing tax preparers of the addition of new product definitions and instructions on how to accurately code the products. (Recommendation 2)

Agency Comments and Our Evaluation

We provided a draft of this report to IRS, FDIC, Federal Reserve, OCC, CFPB, and FTC for review and comment. IRS provided written comments, which are reproduced in appendix IV and discussed below. FDIC, Federal Reserve, OCC, CFPB, and FTC provided technical comments, which we incorporated as appropriate.

In its comments, IRS concurred with both recommendations, and described how it planned to address them. In response to our first recommendation, IRS stated that it plans to provide the appropriate notations with the datasets. In response to our second recommendation, IRS stated that it plans to pursue programming changes and clarify instructions for tax return preparers to promote accurate coding of refund-related products. We believe that these actions, if implemented, would address our recommendations and improve the quality of data IRS reports on these products.

As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the appropriate congressional committees and IRS, FDIC, Federal Reserve, OCC, and FTC. This report will also be available at no charge on our website at http://www.gao.gov.

If you or your staff have any questions about this report, please contact me at (202) 512-8678 or clementsm@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix V.

Michael Clements
Director, Financial Markets
and Community Investment

---

Appendix I: Objectives, Scope, and Methodology

This report (1) describes trends in the market for tax-time financial products and product fees and examines the reliability of IRS data on these trends, (2) describes characteristics of those who use tax-time financial products and factors that influence the decision to obtain the products, and (3) describes regulatory oversight of industry participants and the disclosure of information on product fees and terms.

To examine trends in the use of tax-time financial products, we used 2008–2018 Internal Revenue Service (IRS) data compiled from tax filings to determine the types and use of these products.¹ We assessed the reliability of these data by interviewing IRS officials about the controls and quality assurance practices they used to compile these data. We determined the data alone did not provide a reliable count of refund transfers, refund anticipation loans, or refund advances in 2016, 2017, and 2018, but were adequate to suggest general trends when supplemented with other information. To supplement the IRS data, we collected information from reports issued by the National Consumer Law Center, reviewed Securities and Exchange Commission filings for two selected tax preparers, and interviewed representatives from National Consumer Law Center and both tax preparers on the offerings of tax-time financial products. We selected these preparers because they are major providers of tax preparation services and tax products.

To identify and review trends in product offerings, we reviewed the websites, promotional materials, and other industry literature including Securities and Exchange Commission filings of a nongeneralizeable selection of four providers of online tax preparation services, three tax preparers with physical locations that also offer services online, and four banks. We also discussed changes in the market and product offerings with nine of the industry providers with whom we met. We accessed provider websites before and during the 2018 tax season. The tax preparation firms were selected because they are national tax preparation chains, and the five banks were selected because they partnered with the national tax preparation chains and major developers of tax preparation software. In addition, we reviewed studies related to these products published by GAO, federal agencies, four consumer advocacy and research groups, and two academic researchers. We used these studies primarily to corroborate findings from our data analysis. We focused on studies from 2010 and later; however, we also reviewed an older report to gain a greater understanding of how the market for tax-time financial products evolved. We identified these studies through expert recommendations and citations in studies.

To examine trends in fees for tax-time financial products, we collected fee-related information from several different sources (because of limited publicly available industry data). All of the information cannot be used to generalize our findings to the retail tax preparation industry.

• Product fees. For 2018, we collected information on product fees from six paid tax preparers and four banks. For tax years 2014 to 2017, we used product fee information as reported by the National Consumer Law Center. For 2018, we also reviewed fee data from six providers of online tax preparation software, two that provide services in person and online, and four that only provide services online. We selected these providers after conducting internet searches and reviewing reports by consumer advocates and federal agencies. Data elements included fees for refund transfers and refund advances. For 2018, data elements also included the dollar amount for the incentives banks offered tax preparers for each refund transfer sold.

• Ancillary product fees. We collected information on ancillary product fees from four tax preparers, four banks, and three software developers for tax years 2017 and 2018. Data elements included fees for disbursement methods such as prepaid cards and paper checks and other charges related to the use of a tax-time financial product such as technology and transmission fees.

• Tax preparation fees. We collected information on tax preparation fees from eight tax preparers with physical locations and eight online providers of tax preparation services for 2018. Data elements included fees for federal and state filing.

• Aggregate fees. We collected aggregate tax-time financial product, ancillary product, and tax preparation fee information from studies issued by consumer protection advocates.

We collected the above information from websites, advertising materials, and public filings with the Securities and Exchange Commission of tax preparers, banks, and software developers.

To identify some of the demographic and economic characteristics of product users, we used data from the Bureau of the Census and the Federal Deposit Insurance Corporation (FDIC) from 2011, 2013, 2015, and 2017 to conduct a multivariate regression analysis to determine the influence of individual characteristics on the decision to obtain a product. We statistically controlled for various income, education, and demographic factors. While the FDIC data contain a rich set of demographic and economic variables, they include limited data on characteristics specifically related to tax filing. To identify specific tax-filing characteristics associated with product use, we also used a probability sample of data from IRS from the 2014, 2015, and 2016 tax years to calculate the percentages of taxpayers who used tax-time financial products according to various tax-filing characteristics, including tax filing status and tax filing method. We also used the sample data to calculate the percentage of taxpayers who used free filing services, including free file software, programs, and fillable forms. We reviewed documentation on and conducted testing of the data we used and determined they were sufficiently reliable for reporting economic, demographic, and tax-filing characteristics associated with product use. For more detailed information on our analysis of characteristics associated with tax-time financial product use, see appendix II.

