This is the accessible text file for GAO report number GAO-08-781 entitled 'Tax Preparers: Oregon's Regulatory Regime May Lead to Improved Federal Tax Return Accuracy and Provides a Possible Model for National Regulation' which was released on September 15, 2008. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. 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. Report to the Committee on Finance, U.S. Senate: United States Government Accountability Office: GAO: August 2008: Tax Preparers: Oregon's Regulatory Regime May Lead to Improved Federal Tax Return Accuracy and Provides a Possible Model for National Regulation: GAO-08-781: GAO Highlights: Highlights of GAO-08-781, a report to the Committee on Finance, U.S. Senate. Why GAO Did This Study: Millions of taxpayers use paid tax return preparers and many of these paid preparers are not subject to any qualification requirements. Paid preparers in California and Oregon are exceptions in that these states have set paid preparer qualification standards. Additionally, two bills before Congress would require national paid preparer regulations. To help Congress better understand the potential costs and revenue effects of regulating paid preparers, GAO was asked to study (1) how IRS, California, Oregon, and other states regulate paid preparers, (2) how the accuracy of federal tax returns from California and Oregon compare to other returns, and (3) state-level costs and benefits of the California and Oregon programs and insights they provide for a possible national program. GAO analyzed IRS research data on tax return accuracy; interviewed IRS officials, state administrators, and preparer community representatives; and reviewed relevant documents. What GAO Found: No federal registration, education, or testing requirements apply to all paid preparers before they can prepare tax returns. California and Oregon have requirements that preparers must meet before preparing returns in those states. California paid preparers who are not attorneys, certified public accountants, enrolled agents (or employed by one of these types of tax practitioners) must complete an education requirement, obtain a bond, pay a fee, and register. In following years, they must complete continuing education requirements, and renew their registration. Oregon has similar, but more stringent requirements. Oregon has a two-tiered licensing system, with an education requirement and examination for Licensed Tax Preparers and work experience and a second examination for Licensed Tax Consultants. Oregon exempts certified public accountants and their employees, as well as attorneys, from these requirements. Oregon requires enrolled agents to take a shorter version of the consultant examination. Fifty- four percent of Oregon applicants passed the state’s basic examination. Recently, Maryland enacted legislation to regulate paid preparers and at least three other states have similar pending legislation. According to GAO’s analysis of the Internal Revenue Service’s (IRS) tax year 2001 National Research Program data, Oregon returns were more likely to be accurate while California returns were less likely to be accurate compared to the rest of the country after controlling for other factors likely to affect accuracy. In dollar terms, the average Oregon return required approximately $250 less of a change in tax liability than the average return in the rest of the country. For Oregon’s 1.56 million individual tax filers, this equates to over $390 million more in federal income taxes paid in Oregon than would have been paid if the returns were as accurate as similar returns in the rest of the country. These results are consistent with, but do not prove, that Oregon’s regulations lead to some increased tax return accuracy. GAO’s analysis could not account for all factors that might affect the accuracy of these tax returns. Because some states without preparer regulation also had tax returns that, on average, were more accurate than the national average, some portion of the increased accuracy of Oregon returns likely is due to other factors. The California and Oregon programs’ costs varied with differences in the programs’ scope. Both programs’ administrative costs are funded primarily from program fees. California’s costs were about $29 per preparer and Oregon’s about $123. GAO estimates that the total annual cost of the ongoing Oregon program, including state costs and the cost to preparers for their time and expense in acquiring required education, likely is about $6 million. Officials in both states believe program benefits like reducing the number of incompetent preparers outweigh costs, although neither state had data on benefits. IRS officials said that a national program’s costs likely would depend on the program’s objectives and features. What GAO Recommends: If Congress judges that the Oregon paid preparer regulations account for even a modest portion of the higher accuracy of Oregon federal tax returns at a reasonable cost, it should consider adopting a similar regime nationwide. If Congress enacts paid preparer legislation, it should also require IRS to evaluate its effectiveness. IRS provided technical comments on a draft of this report which were incorporated. To view the full product, including the scope and methodology, click on [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-781]. For more information, contact Michael Brostek at (202) 512-9110 or brostekm@gao.gov. [End of section] Contents: Letter: Results in Brief: Background: IRS, California, and Oregon Differ Significantly in How Each Regulates Paid Tax Preparers: Oregon's Regulatory Regime May Lead to More Accurate Federal Tax Returns: Costs and Benefits of the California and Oregon Programs Provide Some Guidance for a National Program: Conclusions: Matter for Congressional Consideration: Agency Comments: Appendix I: Objectives, Scope, and Methodology: Appendix II: Comments from the Internal Revenue Service: Appendix III: GAO Contact and Staff Acknowledgments: Tables: Table 1: Overview of the California and Oregon Paid Tax Preparer Programs: Table 2: Comparison of State-Level Paid Preparer Requirements and Pending Legislation: Table 3: Percentages, Odds, and Odds Ratios for Return Accuracy, before and after Controlling for Other Factors: Abbreviations: CPA: Certified Public Accountant: CRTP: CTEC Registered Tax Preparer: CTEC: California Tax Education Council: EIC: Earned Income Credit: FTB: Franchise Tax Board: IRS: Internal Revenue Service: LTC: Licensed Tax Consultant: LTP: Licensed Tax Preparer: NRP: National Research Program: OBTP: Oregon Board of Tax Practitioners: OPR: Office of Professional Responsibility: [End of section] United States Government Accountability Office: Washington, DC 20548: August 15, 2008: The Honorable Max Baucus: Chairman: The Honorable Charles Grassley: Ranking Member: Committee on Finance: United States Senate: Nearly 78 million of the 127 million individual income tax returns filed during the 2006 filing season were prepared by paid tax return preparers.[Footnote 1] Paid preparers are such an important part of the federal tax administration system that the Internal Revenue Service (IRS) sometimes refers to them as "partners." However, we testified in 2006 on the serious errors that paid preparers can make--errors that cause taxpayers to underpay their taxes, exposing themselves to IRS enforcement action, or that lead taxpayers to not take advantage of available credits or deductions and, as a result, they end up paying too much.[Footnote 2] Tax return preparers may be self employed or may work in a variety of business settings, including large companies, franchises, and small businesses. Most paid preparers are not subject to any education, testing, or registration requirements. Two states, California and Oregon, are exceptions in that for many years they have had their own requirements that apply to paid preparers working in their states. To help Congress better understand the potential costs and revenue effects of establishing regulation at a federal level for all paid preparers, you asked us to answer the following questions: (1) How do IRS, California, Oregon, and other states regulate paid preparers? (2) Using available IRS data, how does the accuracy of federal tax returns in California and Oregon compare to that of returns in the rest of the country, after accounting for other factors that might influence accuracy? (3) What are the state-level costs and benefits of the paid preparer programs in California and Oregon and what insights do they provide for possible benefits if Congress were to enact national paid preparer registration or licensing requirements? To answer these questions, we reviewed relevant documents from California, Oregon, and IRS, including budget and legislative material. We interviewed California and Oregon state program administrators and paid preparer industry representatives in those states and nationwide. We also searched legal databases for examples of newly enacted paid preparer laws in other states and pending legislation. We also interviewed IRS officials to discuss the implications of using the California or Oregon regulatory regimes as possible models for federal- level paid preparer legislation. To compare tax return accuracy, we analyzed data from the National Research Program (NRP), an IRS study of reporting compliance for a random sample of individual tax returns filed for tax year 2001.[Footnote 3] In most cases, the returns were audited to determine whether income, expenses, and other items were reported accurately by the taxpayers. We determined that the data used to characterize tax return accuracy differences between California, Oregon, and the rest of the country and to describe the costs of the two state programs were sufficiently reliable for the purposes of this report; we determined this after assessing the reliability of NRP data, reviewing California and Oregon financial reports, and interviewing state program administrators. We conducted this performance audit from September 2007 through July 2008 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. For a more detailed discussion of our scope and methodology, see appendix I. Results in Brief: Oregon has more requirements for paid tax return preparers than California, and both states have more paid tax return preparer requirements than the federal government. Only a few federal laws apply to all paid preparers and these laws concern tax preparer conduct rather than qualification requirements. Only a small portion of paid preparers--enrolled agents[Footnote 4]--have any federal registration, testing, or fee requirements. California paid preparers who are not attorneys, certified public accountants (CPA), enrolled agents (or employed by one of these types of tax practitioners) must complete 60 hours of qualifying education, obtain a surety bond, register with the California Tax Education Council (CTEC), and pay a fee to become a CTEC Registered Tax Preparer (CRTP), and they must complete 20 hours of continuing education and reregister in each subsequent year. Paid preparers who fail to register can be fined up to $5,000. Oregon has a two-tiered licensing program. Oregon requires prospective paid preparers to complete 80 hours of qualifying education, pass a state- administered examination, register, and pay a fee to be initially certified as a Licensed Tax Preparer (LTP), and they must complete 30 hours of continuing education and pay a fee to reregister in each subsequent year. Oregon also requires that all LTPs work under the supervision of a Licensed Tax Consultant (LTC), CPA, or attorney. To become an LTC, a preparer must meet specific work experience requirements and pass a second, more advanced examination. Oregon can impose fines of up to $5,000 per return for unlicensed tax return preparation and for certain conduct on the part of LTPs and LTCs. The Oregon tests are notable in that they have low passing rates--54 percent for the LTP examination and 30 percent for the LTC examination. In May 2008, Maryland enacted legislation to regulate paid preparers and at least three other states have pending legislation to regulate paid preparers. When controlling for other factors likely to affect tax return accuracy, our analysis of IRS data showed that tax year 2001 federal tax returns filed in Oregon were more likely to be accurate than returns in the rest of the country, which is consistent with but not sufficient to prove that Oregon's regulatory regime leads to some increased tax return accuracy. On average, returns filed in California were less likely to be accurate than returns filed in the rest of the country.[Footnote 5] This indicates that California's paid preparer regulatory regime may not improve the likelihood that returns are accurate, relative to the rest of the country. Including both self- prepared and paid prepared returns, Oregon's 2001 federal returns were on average about $250 dollars more accurate than returns in the rest of the country. With about 1.56 million individual tax filers in Oregon in 2001, this translates into over $390 million more in income taxes paid in Oregon than would have been paid if Oregon returns were prepared at the level of accuracy seen on similar returns in the rest of the country. While some portion of this difference might be due to preparer regulations, we cannot rule out that other factors may influence accuracy, such as whether Oregon paid preparers were more likely to be attorneys or accountants than were paid preparers elsewhere in the country. Also, we cannot compare the before and after effects of either state's regulatory regime. Furthermore, some states without paid preparer regulation also had tax returns that, on average, were more accurate than the national average. Consequently, we cannot rule out the possibility that Oregon or California returns were no more or less likely to be accurate than they would have been without regulation of paid preparers. The costs and benefits of the California and Oregon programs vary in terms of scope and point to factors that would have implications for any proposed national paid preparer regulatory program. In both programs, direct administration costs are funded principally through fees with no direct cost to the states. California's program is focused on ensuring that paid preparers have received required education and its cost is relatively low, with direct costs of about $29 per CRTP per year, according to our analysis of the CTEC budget. Oregon's per paid preparer costs are higher than California's--about $123 per LTC and per LTP, according to our analysis of the Oregon Board of Tax Practitioners (OBTP) budget. Oregon's higher cost per preparer is partly because the Oregon program includes testing and also because Oregon's costs are spread among far fewer paid preparers than California's. Nevertheless, we conservatively estimate that the total cost of the Oregon program-- including both fees paid to the state and the time and expense that preparers incur to comply with Oregon's education requirements--was about $6 million in 2007. If only a small portion of the increased revenue that we found in Oregon is attributable to the Oregon regulatory regime, the regime would compare favorably to IRS's overall efforts to increase reporting accuracy. Program administrators and preparer community representatives we spoke to in both states said they believe that the programs are beneficial because they reduce the number of incompetent paid preparers, professionalize the industry, and have benefits that outweigh the costs, although neither state has conducted research into this latter question. Costs and benefits of paid preparer regulation at the federal level would similarly be driven by program features--the more a program would be expected to accomplish, the more it would likely cost to design, implement, and administer. If Congress judges that the Oregon paid preparer regulatory regime is likely to account for at least a modest portion of the higher accuracy of Oregon federal tax returns and could be implemented nationwide at a favorable cost compared to the potential benefits of improved accuracy, it should consider adopting a similar regime nationwide. In light of the uncertainty about the extent to which Oregon's regime improves tax return accuracy, if Congress enacts national paid preparer legislation, it should also require IRS to evaluate its effectiveness. We provided the Commissioner of Internal Revenue with a draft of this report for review and comment and IRS provided technical comments which we incorporated. The Commissioner's letter is reprinted in appendix II. Background: A paid tax return preparer is anyone who is paid to prepare, assist in preparing, or review a taxpayer's tax return.[Footnote 6] In this report, we refer to two categories of paid preparers--tax practitioners and unenrolled preparers. CPAs, attorneys, and enrolled agents are tax practitioners. Tax practitioners can practice before IRS; practicing before IRS includes the right to represent a taxpayer before the IRS, to prepare and file documents with IRS for the taxpayer, and to correspond and communicate with IRS. Individuals can become enrolled agents by passing a 3-part examination; IRS waives the examination requirement for people with specific prior work experience at IRS. Department of the Treasury Circular 230, Regulations Governing the Practice of Attorneys, Certified Public Accountants, Enrolled Agents, Enrolled Actuaries, and Appraisers before the Internal Revenue Service, applies to tax practitioners and governs their duties, restrictions, sanctions, and disciplinary proceedings. IRS's Office of Professional Responsibility (OPR) has responsibility for administering and enforcing Treasury Circular 230. We use the term unenrolled preparer to describe the remainder of the paid preparer population. In most states, anyone can be an unenrolled preparer regardless of education, experience, or other standards.[Footnote 7] Paid preparers are a critical part of the nation's tax administration system because of the wide variety of services they offer and their unique relationship with taxpayers. Paid preparers may combine several taxpayer services, including help understanding tax obligations, answering tax law questions, and providing tax forms and publications, return preparation, and electronic filing. IRS regards tax professionals as a critical link between taxpayers and the government. For example, IRS has a section of its Web site dedicated to providing information directly to tax professionals. IRS also sponsors the Nationwide Tax Forums, annual conferences in several cities every year to provide tax education to paid preparers. The Web site of the National Association of Tax Professionals also points out the shared responsibility of paid preparers to represent their clients while respecting the law, listing among its professional standards one that says "Should the client insist upon [an] item being stated on the return incorrectly, the member should withdraw and refuse to prepare the return." The number of active paid preparers is unknown. In 1999, IRS estimated there were up to 1.2 million paid preparers, but IRS officials acknowledge that the actual number could be significantly higher or lower. The total number of active paid preparers is unknown because only a small portion of all paid preparers--enrolled agents--are licensed directly by IRS to practice before the IRS. As of June 2008, about 43,000 tax preparers were actively enrolled to practice before the IRS.[Footnote 8] IRS officials said that the number of new enrolled agent applications and the number of people taking the examination have declined in recent years. They noted that these declines followed increases in enrolled agent application and examination fees.[Footnote 9] Similarly, the number of attorneys and accountants who make tax return preparation a part of their practice is unknown. Millions of tax returns prepared by paid preparers have serious compliance problems, which often leave taxpayers owing or overpaying by hundreds or thousands of dollars. As we have previously reported, [Footnote 10] IRS's tax year 2001 NRP data indicate that tax returns prepared by paid preparers had a higher error rate--56 percent- -than returns prepared by taxpayers--47 percent.[Footnote 11] In 2002, we estimated that on as many as 2.2 million tax returns, taxpayers claimed the standard deduction when their potential itemized deductions were greater, and that about half of these taxpayers had returns prepared by another person.[Footnote 12] In 2005, we reported that many tax returns included claims for one of three available postsecondary education tax preferences that resulted in higher overall tax liability than if one of the other preferences had been taken, and that over half of these returns were prepared by paid preparers.[Footnote 13] However, the fact that errors were made on a return done by a paid preparer does not necessarily mean the errors were the preparer's fault; the taxpayer may be to blame. The preparer must depend on the information provided by the taxpayer. On the other hand, some mistakes are clearly the fault of the preparer. In 2006 we reported on the results of an investigation where we identified mistakes in 19 out of 19 visits to paid preparers working in preparer chain offices. Some of the mistakes were significant, either exposing the taxpayers to serious IRS enforcement action or costing taxpayers over $1,500 in overpaid taxes.