To better understand user characteristics associated with the decision to obtain a tax-time financial product identified by our analysis, we reviewed relevant federal and industry reports on the financial needs of individuals with characteristics similar to taxpayers who obtained these products. We focused on reports from 2010 and later. We also reviewed our prior studies and studies from the Consumer Financial Protection Bureau (CFPB) on alternative credit products and compared their features and fees to those of tax-time financial products. In addition, we interviewed representatives from consumer groups, four Low-Income Taxpayer Clinics, and IRS's Taxpayer Advocate Service to obtain their perspectives on characteristics associated with tax-time financial product users.

To describe the regulatory oversight of industry participants associated with tax-time financial products, we reviewed relevant federal laws and regulations, and reports and guidance documents from IRS and federal regulators, including the CFPB, FDIC, the Board of Governors of the Federal Reserve System, Office of the Comptroller of the Currency (OCC), and Federal Trade Commission. We inquired about consumer complaint data related to tax-time financial products at the federal regulators and interviewed officials from the federal agencies and representatives from five tax preparation providers, five banks and bank affiliates such as settlement service providers, four consumer advocacy organizations, three software developers, two researchers, one provider of alternative financial services, and one industry group to gain their perspectives on the benefits and risks of the tax-time financial products and how any related concerns were being addressed. The tax preparation firms were selected because they are national tax preparation chains, and the five banks and three software developers were selected because they partnered with the national tax preparation chains. The four consumer advocacy organizations, two researchers, alternative financial service provider, and industry group were selected for their experience and to provide a range of perspectives.

To review how product terms and fees are disclosed by tax preparers, in February 2018 GAO investigators acting in an undercover capacity visited a nongeneralizeable sample of nine randomly selected tax preparers in Washington, D.C., Maryland, and Virginia to inquire about tax-time financial products. We selected the two states and Washington, D.C. to ensure a mixture of state and local laws governing the products and providers. From the two states and Washington, D.C., we selected one metropolitan statistical area based on the concentration of product users and the proximity to lower-income households. We randomly selected three individual tax preparers in each of the three metropolitan statistical areas to visit, based on proximity to taxpayers in lower-income households and to ensure a mixture of urban and rural communities and company sizes. We visited offices of large tax preparation chains and single-office tax preparation businesses. Results cannot be used to generalize our findings to the retail tax preparation industry. Our investigators posed as taxpayers seeking tax preparation services who wanted to pay for the tax preparation fees with the expected refund or obtain an advance based on their anticipated tax refund. They requested available documents associated with tax preparation, refund advance and refund transfer products, and different disbursement options and fees. Because GAO investigators did not experience the tax preparation or the product application process, we were not able to assess the timing of any disclosures typically made after the tax return preparation process would begin. In addition, we received some consumer-facing disclosures and product agreements that were typically provided during the product application process from two tax preparers and two banks.

We also conducted a content analysis of websites of eight selected tax preparers that offer tax-time financial products. The tax preparers were selected as national providers of tax preparation services with an online presence, and the results are not generalizeable to the retail tax preparation industry. Three of the providers offer tax preparation services online and through physical retail locations and five of the providers offer their services online only. We reviewed these websites to understand the extent to which they disclose fees to the taxpayer for tax preparation services, tax-time financial products, disbursement, and additional products or services, and to review the ease with which these disclosures are accessible.

In addition to consumer-facing disclosures we received from providers with whom we met, we searched online for additional disclosures provided by the tax preparers and banks in our review and reviewed seven disclosures from two national tax preparation chains and 12 disclosures from five banks offering tax-time financial products. We then compared the disclosures against IRS and OCC requirements for disclosure for product terms and conditions. IRS established certain disclosure requirements for authorized IRS e-file providers.² OCC instructs banks it supervises to make certain disclosures to product consumers.³ More specifically, we analyzed tax products and fee disclosures obtained from our undercover visits of selected tax preparers, online reviews, and directly from tax preparers and banks to determine the type and timing of disclosures made in these instances and whether they were consistent with IRS disclosure requirements and followed OCC guidance.

We conducted this performance audit from July 2017 to April 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. We conducted our related investigative work in accordance with standards prescribed by the Council of the Inspectors General on Integrity and Efficiency.

---

Appendix II: Analysis of Characteristics Associated with Tax-Time Financial Product Use

This technical appendix outlines the development, estimation, results, and limitations of the econometric model and other data analysis we described in the report. We undertook this analysis to better understand the characteristics associated with the decision to obtain a tax-time financial product.

Data

Federal Deposit Insurance Corporation. To assess the characteristics associated with tax-time financial product use, we used data from the Federal Deposit Insurance Corporation's (FDIC) National Survey of Unbanked and Underbanked Households for 2011, 2013, 2015, and 2017, which is a supplement of the Current Population Survey. We used the following variables on households and heads of households to examine how various demographic and economic characteristics are related to the use of tax-time financial products:

• Household income.