[Footnote 14] In 2007, the Department of Justice took action against corporations operating franchises of a major tax preparation chain. The government complaints alleged that the franchisee corporations created and fostered a business environment "in which fraudulent tax return preparation is encouraged and flourishes."[Footnote 15] The corporations that owned the franchises agreed to sell the franchises to new owners and to be permanently barred from preparing federal income tax returns.[Footnote 16] When mistakes or deliberate noncompliance by paid preparers result in taxpayers underreporting their tax liabilities, it adds to the tax gap. The net tax gap is an estimate of the difference between the taxes owed--including individual income, corporate income, employment, estate, and excise taxes--and what was eventually paid for a specific year. IRS most recently estimated the net tax gap to be $290 billion in 2001. In March 2008, we recommended that IRS develop a plan to require a single identification number for paid preparers, including assessing the feasibility of options, their benefits and costs, as well as their usefulness for enforcement and research, on paid preparer behavior. [Footnote 17] Also, as of July 2008 there were similar bills pending before Congress calling for national paid preparer regulation. Senate Bill 1219 and House of Representatives Bill 5716 would require members of the current community of unenrolled paid preparers to pass an initial qualifying examination and meet continuing annual education requirements. Support for legislation such as this can be found in the National Taxpayer Advocate's 2002 and 2003 Annual Reports to Congress, which recommended Congress create a designation called a "Federal Tax Return Preparer," defined as someone other than an attorney, CPA, or enrolled agent, who prepares more than five federal tax returns in a calendar year and satisfies registration, examination, and certification requirements.[Footnote 18] IRS, California, and Oregon Differ Significantly in How Each Regulates Paid Tax Preparers: Only a few Internal Revenue Code provisions apply to all paid preparers and only a small portion of paid preparers--enrolled agents--have any federal registration, testing, or fee requirements. All paid preparers are subject to a few Code provisions and may be penalized if they fail to follow them. For example, the Internal Revenue Code imposes monetary penalties on paid preparers who (1) understate a taxpayer's liability due to a position that fails to meet the applicable legal standard, (2) fail to provide a copy of the return to the taxpayer, or (3) fail to identify themselves on the returns they prepare. Additionally, for returns that include the Earned Income Credit (EIC), paid preparers must ask specific questions to determine a taxpayer's eligibility for the credit. Also, all paid preparers who choose to file electronically are subject to IRS Electronic Return Originator rules. Both California and Oregon began to regulate paid preparers in the 1970s. California's program was first administered by the state's Department of Consumer Affairs, and legislation transferred oversight responsibility to CTEC in 1997. Oregon's program was established by the 1973 Oregon Legislative Assembly after representatives of the state's paid preparer community recommended that the legislature regulate the profession. According to a preparer involved at the time, the Oregon Legislative Assembly was responding to a report that there were many dishonest or incompetent paid preparers working in the state. The main features of California's paid preparer program are qualifying and continuing education and registration. To become a CRTP, individuals initially register with CTEC by completing a 60-hour qualifying education course, purchasing a $5,000 surety bond, completing an application, and paying a $25 registration fee. CTEC may waive some of the qualifying education requirements for individuals with 2 recent years experience in the preparation of personal income tax returns.[Footnote 19] In each subsequent year, CRTPs must complete 20 hours of continuing education, ensure their bond remains in full force, submit a renewal application, and pay a $25 renewal fee. As of June 6, 2008, 41,755 paid preparers were registered with CTEC. CPAs, attorneys, enrolled agents, and employees of any of these types of tax practitioners are exempt and not required to register.[Footnote 20] California does not require prospective CRTPs to pass a criminal background check or to report past criminal convictions or current legal issues. This means that prior questionable or illegal conduct is not known to program administrators. Moreover, CTEC does not have the authority to deny a preparer's registration application based on known illegal conduct, nor does the California Code include provisions for refusing to renew a CRTP's registration as long as the CRTP meets the continuing education requirement and pays the annual registration fee. The 60-hour qualifying education requirement is intended to ensure paid preparers have a basic knowledge of federal and California tax laws. According to the CTEC policy manual, the intent of the annual continuing education requirement is to enhance the paid preparer's skill in tax matters above the basic knowledge they have already acquired. CTEC approves an education provider's curriculum based on an independent review of one of the prospective provider's courses at least once every 3 years. People who are not one of the types of exempt tax practitioners who prepare tax returns in California without becoming CRTPs can be fined. Under a Memorandum of Understanding between CTEC and the California Franchise Tax Board (FTB),[Footnote 21] the FTB is reimbursed by CTEC for providing staff to identify unregistered tax preparers. In 2007, FTB provided one full-time and one part-time employee and CTEC reimbursed FTB $270,000. Persons suspected of illegally preparing tax returns are first issued penalty letters and encouraged to become registered. If they do not register within 90 days, the FTB can levy fines of up to $5,000. An FTB official said that between July 1, 2005, and June 30, 2006, FTB identified 77 individuals as unregistered. [Footnote 22] Many of these persons were identified by the 2 FTB staff members who visited the Los Angeles and San Francisco Bay areas--where there are large numbers of paid preparer offices--met with paid preparers, and asked to see evidence of registration. Noncompliant paid preparers have also been identified through complaints sent to CTEC and passed along to FTB. Oregon requires paid preparers who are not already licensed by the state as CPAs or attorneys, or working for a CPA, to obtain a state license to prepare tax returns.[Footnote 23] Enrolled agents-- practitioners licensed by Treasury--must also obtain an Oregon license, but they are subject to fewer qualifying requirements than other individuals who are seeking an LTC license. The state board that administers the program--the Oregon Board of Tax Practitioners--issues two levels of paid preparer licenses: the Licensed Tax Preparer (LTP) license and the Licensed Tax Consultant (LTC) license.[Footnote 24] To become an LTP, a person must have a high school diploma or the equivalent, complete 80 hours of approved qualifying education, pass a state-administered examination with a score of 75 percent or better, and pay an $80 registration fee. To continue as an LTP in following years, individuals must annually renew their license by completing 30 hours of approved continuing education and paying an $80 renewal fee. An LTP in Oregon may only prepare tax returns for Oregon residents under the supervision of an LTC, CPA, or attorney.[Footnote 25] A person can become an LTC after working as a tax preparer for a minimum of 780 hours during 2 of the prior 5 years, completing a minimum of 15 hours of continuing education within 1 year of submitting an application, and passing a more advanced examination with a score of 75 percent or better.[Footnote 26] LTPs and LTCs must disclose on their initial license and license renewal applications if they have been convicted of a crime or are under indictment for criminal offenses involving dishonesty, fraud, or deception. According to the Oregon statute, OBTP can consider the circumstances in particular cases and still approve an application when the applicant has disclosed a legal issue. Many applicants do not pass the LTP or LTC examinations. For instance, from March 1, 2006, to February 28, 2007, 54 percent of test takers passed the LTP examination and 30 percent passed the LTC examination. [Footnote 27] The OBTP updates both examinations yearly. The examinations cover specific Oregon and federal personal income tax laws as well as tax theory and practice. The LTC examination also includes questions on corporation and partnership income as they relate to personal income tax returns. The examination questions pertain to approximately 75 percent federal and 25 percent state law. IRS enrolled agents in Oregon who wish to become LTCs must pass a shorter version of the LTC examination that is limited to Oregon state laws. The intent of Oregon's education and examination requirement is to ensure paid preparers comprehend the state and federal tax codes. OBTP reports that in March 2008, 3,993 paid preparers held one of these two licenses-- 1,916 LTPs and 2,077 LTCs. The Oregon statute includes fines for preparing tax returns without a license. Each return prepared can generate a separate fine, so the total penalty for working as an unlicensed preparer can be very large. OBTP also has the authority to assess civil penalties of up to $5,000, or suspend or revoke the license of LTCs and LTPs who engage in fraudulent or illegal conduct, or who violate other provisions of the Oregon statutes or OBTP rules. Additionally, the board may order restitution to consumers harmed by tax preparation fraud. From March 2001 to November 2007, OBTP took disciplinary action 48 times, with fines totaling about $2 million. The largest fine for one individual was in April 2002 for $805,700. Only a fraction of fines are eventually collected however--while about $867,000 in fines were levied from July 2005 through June 2007, about $69,000 in fines and $6,000 in interest was collected during the same period. Persons penalized by the OBTP can appeal these decisions and OBTP has an arrangement with the Oregon Office of Administrative Hearings to provide an administrative law judge to hear these cases. Individuals can also appeal their cases to the Oregon Court of Appeals. Both California and Oregon use their registered or licensed paid preparer lists to contact preparers to remind them about requirements and to inform them about changes to the tax code or other matters they should know about. However, neither state uses their preparer information to track paid preparer accuracy or for enforcement purposes. California does not require CRTPs to include their CTEC registration number on either the state or federal tax returns that they prepare. Oregon requires LTCs and LTPs to include their license number on both types of returns, but officials told us that this requirement is not consistently followed as some licensees incorrectly put down their Preparer Tax Identification Number, Social Security Number, or an employer's Employer Identification Number. Consequently, neither state has a reliable means to track or analyze returns prepared by registered or licensed paid preparers in their states. Table 1 illustrates some of the highlights of the California and Oregon regulatory programs. Table 1: Overview of the California and Oregon Paid Tax Preparer Programs: Requirement: Experience; California: May consider work experience in lieu of education; Oregon: Licensed Tax Preparer (LTP): None; Oregon: Licensed Tax Consultant (LTC): Prior experience as an LTP or submit petition form of all past tax preparation experience. Requirement: Education; California: Complete a 60-hour qualifying education course; Oregon: Licensed Tax Preparer (LTP): (1) Hold a high school diploma or pass equivalency exam; (2) Complete 80 hours of qualifying education; Oregon: Licensed Tax Consultant (LTC): Oregon: If currently an LTP, complete at least 15 hours of continuing education. Otherwise, complete 80 hours of education on income tax law. Requirement: Examination; California: None; Oregon: Licensed Tax Preparer (LTP): Pass exam with a score of at least 75 percent; Oregon: Licensed Tax Consultant (LTC): Pass exam with a score of at least 75 percent; Enrolled agents take only the sections of the LTC examination focused on Oregon laws. Requirement: Exempted individuals; California: CPAs, attorneys, enrolled agents, and anyone employed by them. Trust company and financial institution employees functioning within the scope of their employment; Oregon: CPAs, public accountants, and their employees; attorneys; employees of businesses who prepare only their businesses' tax returns; fiduciaries and their employees while acting on behalf of estates; and employees of governmental agencies while performing official duties. Requirement: Is criminal background relevant to registration or licensing? California: No; Oregon: Yes. OBTP makes case-by-case decisions. Requirement: Other; California: Purchase a $5,000 surety bond; Oregon: Must be 18 years old. Requirement: Fees; California: $25 (initial registration and annual renewal); Oregon: Licensed Tax Preparer (LTP): $80 (Initial issuance or renewal); Oregon: Licensed Tax Consultant (LTC): $95 (initial issuance and renewal), $65 (if currently an LTP). Requirement: Renewal; California: Annual. Complete 20 hours of continuing education and ensure bond remains in force; Oregon: Annual. Complete 30 hours of continuing education. Requirement: Penalties for failing to register; California: Unregistered individuals may be fined $2,500, but fine may be waived if they register within 90 days. If they fail to comply, the fine may be increased to $5,000; Oregon: Civil penalties range from $50 to $5,000 per violation. Sources: California and Oregon paid preparer regulatory programs. [End of table] In May 2008, Maryland also enacted paid preparer legislation that will require tax preparers to pass an examination, pay a registration fee, and subsequently comply with continuing education requirements. Also, New York, Oklahoma, and Arkansas all have legislation pending that would create tax preparer programs. All three pending bills create an oversight regime, which would include tax preparer registration and education requirements, both initial and continuing.[Footnote 28] The Oklahoma and Arkansas bills require that preparers pass an examination to register. Arkansas's pending legislation closely models the Oregon regime, with requirements for both preparers and consultants. New York's pending legislation is similar to California's paid preparer program, requiring preparers to maintain surety bonds but having no provision for preparer testing. The enacted Maryland program and the pending legislation in New York and Oklahoma exempt CPAs, attorneys and their employees, and enrolled agents from the requirements. The Arkansas bill would exempt CPAs and attorneys and their employees, and would require enrolled agents to pass a test only on Arkansas tax law issues.[Footnote 29] Table 2 provides an overview comparison of the California and Oregon requirements with the Maryland requirements and the pending legislation in the other states. Table 2: Comparison of State-Level Paid Preparer Requirements and Pending Legislation: Registration: California: Yes; Oregon: Yes; Maryland: Yes; Arkansas (pending legislation): Yes; New York (pending legislation): Yes; Yes; Oklahoma (pending legislation): Yes. Qualifying education: California: Yes; Oregon: Yes; Maryland: No; Arkansas (pending legislation): Yes; New York (pending legislation): Yes; Oklahoma (pending legislation): Yes. Testing: California: No; Oregon: Yes; Maryland: Yes; Arkansas (pending legislation): Yes; New York (pending legislation): No; Oklahoma (pending legislation): Yes. Continuing education: California: Yes; Oregon: Yes; Maryland: Yes; Arkansas (pending legislation): Yes; New York (pending legislation): Yes; Oklahoma (pending legislation): Yes. Sources: GAO review of state laws and pending legislation. [End of table] IRS officials noted that continued growth in the number of different paid preparer registration or licensing regimes in different states could become a problem if the requirements differ from state to state. The officials described this as primarily a problem for the tax preparation industry in that a variety of regulatory regimes across many different states could make it complicated, for example, for paid preparers to move their practice from one state to another or for a tax preparation chain to move employees or expand their operations. Oregon's Regulatory Regime May Lead to More Accurate Federal Tax Returns: When controlling for other factors likely to affect tax return accuracy, our analysis of IRS data showed that tax year 2001 federal tax returns filed in Oregon were more likely to be accurate than returns in the rest of the country, which is consistent with but not sufficient to prove that Oregon's regulatory regime improves tax return accuracy. Relative to the rest of the country, Oregon paid preparer returns had a greater likelihood of being accurate and California paid preparer returns were less likely to be accurate. Specifically, we found that the odds that a return filed by an Oregon paid preparer was accurate were about 72 percent higher than the odds for a comparable return filed by a paid preparer in the rest of the country. Conversely, the odds that a paid preparer return in California was accurate were about 22 percent lower than for paid preparer returns in the rest of the country.[Footnote 30] This indicates that California's paid preparer regulatory regime may not improve the likelihood that returns are accurate, relative to the rest of the country. Our analysis controlled for factors such as the complexity of tax returns in comparing California and Oregon to the rest of the country. However, our analysis cannot rule out the possibility that factors for which we could not control affected the accuracy of tax returns in either state. To determine the relative likelihood that Oregon and California returns were accurate, we used multivariate logistic regression to compare the odds of return accuracy in these states compared to odds in the rest of the country, controlling for other characteristics that might influence return accuracy.[Footnote 31] To make these accuracy comparisons, we used data from IRS's NRP, which assessed the accuracy of individual tax returns from tax year 2001. We defined a return as accurate if it required less than $100 absolute value in changes.[Footnote 32] As an illustration of the differences among paid preparer returns in California and Oregon, we computed the probability of accuracy for a medium complexity, form 1040, U.S. Individual Income Tax Return, for a taxpayer with income over $100,000. While a return with these characteristics prepared by a paid preparer in Oregon would have a 74 percent probability of being accurate, a similar return prepared by a paid preparer in California would have a 55 percent probability of being accurate. In addition to having a higher likelihood of accuracy than the rest of the country, on the average Oregon 2001 federal tax return--regardless of whether it was self prepared or from a paid preparer--auditors identified a smaller increase in taxes owed.[Footnote 33] In Oregon, the average return required approximately $250 less of a change in tax liability than the average return in the rest of the country. Our $250 estimate is conservative in that it does not incorporate the limited number of cases with relatively large liability changes. With about 1.56 million individual tax filers in Oregon in 2001, this translates into over $390 million more in income taxes paid in Oregon than would have been paid if Oregon returns were prepared at the level of accuracy seen on similar returns in the rest of the country. The average tax liability change in California was higher than the average in the rest of the country by approximately $90. Although the differences we observed in the states' regulatory programs and in how likely California and Oregon returns were to be accurate compared to the rest of the country are consistent with the Oregon regime leading to some improved federal tax return accuracy, the analysis cannot rule out that the regime did not have such an effect. We could not control for other factors that may influence accuracy, such as whether Oregon paid preparers were more likely to be attorneys or CPAs than preparers elsewhere in the country. Also, data are not available on return accuracy prior to the existence of each state's program, so we cannot compare the before and after effects of the regimes. Before and after data might have shown, for instance, whether the California regime leads to improved tax return accuracy compared to what it otherwise would have been even though California's returns in 2001 were less accurate, on average, than returns in the rest of the country. Also, we considered the accuracy of tax returns in other states and found that some states without paid preparer laws had more accurate tax returns than the national average, after controlling for the factors in our model.