• Household type.

• Homeownership status.

• Race and ethnicity of the head of household.

• Educational attainment of the head of household.

• Age of the head of household.

• Head of household has children.

• Household used refund anticipation loan or a tax preparation service to receive a tax refund faster than the Internal Revenue Service (IRS) would provide it in the past 12 months. This is a dummy variable, which equals 1 if the household used products and 0 otherwise.

  • A refund anticipation loan is a tax-time financial product. Based on our interviews and other research reports, refund anticipation loans and other tax-time financial products (including refund anticipation checks) may be used by consumers to get their tax return faster than IRS could provide it. We refer to this variable as “used tax-time financial product” for simplicity in the report, and we explain the relevant caveats and limitations below.

  • This variable is the basis for the sample used for this analysis.

See table 3 for the estimated distributions of these variables for all households, as well as households that used tax-time financial products in 2017.

Table 3: Characteristics of Households and Heads of Households, 2017

Household characteristics		Estimated percentage of population	Estimated percentage of population that used tax-time financial products
Income	Less than $10,000	6.6	7.7
	$10,000 to $19,999	10.3	12.3
	$20,000 to $29,999	10.6	12.9
	$30,000 to $39,999	11.5	16.9
	$40,000 to $49,999	8.2	8.4
	$50,000 to $59,000	8.1	8.4
	More than $60,000	44.6	33.5
Household type	Married couple	47.4	36.2
	Unmarried male-headed family	5.1	6.7
	Unmarried female-headed family	11.8	22.9
	Single male	16.9	17.7
	Single female	18.8	16.5
Homeownership	Homeowner	63.9	44.4
	Non-homeowner	36.1	55.6
Head of household characteristics
Race/ethnicity	White, non-Hispanic	66.7	55.6
	African American, non-Hispanic	12.7	20.6
	Asian, non-Hispanic	4.9	5.8
	American Indian/Alaska Native, non-Hispanic	0.8	1.0
	Hispanic, any race	13.5	15.6
	Mixed race/other, non-Hispanic	1.5	1.5
Education	Less than college education	35.2	37.9
	Some college education or more	64.8	62.1
Age	15–29 years	12.8	23.2
	30–39 years	17.0	25.6
	40–49 years	16.9	18.3
	50–59 years	18.9	15.6
	60 years and older	34.3	17.3
Children	Has own children	26.7	37.5
	Has no own children	73.3	62.5
Sample size	33,561
Estimated population size	123 million
Source: GAO analysis of Federal Deposit Insurance Corporation data. | GAO-19-269 Notes: We used data from the 2017 Federal Deposit Insurance Corporation's National Survey of Unbanked and Underbanked Households. The sample size is 33,561, representing an estimated population of about 123 million. Of the households sampled, 798 used tax-time financial products, representing an estimated population of about 3 million, and 32,372 did not use the products, representing an estimated population of about 118 million. We estimated that 2.4 percent of households used the products, plus or minus 0.2 percentage points. The first column is the estimated percentage of households and heads of households in the sample, conditional on being part of various demographic subgroups. These statistics are weighted using household-level weights. All estimates in the first column have relative standard errors of about 12.5 percent or less. The second column is the estimated percentage of households and heads of households who used tax-time financial products in the past 12 months, conditional on being part of various demographic subgroups. These statistics are also weighted using household-level weights. Estimates in the second column generally have relative standard errors of about 20 percent or less, with the exception of American Indian/Alaskan Natives and mixed race/other non-Hispanics due to relatively small sample sizes, but estimates for these subgroups are statistically significant.

We also examined the relationship between the use of tax-time financial products and being unbanked, as well as the association between using tax-time financial products and alternative financial services (those offered outside the banking system). We used additional data from FDIC's National Survey of Unbanked and Underbanked Households on the following variables:

• Household used other alternative financial services in the past 12 months, including nonbank check cashing, nonbank money orders, payday loans, and pawn shops.

• Household used prepaid card(s) in the past 12 months.

• Household was unbanked in the past 12 months.

See table 4 for estimated distributions of household responses to questions related to unbanked status and usage of other alternative financial services for all households, as well as households that used tax-time financial products in 2017.

Table 4: Household Responses to Questions Related to Unbanked Status and Alternative Financial Services Usage, 2017