[Footnote 34] This indicates that regulation over paid preparers alone does not explain the differences that we found. Further, to the extent that the Oregon regime does improve tax return accuracy, our methodology does not identify whether any part of the regime is most important to that result. Our methodology only takes into account the entire regimes as implemented in Oregon and California. Costs and Benefits of the California and Oregon Programs Provide Some Guidance for a National Program: Both California and Oregon support their programs almost entirely through fees, with state program costs averaging about $29 and $123 per year, respectively, per registered paid preparer. In addition to the fees charged to paid preparers, the preparers or their employers bear other costs, such as those associated with taking courses on tax law and return preparation. Program administrators and preparer community representatives in both states said that there are intangible benefits from their regulatory regimes, although there are no studies quantifying outcomes in either place. The California and Oregon paid preparer registration programs include differing design features, such as on testing applicants and how much enforcement is deemed desirable, that show, not surprisingly, that more extensive programs cost more. California's Less Extensive Program Costs Less Than Oregon's: California's paid preparer program is more limited in scope than Oregon's, and has lower direct administration costs per registered preparer. Because neither state provides funding for the programs above the fees collected, the entire cost of both programs are borne directly or indirectly by the regulated paid preparer communities. As noted previously, California's program primarily requires unenrolled preparers to register with the state and meet minimum education requirements. The total direct budgeted cost of the California program was about $1.2 million in fiscal year 2007, with most of the funding coming from the $25 registration fees that CRTPs must pay, with additional funds coming from late registration fees and other income such as fees paid by education providers that apply to be approved as CTEC education providers.[Footnote 35] CTEC's total budget in 2007 was $1.2 million and CTEC reported 41,755 CRTPs in June 2008, so the cost per CRTP was about $29. According to CTEC officials, no funds from state tax revenues are used to pay for administering or enforcing California's paid preparer laws. Like California, Oregon also registers preparers and seeks to ensure that paid preparers meet minimum education requirements, but it also tests prospective LTPs and LTCs, adding to the administration cost of the Oregon program. In 2008, prospective LTPs pay $50 and prospective LTCs pay $85 to take the examinations. Also as of 2008, LTPs pay $80 and LTCs pay $95 to obtain their initial license and in each subsequent year to renew their license. The registration fee for a new LTC who had been an LTP is $65. OBTP also collects fines and penalties from both unlicensed tax return preparers and licensed paid preparers who violate Oregon laws--averaging about $38,000 per year in the 2005 through 2007 period. OBTP's administrative expenses amounted to about $490,000 in 2007--divided by the 3,993 LTCs and LTPs OBTP reported in March 2008, this is about $123 per licensee.[Footnote 36] According to OBTP officials, OBTPs operating funds come from the fees and fines described above and none come from the state's general revenues. Administrative functions of CTEC and OBTP include communicating with paid preparers and the public at large about their regulations, informing the paid preparer community about tax law and processing changes, evaluating education providers, recordkeeping related to registration and licensing, maintaining a Web site that taxpayers can use to find a paid preparer or check that a particular paid preparer is properly registered or licensed, and working with the state legislature and the rest of the state government. Some of the difference in the administrative cost per registered or licensed preparer between the two states may be attributed to economies of scale in the registration of paid preparers that California has relative to Oregon. While California's direct operating budget is about twice the size of Oregon's, the number of preparers that it registers is more than 10 times greater. Enforcement-related expenses take up a share of the CTEC and OBTP budgets. In California, CTEC paid the FTB $270,000 in fiscal year 2007 to conduct enforcement targeted at identifying unregistered preparers and either bringing them into compliance or fining them. CTEC is not involved in imposing fines on unregistered preparers and has no means of taking enforcement action against a CRTP for misconduct, and it has never incurred litigation expenses associated with someone appealing a CTEC decision. In Oregon, the OBTP has a full-time investigator on its staff and directly imposes fines on both licensed and unlicensed paid preparers for misconduct. As discussed previously, these fines can be appealed, so OBTP arranges with the Oregon Office of Administrative Hearings for an administrative law judge to hear cases, and reimburses the Oregon Attorney General's Office for counsel to handle legal aspects of disputed cases. In 2007, OBTP expenses for its investigator and costs related to litigation were about $93,000.[Footnote 37] The regulatory programs in the two states impose additional costs beyond the direct administration expenses found in the CTEC and OBTP budgets. In both states, prospective paid preparers must meet qualifying education requirements and the financial and time costs of obtaining this education are directly borne by either the individual or his or her employer. We contacted frequently used education providers in both states and found costs were typically in the $200 to $300 range, although one was $614. According to paid preparers we spoke to, the cost of obtaining continuing education was sometimes fairly low, especially when continuing education was obtained through participation in professional associations. In some associations, monthly meetings usually include a presentation that qualifies for continuing education credit. Other preparers, however, may choose to travel to conferences or training sessions, such as an IRS Nationwide Tax Forum, to obtain their continuing education over just a few days. The registration fee for the IRS forums is fairly low--$179 for early registration in 2008. Out-of-town travel, when necessary, adds to the cost of obtaining required continuing education. Continuing education can also be obtained from state-approved education providers in both classroom settings and over the Internet. Because results for the Oregon regime are consistent with some positive effect on federal tax return accuracy, the cost of that regime is of particular interest. We conservatively estimated the total costs associated with Oregon's regulation to be about $6 million in 2007. This estimate includes the regime's direct administrative costs as well as an estimate of the cost of licensees obtaining qualifying and continuing education from education providers, the value of the time they spend in those classes and studying outside of class, and the same education-related costs for all unsuccessful test takers. This estimate is conservative because it counts preparer education time and expense for all licensees, including enrolled agents, who have continuing education requirements under that program, and employees of tax preparation chains that require similar education for all of their preparers. Appendix I describes how we made our estimate. IRS has developed rough measures of return on investment in terms of tax revenue that it assesses from uncovering noncompliance. Generally, IRS cites an average return on investment for enforcement of 4:1, that is, IRS estimates that it collects $4 in revenue for every $1 of funding.[Footnote 38] For the Oregon paid preparer regulatory regime to be considered a reasonably cost-effective tax administration policy by this standard, it would have to account for only a small share of the $390 million in higher federal tax revenue we estimated came in from Oregon compared to the rest of the country.[Footnote 39] It is important to note that the 4:1 IRS average return is based on administrative spending and such expenses are less than 10 percent of our approximately $6 million annual total cost estimate for the Oregon program. Regulation of preparers can also have the effect of increasing the price of tax preparation services by reducing the supply of paid preparers. A California tax preparer association representative said that the costs to obtain and maintain CRTP status are fairly low and likely do not have much of an impact on prices consumers pay, and that the requirements to become a paid preparer are not so great that the number of paid preparers in the state is being held lower than it would be without any regulation. In Oregon, however, direct costs to become a paid preparer and to maintain licensed status are somewhat higher. Potentially more important, however, is the requirement that LTPs only work in offices supervised by an LTC, attorney, or CPA, and that LTCs may not supervise more than two offices. This means that there can be a substantial bar to the opening of a new tax preparation business if the owner cannot find and recruit an LTC. We were told by a representative of a tax preparation chain that he had experienced difficulty in opening a new rural office because he could not find an LTC to supervise LTPs. However, since there are somewhat more LTCs in Oregon than LTPs, such problems may be limited.