		Estimated percentage of population	Estimated percentage of population that used tax-time financial products
Used nonbank check cashing in past 12 months	Used Did not use Did not know/refused to answer	6.2 92.2 1.6	15.5 84.4 0.0
Used nonbank money order in page 12 months	Used Did not use Did not know/refused to answer	14.0 84.2 1.7	32.7 67.3 0.0
Used payday loan in past 12 months	Used Did not use Did not know/refused to answer	1.8 96.6 1.7	7.9 7.9 0.1
Used pawn shop in past 12 months	Used Did not use Did not know/refused to answer	1.5 96.9 1.6	6.7 93.3 0.0
Used prepaid card in past 12 months	Used Did not use Did not know/refused to answer	9.5 88.9 1.6	21.4 78.4 0.2
Unbanked in past 12 months	Unbanked Banked Did not know/refused to answer	3.9 89.4 0.4	5.4 85.2 0.1
Source: GAO analysis of Federal Deposit Insurance Corporation data. | GAO-19-269 Notes: We used data from the 2017 Federal Deposit Insurance Corporation's National Survey of Unbanked and Underbanked Households. The sample size is 33,561, representing an estimated population of about 123 million. Of the households sampled, 798 used tax-time financial products, representing an estimated population of about 3 million and 32,372 did not use the products, representing an estimated population of about 118 million. We estimated that 2.4 percent of households used the products, plus or minus 0.2 percentage points. Sample size is slightly different for unbanked status (N = 31,653) due to nonresponse, and unbanked means no one in the household had a checking or savings account in the past 12 months. As a result, the percentages for this variable do not add to 1. The first column is the estimated percentage of households in the full sample that used alternative financial services, prepaid cards, or were unbanked in the past 12 months. All estimates in the first column have relative standard errors of about 10 percent or less. The second column is the estimated percentage of households that used tax-time financial products in the past 12 months and used alternative financial services, prepaid cards, or were unbanked in the past 12 months. All statistics are weighted using household-level weights. Estimates in the second column generally have relative standard errors of 20 percent or less with some exceptions. Estimates of those who did not know or refused to answer about nonbank check cashing, payday loan, pawn shop, and prepaid card use and unbanked status are not statistically significantly different from zero.

IRS. To further identify tax-filing characteristics associated with tax-time financial product use and trends, we also used data from a probability sample of 2 percent of all electronically filed tax returns from IRS for tax years 2014, 2015, and 2016. In 2016, the sample size was 2,952,418, representing a population of 147,625,598 tax returns. According to IRS, the sample is representative of all electronically filed tax returns for the relevant tax years. In this sample, IRS provided data on the following variables:

• Tax filing method, including online (self-filed using tax software) or through a paid practitioner (including tax preparers with physical storefronts).

• Taxpayer used free filing services from IRS, including the Free File program and free fillable forms.

• Tax filing status, including single, married, and head of household.

• Disbursement options for tax refunds (direct deposit or paper check) or tax balance due.

• Tax refund amount.

• Tax year.

• Tax-time financial product use, including refund anticipation loans, refund anticipation checks, or no tax-time financial products. In tax year 2016, we estimated that about 18 percent of taxpayers used a tax-time financial product, plus or minus less than 1 percentage point.

We also used IRS data from the Statistics of Income division for tax year 2016 to assess the geographical concentration of product use at the zip-code level. Zip code data from the IRS Statistics of Income division are based on population data that was filed and process by IRS in tax year 2016. Due to some data suppression from IRS for privacy purposes, zip codes with less than 100 tax returns are excluded from the data. As a result, in 2016 the total returns represented in the IRS zip code data are 145,302,140 and the number of tax returns with a tax-time financial product was 21,654,760, meaning about 15 percent of tax filing units in these data used a tax-time financial product.

Methodology

Regression analysis using FDIC data. Using FDIC data, we conducted a multivariate regression analysis to examine the relationship between each explanatory variable and tax-time financial product use. Specifically, we estimated multivariate logistic regression models. Regression models allow us to test significant relationships between economic and demographic variables and the likelihood of using tax-time financial products, while controlling for other factors.

We used logistic regression models because our dependent variable is binary. The dependent variable represents whether a household used tax-time financial products. We collapsed “no” and “did not know/refused” into a single category for our regression analysis, so that the dependent variable is equal to 1 if the household used tax-time financial products and 0 otherwise.

Logistic regressions allow the relationships between various characteristics and tax-time financial product usage to be described as odds ratios. Odds ratios that are statistically significant and greater than 1.00 indicate that households or heads of households with those characteristics are more likely to use tax-time financial products. Odds ratios that are less than 1.00 indicate that households or heads of households with those characteristics are less likely to use tax-time financial products. For categorical variables, this increase or decrease in the likelihood of product use is in comparison to an omitted category, or reference group. For example, the odds ratio for households headed by African Americans is statistically significant and 1.36. This implies that the odds of tax-time financial product use for households headed by African Americans are 1.36 times the odds of use for households headed by whites, holding other factors constant. Put another way, households headed by African Americans are about 36 percent more likely to use tax-time financial products than households headed by white individuals, if other conditions remain constant. This result and others are discussed further in the results section below. We also present 95 percent confidence intervals, which helps clarify the statistical significance of the odds ratios.

Our baseline estimates were derived from logistic regressions that accounted for the survey features of the FDIC data. Our main regression results used data from the 2017 survey year. We also estimated logistic regressions using data from the 2015, 2013, and 2011 survey years, using the same variables when possible. Our baseline specification includes explanatory variables for race and ethnicity, education, age, household type, income, and homeownership. We used groups of indicator variables or categorical variables to control for all characteristics. In other specifications, we included controls for children, unbanked status, use of alternative financial services other than tax-time financial products, state indicators, and region indicators to check the robustness of our results.

We also assessed the sensitivity of our analyses by restricting the analysis to households that only answered “yes” or “no” to tax-time financial product use. We excluded answers of “did not know/refused,” so that the dependent variable is equal to 1 if the household used tax-time financial products and 0 if the household did not use tax-time financial products.

In a more limited analysis, we merged data from the 2017 FDIC data, which is the June 2017 supplement of the Current Population Survey, with the 2017 Annual Social and Economic Supplement, which is the March 2017 supplement of the Current Population Survey. We performed the additional analysis because the March 2017 supplement has data on tax-filing characteristics, including tax credits used by households. Given the structure of the Current Population Survey, some households were surveyed in both the March and June 2017 supplements, and those households comprise the sample used in this part of the analysis. We identified those represented in both supplements using household and person identifiers, as well as data on sex, race and ethnicity, and age. Using this merged sample, we estimated logistic regressions that both did and did not account for the survey features of the data. We included the same explanatory variables as our baseline estimates, along with indicators for use of the Earned Income Tax Credit, Additional Child Tax Credit, and Child Tax Credit.

Analysis of IRS data. Using the 2 percent sample of IRS data, we estimated the percentages of tax filers with varying tax-filing characteristics by year and average refund amounts by year. All estimates are weighted at the tax filing unit level. Using the IRS's zip code data from the Statistics of Income division for 2016, we calculated the number of total tax filing units and tax filing units who used tax-time financial product at the zip code level.

Caveats and Limitations

Regression analysis using FDIC data. Our results have limitations and should be interpreted with caution. For example, our analysis identifies correlations between characteristics and tax-time financial product use and not causal relationships. Moreover, there may be variables that are correlated with tax-time financial product use that are not included in our models. For example, we are not able to account for community characteristics that may influence the decision to use the products due to data limitations. We used statistical tests for multicollinearity (high intercorrelations among two or more independent variables) and goodness of fit to check the validity of the model to the extent possible, given the use of complex survey data.

Our analysis of the characteristics associated with the use of tax-time financial products uses a relatively small number of observations. For example, we observe 798 households that used these products in the 2017 survey year, representing about 2.4 percent of households (plus or minus 0.2 percentage points), and that is the benchmark utilization rate against which the results should be interpreted. Moreover, IRS data indicate that more than 20 million tax filers used tax-time financial products in 2016, representing about 20 percent of tax filers who filed their taxes electronically. These data sets use different units of analysis, and there can be multiple tax filers in one household, especially for those who use Earned Income Tax Credit. However, comparing the two suggests that the survey data may not include all users of tax-time financial products. Given the question used to measure the dependent variable, our analysis focuses on those who use tax-time financial products to get their tax refund more quickly. While a key reason people use tax-time financial products is to meet cash needs, there may be other reasons people use the products, including covering the cost of tax preparation.

Our results may not generalize to other time periods. There have been a number of changes in the market for tax-time financial products in recent years. Our results may not generalize to all products currently available in the market. However, our results from 2017 are generally similar with the 2015, 2013, and 2011 survey years, despite a number of changes to the tax-time financial product market during these years. Our findings suggest that similar types of households have utilized tax-time financial products regardless of industry and market changes, particularly if households used paid preparers and tax-time financial products to expedite their tax refunds.

Our analysis focuses on households that used tax-time financial products and accessed them through paid preparers. However, taxpayers also may have accessed specific types of tax-time financial products when they used online software to file their own taxes. For example, individuals who file their own taxes online may use the products to cover the cost of the software that helps them prepare their taxes. The characteristics of people who use products for these reasons may be different than what we found in our analysis.

Analysis of IRS data. The IRS data are representative of tax returns filed electronically and not of tax returns filed by other means, including by paper. The results may not generalize to years for which we do not have data.

The indicators in the data for specific types of tax-time financial products, including the indicators for refund anticipation loans and refund anticipation checks have some significant limitations. In tax years 2014– 2016, IRS only allowed tax-time financial products to be coded as refund anticipation loans or refund anticipation checks (that is, there was no code to indicate that two or more products were used together). However, there were some major changes in the industry during this period, particularly with regards to refund anticipation loans, that suggest that these indicators do not measure the same types of products over time. Given the limitations of the definitions of specific tax-time financial products, most of our analysis focuses on the universe of tax-time financial products in the IRS data and not on differences by specific types of products.

Results

Regression analysis using FDIC data. Our analysis suggests a number of economic and demographic characteristics are associated with tax — time financial product use, particularly when purchased through a tax preparer to expedite the tax refund, after controlling for other factors. In 2017, relatively lower-income households were more likely to use the products than higher-income households. Households headed by single women with families were more likely to use tax-time financial products than households headed by married couples. Furthermore, householders who owned their homes were less likely to use tax-time financial products. African American households were more likely to use the products compared to white households. Finally, relatively younger households were more likely to use the products than older ones. The results of the main specification of our logistic regression are presented in table 5.