[Footnote 40] Data that could be used to analyze prices charged by paid preparers in California or Oregon, or to compare prices charged in those states with the rest of the country, are not available. NRP data, however, provide a related point of comparison on the use of paid preparers. NRP data show that taxpayers in Oregon are somewhat less likely to use a paid preparer than taxpayers in the rest of the country and even less likely to use paid preparers than taxpayers in California. NRP data show that about 58 percent of individual taxpayers used paid preparers nationally, while only 49 percent of Oregon taxpayers did so. About 64 percent of California tax returns were prepared by paid preparers. It is possible that the Oregon regulatory regime has had the effect of reducing the supply of paid preparers, leading to an increase in the price charged for the service. California and Oregon Officials Consider Their Programs to Be Beneficial: Program administrators and preparer community representatives in both California and Oregon described their programs as having benefits that outweigh their costs. Officials in both states also said they believe that paid prepared tax returns are more accurate due to their paid preparer regulatory regimes. However, neither California nor Oregon program administrators have analyzed tax returns to see if this is the case. Representatives also noted that registration facilitates communication with paid preparers that are registered or licensed, so notifying them about, for example, recent changes in tax rules or forms, can be done fairly easily. Program administrators and paid preparer community representatives in California and Oregon also told us education requirements likely reduce the number of incompetent paid preparers and have led to a more professional tax preparation industry. California and Oregon program administrators also said that consumers benefit from the ability to go online and verify whether a paid preparer is registered or licensed. Both state programs also give taxpayers the ability to seek restitution when wronged by a paid preparer[Footnote 41]. A benefit of the Oregon program is that prospective preparers who cannot pass the state examination are not allowed to prepare tax returns in that state. As noted previously, the Oregon LTP examination has only a 54 percent passing rate. This means that many people who want to become paid preparers but lack the knowledge and skills necessary to pass the Oregon exam are not legally preparing tax returns. People in every other state with a similar desire to become a paid preparer--and a similar lack of skill--are presumably preparing tax returns. Occupational licensing of other professions has been shown to have costs and benefits to the consumer[Footnote 42]. As with other markets for services, licensing paid preparers might be expected to have several potential effects depending on how licensing requirements are designed. Depending on the level of education or expertise required to obtain a license, some preparers who become licensed may acquire additional knowledge, which helps them better prepare returns or expand their expertise to additional types of returns. In Oregon, officials said that they believe unlicensed tax preparers cost the consumer money when they prepare incorrect or inaccurate tax returns. Occupational licensing of other professions suggests that taxpayers may be willing to pay more to have their returns prepared by registered or licensed paid preparers if the regulatory requirements (i.e., education requirements) provide greater assurance of a higher quality prepared return. Consumers who continue to use these paid preparers may benefit as a result and some taxpayers who previously self prepared their own returns may switch to a licensed or registered preparer because of additional assurance of quality service. On the other hand, if the licensing requirements cause some preparers to no longer offer services, prices may rise and some taxpayers may switch to self preparation. Implications for a National Regulatory Program: The California and Oregon paid preparer regulation programs provide reference points for national policymakers when considering a national paid preparer regulatory regime. In both cases, program costs are driven by the scope of the program. As with the differences we identified in California and Oregon, a more extensive national program will likely cost more to administer than a less extensive one. An additional point of comparison for policymakers considering a potential national paid preparer program is IRS's enrolled agent program. Enrolled agents are paid preparers who are permitted to represent their clients in matters before IRS. Enrolled agents have to either pass a 3-part examination covering individual income taxes, business taxes and representation, and practices and procedures, or have specific IRS experience.[Footnote 43] During the period May 2007 through April 2008, the overall passing rate for the three parts of the examination was 48 percent.[Footnote 44] Prospective enrolled agents also have to meet continuing education requirements and pay a $125 registration fee every 3 years. One area in which the enrolled agent program parallels the two state programs we studied is that the examination is handled through a contract that is of no direct cost to the government. A private company developed the tests and administers them at sites around the country and it is compensated entirely through fees of about $100 that test takers pay to take each part of the 3-part examination. Most of the test taking fee is retained by the contractor, but $11 is remitted to IRS. Applicants are also required to allow IRS to conduct a background check.[Footnote 45] IRS officials in OPR said that the more a national program is expected to accomplish, the more expensive it will likely be to design, implement, and administer.[Footnote 46] Enforcement is a key consideration, as even the fairly modest enforcement efforts in the two states we reviewed took up 19 percent of total administrative costs in Oregon and 23 percent in California. IRS officials said that more extensive enforcement nationwide could be very costly. IRS officials said they have not developed specific costs for a national regime, in part because they are uncertain which of the many potential elements the program would include. Conclusions: The California and Oregon regulatory regimes point to the feasibility of a nationwide regulatory regime involving paid preparer education, registration, and, as in Oregon's case, testing. Both states have enacted registration and other requirements while funding the administration of their programs through relatively modest fees paid by paid preparers, similar to the way that IRS sees to the testing of enrolled agents. A key benefit from the Oregon approach is the apparent rigor of its qualifying examinations. Just under half of the people who take the Oregon LTP examination fail to pass. These people are not legally preparing tax returns in Oregon today, at least not until they are able to pass the examination. Paid preparers with an equivalent lack of demonstrated ability may well be working as paid preparers in other states. Available data do not conclusively support or refute the idea that adopting some or all of the California or Oregon program elements at the national level would improve the accuracy of paid prepared returns or reduce the tax gap. However, the more stringent requirements of the Oregon regime along with our modeling results suggest that an Oregon- style approach to paid preparer regulation may be beneficial. The higher level of accuracy found on Oregon returns meant $390 million more in income taxes paid in Oregon than would have been paid if Oregon returns were as accurate as returns everywhere else. The cost of the Oregon program is quite small in comparison, about $490,000 per year in administrative expenses and an estimated total of about $6 million after including the time and expense associated with paid preparers meeting their education and testing requirements. If only a small share of the increased revenue is attributable to the Oregon regulatory regime, it would compare favorably to IRS's overall efforts to increase reporting accuracy. With over half of individual taxpayers using paid preparers, it may be possible to make meaningful progress towards narrowing the tax gap by requiring all paid preparers to demonstrate competence before being allowed to prepare other people's tax returns. However, because the extent, if any, to which the Oregon regulatory regime improves federal tax return accuracy, is uncertain, if a similar regulatory regime is adopted at the federal level, its effect on tax return accuracy should be assessed. Because IRS has resumed periodic studies of tax return accuracy, such a study could compare accuracy of returns before and after implementation of a federal regime. Matter for Congressional Consideration: If Congress judges that the Oregon paid preparer regulatory regime is likely to account for at least a modest portion of the higher accuracy of Oregon federal tax returns and could be implemented nationwide at a favorable cost compared to the potential benefits of improved accuracy, it should consider adopting a similar regime nationwide. In light of the uncertainty about the extent to which Oregon's regime improves tax return accuracy, if Congress enacts national paid preparer legislation, it should also require IRS to evaluate its effectiveness. Agency Comments: In a letter commenting on a draft of this report dated August 1, 2008, the Commissioner of Internal Revenue noted the important role that paid preparers play in supporting a fair, efficient, and effective system of tax administration. His letter also notes IRS's strategy of working with paid preparers and curbing abuses by unscrupulous preparers. IRS also provided technical comments which we incorporated. The Commissioner's letter is included in appendix II. As agreed with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days after its date. At that time, we will send copies of this report to the Secretary of the Treasury, the Commissioner of Internal Revenue, and other interested parties. This report is available at no charge on GAO's Web site at [hyperlink, http://www.gao.gov]. If you or your staff have any questions about this report, please contact me at (202) 512-9110 or brostekm@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix III. Signed by: Michael Brostek: Director, Tax Issues Strategic Issues Team: [End of section] Appendix I: Objectives, Scope, and Methodology: Our objectives were to answer the following questions: (1) How do IRS, California, Oregon, and other states regulate paid preparers? (2) Using available IRS data, how does the accuracy of federal tax returns in California and Oregon compare to that of returns in the rest of the country, after accounting for other factors that might influence accuracy? (3) What are the state-level costs and benefits of the paid preparer programs in California and Oregon and what insights do they provide for possible benefits if Congress were to enact national paid preparer registration or licensing requirements? To answer the first and third objectives we conducted a literature review of both the California and Oregon paid preparer programs, including a review of applicable laws and budget documents. We also interviewed state program administrators from the California Tax Education Council and the Oregon Board of Tax Practitioners (OBTP); officials from the California Franchise Tax Board and the Oregon Department of Revenue; and leaders in each state's paid preparer community, and reviewed documents provided to us by