Table 5: Factors Associated with Tax-Time Financial Product Use, 2017

Explanatory variables	Odds ratios	95% confidence interval lower bound	95% confidence interval upper bound
Income (omitted — income $60,000 or more)
Income less than $10,000	1.02	0.68	1.52
	(0.21)
Income between $10,000 and $19,999	1.32	0.94	1.86
	(0.23)
Income between $20,000 and $29,999	1.34*	1.00	1.81
	(0.20)
Income between $30,000 and $39,999	1.61***	1.24	2.09
	(0.22)
Income between $40,000 and $49,999	1.19	0.87	1.62
	(0.19)
Income between $50,000 and $59,999	1.22	0.86	1.74
	(0.22)
Household type (omitted — married)
Unmarried male head of household with family	1.25	0.88	1.79
	(0.23)
Unmarried female head of household with family	1.76***	1.36	2.30
	(0.24)
Single male	1.11	0.83	1.47
	(0.16)
Single female	1.12	0.82	1.54
	(0.18)
Homeownership (omitted — non-homeowner)
Head of household is homeowner	0.66***	0.52	0.82
	(0.08)
Children (omitted — no children)
Head of household has children present	1.13	0.88	1.44
	(0.14)
Race and ethnicity (omitted — white, non-Hispanic head of household)
African American, non-Hispanic head of household	1.36**	1.07	1.73
	(0.17)
American Indian/Alaskan Native, non-Hispanic head of household	1.17	0.52	2.64
	(0.49)
Asian, non-Hispanic head of household	1.18	0.74	1.88
	(0.28)
Hispanic, any race of head of household	0.93	0.73	1.20
	(0.12)
Mixed race/other non-Hispanic head of household	0.84	0.39	1.82
	(0.33)
Education (omitted — no college education)
Head of household has some college education or more	0.96	0.81	1.13
	(0.08)
Age (omitted — 60 years or older)
Age of head of household between 15 and 29	2.55***	1.87	3.48
	(0.41)
Age of head of household between 30 and 39	2.36***	1.72	3.23
	(0.38)
Age of head of household between 40 and 49	1.89***	1.36	2.64
	(0.32)
Age of head of household between 50 and 59	1.62***	1.21	2.17
	(0.24)
Number of observations	33,561
Legend: *** = p<0.01; ** = p<0.05; and * = p<0.1. Source: GAO analysis of Federal Deposit Insurance Corporation data. | GAO-19-269 Notes: We used from data from the 2017 Federal Deposit Insurance Corporation's National Survey of Unbanked and Underbanked Households. Odds ratios are estimated from a multivariate logistic regression that accounted for the survey features of the data. Standard errors are calculated using successive difference replication based on the household weight and replicate weights are in parentheses. The baseline household characteristics (omitted categories) are households with incomes over $60,000, married couples, non-homeowners, white and non-Hispanic heads of households, heads of household with no college education, heads of household over 60 years old, and heads of household with no children.

Our results for other specifications using 2017 data were generally similar. For example, adding an additional control for unbanked status did not substantively change the results. In alternative specifications that included an indicator for use of other alternative financial services, we found a significant and positive correlation between using tax-time financial products and other alternative financial services, including nonbank check cashing, nonbank money orders, payday loans, and pawn shops. Moreover, including state and region indicators did not substantively affect the results. Using the sample restricted to just “yes” and “no” responses also did not substantively change the results.

Our results for other years were generally similar, with some exceptions. For example, in other survey years prior to 2017, we found that in addition to African American households, Native American households also were more likely to use tax-time financial products than white households. Moreover, education and children were significant correlates in prior survey years.

Analysis of IRS data. We found that nearly 1 in 5 taxpayers who filed their taxes electronically used tax-time financial products each year from 2014 to 2016, while less than 3 percent of filers used free filing services available through IRS during the same period.

We also found that in 2016, tax-time financial product use was associated with receiving tax refunds through direct deposit, which is a faster way to receive a tax refund than paper check. Users of tax-time financial products also were more likely to file as heads of household (tax filing status) than taxpayers who did not use tax-time financial products. Moreover, taxpayers who used the products received higher tax refunds on average than taxpayers who did not use the products, especially when they used paid tax preparers to file their taxes.

Finally, analyzing the zip code of the filers, we found that use of tax-time financial product was concentrated in some areas of the South and the West.

---

Appendix III: Disclosure of Product and Related Fees and Terms

Disclosure of Product Fees and Terms

Our limited nongeneralizeable review of documents received from selected banks and tax preparers found disclosures generally followed Office of the Comptroller of the Currency (OCC) guidance or Internal Revenue Service (IRS) requirements for fees disclosure, respectively. However, we noted from our undercover visits of selected tax preparers that the extent and clarity of the disclosures offered to customers varied. Furthermore, in our review of selected tax preparers' websites, we found that fees and information about products were not always clearly disclosed.

Undercover Visits

All nine tax preparers we visited offered the option to pay for the tax preparation fees with the tax refund by using a refund transfer, but they did not always clearly communicate how these options work.¹ For example, three preparers did not disclose the refund transfer fee, and in a few instances, the refund transfer was provided alongside a refund advance and we were not given the option to pay for the tax preparation fees out of pocket. In other cases, the refund transfer fee was disclosed, but the product was not always identified as optional (that is, not required for tax preparation).

During six of our undercover visits, our investigators explicitly requested literature on product fees. However, the preparers either stated they did not have the literature available or only provided us with business cards and promotional material. The other three times we did not ask for, and were not offered literature on product fees, features, or terms.

In two of our visits, the tax preparers offered our investigators a refund advance after we expressed an interest in getting the refund quickly. In another two visits, we were offered unsolicited refund advances. When offering the product, these four tax preparers bundled the refund advance with a refund transfer (an optional product). By adding a refund transfer, the tax preparer effectively added a fee-based product to the refund advance, a product that otherwise is free to the taxpayer. During one of the visits, we were offered a refund advance only after we specifically asked for it.

Website Content Analysis

We reviewed the websites of eight selected providers of tax preparation services. We found that while these providers generally disclosed product fees, these disclosures were not made in a consistent manner. For example, all eight of the websites we reviewed offered taxpayers the option to use the expected refund to pay for tax preparation fees. Most of the time, the fee associated with this option was not clearly disclosed on the website. Only two of the eight providers clearly disclosed this fee on the products page; the other six did not disclose the fee in a prominent way or at all. In addition, all five providers that offered refund advances fully disclosed fee information for this product.

Three of the eight online tax preparation service providers had physical locations in addition to their online presence. Of these three, only one disclosed on its website the refund transfer fee for taxpayers who filed a return in-person at one of their offices. For the second preparer with a physical presence, the refund transfer fee quoted for the online service was significantly lower than the fee we were quoted for in-person services at an office. The third preparer with a physical and online presence did not disclose the refund transfer fee for either the in-person service or online filing.

Document Review

We received and reviewed seven disclosure documents originated by two national tax preparation companies both of which are electronic return originators (ERO) and 12 bank documents from five banks in the industry. We compared the disclosure documents against IRS requirements for disclosure of fees for tax products and we compared the bank documents to OCC guidance related to disclosure of product, disbursement, and additional fees.² Both sets of documents in our nongeneralizeable review generally disclosed the product fees in accordance with IRS requirements or OCC guidance as appropriate. Bank forms, including disclosures, are presented to taxpayers once they have decided to apply for a tax product. This practice is consistent with OCC's guidance, which states that the details of a product should be provided to consumers before they apply for it. However, our analysis found that almost all of these documents are presented to taxpayers after returns have been prepared and tax preparers have determined the taxpayers were qualified for a tax-time financial product. The timing of when a tax preparer make these disclosures would make it challenging for a taxpayer to compare product prices from different providers or make more informed purchasing decisions.

Moreover, all the ERO documents we reviewed with information on refund advances disclosed that the taxpayer would be receiving a loan and not a refund. However, of the six ERO disclosure documents that disclosed fees, four disclosed additional fees that might be associated with tax refund products, such as disbursement fees.

Of the 12 bank documents we reviewed, all disclosed that funds would be sent to the bank if taxpayers used a tax product. Almost all the documents disclosed the fees associated with the tax product and that the fees would be deducted from the refund. And four of five documents related to a loan product disclosed that the taxpayer would be receiving a loan and not a tax refund. The majority of the documents also disclosed that the taxpayer may receive the refund directly from the taxing authority without incurring additional costs and within the same time frame without using a tax product.

All the tax preparer documents and the banks' disclosure documents were brief and written in plain language. However, almost all the bank application documents were longer than four pages and included technical and industry language.

Disclosure of Disbursement Fees, Including on Prepaid Cards

Based on our document reviews of selected tax preparers and banks and as suggested by our undercover visits of nine selected tax preparers, the disclosure of fees for disbursing funds was inconsistent, particularly around prepaid cards. Prepaid cards are often used to disburse funds from a tax-time product. Based on our analysis of providers' promotional content, in some cases a tax preparer will offer prepaid cards as the only disbursement option. The cards generally carry additional fees for long-term use (such as monthly, withdrawal, reload, and inactivity fees). Prepaid cards usually are reloadable and can be used to pay bills and make retail purchases. IRS does not have guidelines for disclosing fees for the long-term use of prepaid card. However, OCC requires that banks disclose if a tax product may be used on a long-term basis and disclose fees associated with extended use of the product.

During our visits, seven of the nine tax preparers provided the option to have the tax refund deposited on a prepaid card.³ However, only two of the seven preparers noted any potential fee information associated with the short or long-term use of prepaid cards. These two preparers said that there was no additional charge to have the taxpayer's refund deposited on a prepaid card, and the other five did not explain whether any fees would be charged for this transaction.

Five of the seven preparers that offered a prepaid card explained that the card could be used for transactions other than receiving the tax refund. However, only two of the five disclosed any fee information associated with long-term use of the card. Another two of the five preparers referred our undercover agents to the issuer of the card for additional information. The remaining preparer did not disclose that additional fees would apply to long-term use of the card.

Four of the eight tax preparation websites we reviewed disclosed partial information about fees related to the disbursement of funds to the taxpayer. Three of the eight websites only disclosed disbursement fee information related to use of prepaid cards. We found fee information in one of the eight websites only after doing a word search. Fees associated with the long-term use of prepaid card fees were not disclosed by three of the six preparers that offered this disbursement option. Two websites disclosed partial fee information and only one disclosed all the fees and terms associated with the long-term use of a prepaid card. Six of these websites advised the taxpayer to see the terms and conditions of the card, four included a link to the terms and conditions of the card, and two did not include a link.

Bank documents generally disclosed the fees associated with different disbursement methods such as paper checks and prepaid cards; however, fees related to the long-term use of prepaid cards were not always disclosed. Almost half of the documents we reviewed that include the use of a prepaid card did not acknowledge that fees were associated with the long-term use of prepaid cards, while others included only partial information or a general statement that “fees may apply.”