them. At the federal level, we reviewed appropriate legislation concerning the regulation of paid preparers, interviewed IRS officials, primarily from the Office of Professional Responsibility, and reviewed documents related to the enrolled agent program. We also interviewed and obtained data from an official from Prometric, the company IRS contracted with to develop and administer the enrolled agent examinations. We interviewed the National Taxpayer Advocate and members of her staff concerning her prior recommendations to regulate paid preparers. We also met with a representative from the National Association of Enrolled Agents to understand their perspective on a more expansive national regulatory regime. Finally, we conducted a literature review of professional occupational regulation to understand the potential effects of occupational regulation on the paid preparer profession. In identifying nonfederal paid preparer regulation programs, we limited our review to state governments and requirements concerning qualification, registration, or licensing of paid preparers and we did not consider possible county or city regulations, or laws dealing with paid tax return preparer conduct. For the discussion of costs and benefits from the Oregon program in the third objective, we also used information from the OBTP about program costs and the number of new and returning licensees in 2007. We obtained information from education providers about the fees that they charge for basic and continuing education. We also used the U.S. Bureau of Labor Statistics national average hourly wage for paid tax return preparers--$16.78 in 2007--the value of the time spent obtaining the education. Using this information, we developed an estimate of the total cost of the Oregon program. In considering costs to include, we included higher-end estimates where possible to ensure that our estimate of the total cost of the Oregon program was conservative. For example, we did not consider the fact that many Oregon licensees are employed by a national tax preparation chain that requires its paid preparers to receive initial and continuing education, so they would be obtaining that education regardless of the Oregon laws. To answer the second objective, we analyzed data from IRS's National Research Program (NRP). The NRP contains detailed tax and audit data from approximately 47,000 randomly selected tax year 2001 returns, and includes extensive compliance data including line-by-line estimates of accuracy.[Footnote 47] Unlike other compliance-related data sets, NRP data are generalizable to the population of individual taxpayers throughout the U.S. While NRP was not designed for specific state-level analysis, in conjunction with IRS's NRP officials, we agreed on the types of analysis that the data would support and which variables could be used. Our analysis comprised four main steps, each of which is explained in more detail below. We first examined the odds that returns from different locations and using different preparation types were accurate. Next, we considered the relative likelihood that a return was accurate, prior to controlling for other factors. Additionally, recognizing that Oregon and California differ from the rest of the country in terms of factors potentially related to a return's accuracy, we developed multivariate statistical models to assess whether returns from these states were more or less likely than returns from other states to require liability changes of $100 or more in absolute value after controlling for other factors. We also assessed differences in the accuracy of self-prepared tax returns. Finally, we estimated potential cost savings using multivariate regression analysis to assess the size of average tax liability changes for Oregon or California returns relative to the returns in the rest of the United States, controlling for other factors.[Footnote 48] In creating our statistical models, we examined a variety of variables on the basis of previous research, our reports, and recommendations from NRP personnel. Our final model included measures of the complexity of the return,[Footnote 49] including whether it was for a sole proprietor or claimed the Earned Income Credit (EIC). We also included the examination class of the return,[Footnote 50] taxpayer adjusted gross income in quartiles, whether the return was e-filed, filing status, and a proxy for a state's aggregate level of English proficiency.[Footnote 51] All models were calculated using sampling weights and robust estimation to account for differential variation among returns in distinct sampling strata. Logistic Regression: Table 3 illustrates differences in likelihood that returns from different locations and using different preparation types were accurate. Column A of table 3 shows that, prior to controlling for other factors, 54 percent of California returns and 71 percent of Oregon returns were accurate compared to 64 percent of returns in the rest of the United States. On average, 58 percent of paid preparer returns were accurate, compared to 70 percent of self-prepared returns. The lower half of table 3 illustrates the combined effect of location and preparation status. Prior to controlling for other factors, 49 percent of California paid preparer returns and 67 percent of Oregon paid preparer returns were accurate, compared to 59 percent of paid preparer returns in the rest of the country. Similarly, without controlling for other factors, 63 percent of California self-prepared returns and 75 percent of Oregon self-prepared returns were accurate, compared to 71 percent of self-prepared returns in the rest of the country. The odds within each category, shown in column C, compare the proportion of returns that were accurate to the proportion of returns that were not accurate. Table 3: Percentages, Odds, and Odds Ratios for Return Accuracy, before and after Controlling for Other Factors: Location: California (average); Percentage of accurate returns (A): 54.1; Percentage of nonaccurate returns (B): 45.9; Odds of accuracy C = (A/B): 1.18; Unadjusted odds ratio prior to controlling for other factors (D): .66[A]; Adjusted odds ratio controlling for other factors (E): .81[A]. Location: Oregon (average); Percentage of accurate returns (A): 71.0; Percentage of nonaccurate returns (B): 29.0; Odds of accuracy C = (A/B): 2.45; Unadjusted odds ratio prior to controlling for other factors (D): 1.37[A]; Adjusted odds ratio controlling for other factors (E): 1.54[A]. Location: Rest of U.S.; Percentage of accurate returns (A): 64.2; Percentage of nonaccurate returns (B): 35.8; Odds of accuracy C = (A/B): 1.79; Unadjusted odds ratio prior to controlling for other factors (D): [B]; Adjusted odds ratio controlling for other factors (E): [B]. Preparation type: Paid preparer (ave); Percentage of accurate returns (A): 58.0; Percentage of nonaccurate returns (B): 42.0; Odds of accuracy C = (A/B): 1.38; Unadjusted odds ratio prior to controlling for other factors (D): .58[A]; Adjusted odds ratio controlling for other factors (E): .89[A]. Preparation type: Self prepared (ave); Percentage of accurate returns (A): 70.3; Percentage of nonaccurate returns (B): 29.7; Odds of accuracy C = (A/B): 2.37; Unadjusted odds ratio prior to controlling for other factors (D): [B]; Adjusted odds ratio controlling for other factors (E): [B]. Location by preparation type: Paid Preparer Returns; California paid preparer; Percentage of accurate returns (A): 49.4; Percentage of nonaccurate returns (B): 50.6; Odds of accuracy C = (A/B): .98; Unadjusted odds ratio prior to controlling for other factors (D): .67[A]; Adjusted odds ratio controlling for other factors (E): .78[A]. Location by preparation type: Paid Preparer Returns; Oregon paid preparer; Percentage of accurate returns (A): 67.1; Percentage of nonaccurate returns (B): 32.9; Odds of accuracy C = (A/B): 2.04; Unadjusted odds ratio prior to controlling for other factors (D): 1.41; Adjusted odds ratio controlling for other factors (E): 1.72[A]. Location by preparation type: Paid Preparer Returns; Rest of U.S. paid preparer; Percentage of accurate returns (A): 59.2; Percentage of nonaccurate returns (B): 40.8; Odds of accuracy C = (A/B): 1.45; Unadjusted odds ratio prior to controlling for other factors (D): [B]; Adjusted odds ratio controlling for other factors (E): [B]. Location by preparation type: Self-prepared returns; California self- prepared; Percentage of accurate returns (A): 62.5; Percentage of nonaccurate returns (B): 37.5; Odds of accuracy C = (A/B): 1.67; Unadjusted odds ratio prior to controlling for other factors (D): .68[A]; Adjusted odds ratio controlling for other factors (E): .85. Location by preparation type: Self-prepared returns; Oregon self- prepared; Percentage of accurate returns (A): 74.8; Percentage of nonaccurate returns (B): 25.2; Odds of accuracy C = (A/B): 2.97; Unadjusted odds ratio prior to controlling for other factors (D): 1.21; Adjusted odds ratio controlling for other factors (E): 1.29. Location by preparation type: Self-prepared returns; Rest of U.S. self- prepared; Percentage of accurate returns (A): 71.1; Percentage of nonaccurate returns (B): 28.9; Odds of accuracy C = (A/B): 2.46; Unadjusted odds ratio prior to controlling for other factors (D): [B]; Adjusted odds ratio controlling for other factors (E): [Empty]. Sources: GAO analysis of IRS's NRP data. [A] Indicates statistical significance at the 95 percent confidence level. The NRP sample is only one of an infinite number of samples that could have been selected to represent the population of taxpayers in the U.S. Statistical significance at the 95 percent level indicates that there is less than a 5 percent chance we would have gotten a result of this magnitude if there were no actual difference between the group of interest and the reference category in the population. [B] Indicates referent category of self-prepared returns and/or returns in the rest of the United States. [End of table] For the next step, we used odds ratios to compare the relative likelihood that returns from different locations or of different preparation types were accurate. The unadjusted odds ratio in column D compares the odds of return accuracy in each specific subgroup to a reference group, prior to controlling for other factors. An odds ratio of 1 illustrates that on average, returns for the two groups have the same odds of being accurate, while odds ratios above 1 indicate a higher likelihood of accuracy and odds ratios below 1 indicate a lower likelihood of accuracy. Column D of table 3 illustrates that, prior to controlling for other factors, California returns on average had lower odds of accuracy than returns in the rest of the country, by a factor of .66 (34 percent lower). Conversely, Oregon