---

Appendix IV: Comments from the Internal Revenue Service

March 22, 2019

Mr. Michael Clements
Director, Financial Markets and Community Investment
U.S. Government Accountability Office
441 G Street, N.W.
Washington, DC 20548

Dear Mr. Clements;

I have reviewed the draft report entitled TAX REFUND PRODUCTS: Product Mix Has Evolved and IRS Should Improve Data Quality (GAO-19-269) and appreciate the opportunity to provide comments. Financial products associated with potential income tax refunds are marketplace offerings provided by the income tax return preparation industry to its customers. The IRS neither administers nor promotes the sale or use of these products. We recognize that variations have occurred in recent years affecting the nature and number of refund-related products and how they are promoted to the public.

As noted in the report, we have expanded the data captured on returns to more accurately reflect the products offered in today's market. We also continue to pursue programming changes that will further improve our ability to identify the use of multiple refund-related products on a single return. Better-defined descriptors of the refund-related products used by taxpayers will be useful for users of that data; however, the presence or absence of that data does not impact our ability to process tax returns or issue refunds. Consequently, due to programming resources being finite and subject to competing priorities, completion of this work is dependent on prioritization and funding.

We agree with your recommendations to communicate the limitations associated with the Refund Anticipation Loan indicators for tax years 2016 and 2017. Additionally, we will update Publication 1345, Handbook for Authorized IRS e-file Providers of Individual Income Tax Returns, to clarify guidance on how the use of refund-related products should be coded on returns.

Responses to your specific recommendations are enclosed. If you have any questions, please contact Michael Beebe, Director, Return Integrity and Compliance Services, Wage and Investment Division, at (470) 639-3250.

Sincerely,

Kirsten B. Wielobob
Deputy Commissioner for Services and Enforcement

Enclosure

---

Enclosure

Recommendations for Executive Action

RECOMMENDATION 1

The Commissioner of Internal Revenue Service should communicate data issues regarding the Refund Anticipation Loan (RAL) indicators for tax year 2016 and 2017 and the refund transfer indicators since tax year 2016 — for example, by attaching explanatory material to the dataset.

COMMENT

We agree with this recommendation and will provide the appropriate notations with the datasets.

RECOMMENDATION 2

The Commissioner of Internal Revenue Service should improve the quality of tax-time financial product data collected; for example, by allowing authorized e-file providers to indicate more than one type of tax-time financial product for each return or by informing tax preparers of the addition of new product definitions and instructions on how to accurately code the products.

COMMENT

We agree with this recommendation. Programming changes will be pursued and instructions for tax return preparers will be clarified to promote accurate coding of refund-related products.

---

Appendix V: GAO Contact and Staff Acknowledgments

GAO Contact

Michael Clements, 202-512-8678, or clementsm@gao.gov

Staff Acknowledgments

In addition to the contact named above, Karen Tremba (Assistant Director), Nathan Gottfried (Analyst in Charge), Jessica Artis, Maurice Belding, Evelyn Calderón, Farrah Stone, Kathleen McQueeney, Marc Molino, Neil Pinney, Barbara Roesmann, Jessica Sandler, Erinn Sauer, Erin Saunders-Rath, Michael Walton, and Helina Wong made significant contributions to this report.

GAO's Mission

The Government Accountability Office, the audit, evaluation, and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO's commitment to good government is reflected in its core values of accountability, integrity, and reliability.

Obtaining Copies of GAO Reports and Testimony

The fastest and easiest way to obtain copies of GAO documents at no cost is through GAO's website (https://www.gao.gov). Each weekday afternoon, GAO posts on its website newly released reports, testimony, and correspondence. To have GAO e-mail you a list of newly posted products, go to https://www.gao.gov and select “E-mail Updates.”

Order by Phone

The price of each GAO publication reflects GAO's actual cost of production and distribution and depends on the number of pages in the publication and whether the publication is printed in color or black and white. Pricing and ordering information is posted on GAO's website, https://www.gao.gov/ordering.htm.

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or TDD (202) 512-2537.

Orders may be paid for using American Express, Discover Card, MasterCard, Visa, check, or money order. Call for additional information.

Connect with GAO

Connect with GAO on Facebook, Flickr, Twitter, and YouTube.

Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts.

Visit GAO on the web at https://www.gao.gov.

To Report Fraud, Waste, and Abuse in Federal Programs

Contact FraudNet:

Website: https://www.gao.gov/fraudnet/fraudnet.htm

Automated answering system: (800) 424-5454 or (202) 512-7700

Congressional Relations

Orice Williams Brown, Managing Director, WilliamsO@gao.gov, (202) 512-4400, U.S. Government Accountability Office, 441 G Street NW, Room 7125, Washington, DC 20548

Public Affairs

Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800, U.S. Government Accountability Office, 441 G Street NW, Room 7149 Washington, DC 20548

Strategic Planning and External Liaison

James-Christian Blockwood, Managing Director, spel@gao.gov, (202) 512-4707, U.S. Government Accountability Office, 441 G Street NW, Room 7814, Washington, DC 20548