returns on average had higher odds of accuracy than the rest of the country, by a factor of 1.37 (37 percent), before we account for other factors that might influence accuracy. This pattern holds when we compare returns using different preparation methods to similarly prepared returns. For example, California paid preparer returns have odds of accuracy approximately 33 percent lower than paid preparer returns in the rest of the country, and Oregon paid preparer returns have odds that are 41 percent higher than similarly prepared returns in the rest of the country, before controlling for other factors. These unadjusted odds do not control for other factors that might differentiate between returns in Oregon and California compared to those in the rest of the country. However, descriptive data reveal that the characteristics of returns filed in California and Oregon differ from the characteristics of returns filed in the U.S. as a whole. For example, a greater proportion of Oregon and California residents file sole proprietor returns than in the U.S., on average. To control for potential differences that might influence the likelihood of filing an accurate return, we used multivariate logistic regression. These models enabled us to compare the adjusted odds of accuracy for returns from Oregon or California with returns in the rest of the country, holding constant the effect of other factors that could affect accuracy. Column E in the upper half of table 3 shows that the odds of accuracy for an average Oregon return were still higher when compared to the rest of the country, and the odds of accuracy for a California return were still lower, after controlling for other factors. Additionally, paid preparer returns, on average, had lower odds of accuracy than self-prepared returns, controlling for other factors including location. As we note previously, not all mistakes on paid prepared tax returns are the fault of the paid preparer. The results for all returns in the upper half of table 3 treat location and preparation type as distinct factors, without considering potential interaction between location and preparation type. To ensure that these estimates did not mask compliance differences between paid preparer and self-prepared returns and to assess the potential impact of regulation on the population directly affected by the regime (paid preparers), we also examined self-prepared and paid preparer returns separately (see the lower half of table 3). These models reveal pronounced effects among paid preparers, after controlling for other factors. Among paid preparer returns, Oregon returns had odds of accuracy 72 percent higher, and California returns had odds of accuracy 22 percent lower, than comparable paid preparer returns in the rest of the country. While self-prepared returns in California had lower odds of accuracy than self-prepared returns in the rest of the country, and Oregon returns had higher odds of accuracy after controlling for other factors, these results were not statistically significant at the 95 percent level. Our estimates of the impact of location on the likelihood that a return was accurate had fairly wide confidence intervals. One reason for this is due to our inability to incorporate the full range of individual or state-level factors that might influence the likelihood of compliance, such as whether a paid prepared return was prepared by an attorney or CPA. Additionally, the NRP sample was designed for purposes other than to compare states, which resulted in wider confidence bounds than would a sample designed specifically for state-level estimates.[Footnote 52] Our analyses identified several factors other than location that influenced the likelihood that a return would require less than $100 in liability changes, both among returns in general and the subpopulation of paid preparer returns. For example, the odds that a return claiming the EIC was accurate were less than half those of returns that did not claim the EIC in all models.[Footnote 53] Similarly, sole proprietor returns (those individual returns that had an attached Schedule C, Profit or Loss from Business) had lower odds of being accurate than other returns. Additionally, returns with a filing status of "married, filing separately" were significantly less likely to be accurate than returns in any other filing status. Overall, 1040 forms with total positive incomes of less than $100,000 had higher odds of accuracy compared to form 1040 returns with total positive income of $100,000 or above. Conversely, among forms with total positive income of $100,000, forms 1040F, Profit or Loss from Farming, and 1040C, U.S. Departing Alien Income Tax Return, were less likely to be accurate. In general, e- filed returns had slightly lower odds of accuracy than paper returns. In addition to our main logistic regression model, we conducted a series of alternative analyses to examine the impact of location and paid preparer status with additional control factors and alternative dependent variables, and found results generally consistent with the models presented in table 3. These included several models with and without various aggregate state factors (such as per capita income and whether a state had an income tax), with alternative measures of complexity (including one based on the number of schedules filed), and with a dummy variable for returns that were software generated but not e-filed.[Footnote 54] Finally, we examined alternative dependent variables, including tax liability changes prior to EIC and additional child credits, and the net sum of dollar values of line item adjustments for each return. These additional analyses give us confidence that our results are robust to a variety of model specifications and different definitions of accuracy. Cost Savings: To identify potential cost savings from an Oregon-style regulatory regime, we used multivariate linear analysis to assess the size of average tax liability changes among all returns, controlling for other factors. We conducted diagnostic analysis to identify and exclude outliers and potentially high-leverage cases--individual cases that have the potential to disproportionately affect our estimate when compared to other cases. Our estimate of savings is thus conservative when compared to an analysis that includes all cases, as it does not incorporate the savings generated by a limited number of cases with relatively large liability changes. After controlling for the other factors described, we found that the average return in Oregon required significantly lower changes in tax liability than returns in California or the rest of the country. The average Oregon return required tax liability increases that were approximately $250 lower than comparable returns in the rest of the country. In contrast, the average California return required tax liability increases that were approximately $90 higher than returns in the rest of the country, controlling for other characteristics. We conducted this performance audit from September 2007 through July 2008 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. [End of section] Appendix II: Comments from the Internal Revenue Service: Department Of The Treasury: Commissioner: Internal Revenue Service: Washington, D.C. 20224: August 1, 2008: Mr. Michael Brostek: Director, Tax Issues: U.S. Government Accountability Office (GAO): 441 G Street, N.W. Washington, D.C. 20548: Dear Mr. Brostek: Thank you for giving us the opportunity to comment on the draft Government Accountability Office (GAO) report titled "Tax Preparers: Oregon's Regulatory Regime May Lead to Improved Federal Tax Return Accuracy and Provides a Possible Model for National Regulations" (GAO- 08-781). According to Internal Revenue Service (IRS) data, the majority of individual taxpayers used a return preparer in 2006. The IRS recognizes the critical role of these preparers in supporting a fair, efficient, and effective system of tax administration. Our strategy is to enhance service to and collaboration with return preparers; ensure coordinated and consistent oversight to curb abuses by unscrupulous preparers; and identify and provide new tools to preparers. If you have any questions, or would like to discuss in more detail, please contact me or Michael Chesman, Director, Office of Professional Responsibility, at (202) 927-3397. Sincerely, Signed by: Douglas H. Shulman: [End of section] Appendix III: GAO Contact and Staff Acknowledgments: GAO Contact: Michael Brostek, (202) 512-9110 or brostekm@gao.gov: Acknowledgments: In addition to the contact person named above, David Lewis, Assistant Director; Crystal Bernard; Amy Bowser; James Cook; John Mingus; Ed Nannenhorn; Karen O'Conor; and Anna Maria Ortiz made key contributions to this report. [End of section] Footnotes: [1] GAO, Tax Administration: 2007 Filing Season Continues Trend of Improvement, but Opportunities to Reduce Costs and Increase Compliance Should be Evaluated, [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO- 08-38] (Washington, D.C.: Nov. 15, 2007). [2] GAO, Paid Tax Return Preparers: In a Limited Study, Chain Preparers Made Serious Errors, [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO- 06-563T] (Washington, D.C.: Apr. 4, 2006). [3] The results of the 2001 NRP are the most recent IRS compliance research data available. [4] Enrolled agents are allowed to represent a taxpayer before the IRS, to prepare and file documents with the IRS for the taxpayer, and to correspond and communicate with the IRS. Individuals can become enrolled agents by passing a 3-part examination; IRS waives the examination requirement for people with specific prior work experience at IRS. [5] We categorize a return as "accurate" if the IRS examination found that it required a change of tax liability of less than $100 in absolute value. [6] See 26 U.S.C. § 7701(a)(36). [7] Regulation of unenrolled preparers is the principal focus of this report. [8] Enrolled agents must complete 72 hours of continuing education and renew their registration every 3 years. [9] IRS officials said that the number of applications for enrollment was 3,108 in fiscal year 2006 and 1,916 in fiscal year 2007. In fiscal year 2005 and 2006, there were over 11,000 candidates taking the examinations per year and 5,847 did so in fiscal year 2007. Enrollment fees increased from $80 to $125 in fiscal year 2007. Total examination fees increased from $55 to $291 in 2006. [10] GAO, Paid Tax Return Preparers: In a Limited Study, Chain Preparers Made Serious Errors, [hyperlink, http://www.gao.gov/cgi- bin/getrpt?GAO-06-563T] (Washington, D.C.: Apr. 4, 2006). [11] All percentage estimates from the NRP files have margins of error of plus or minus 5 percentage points or less, unless otherwise noted. All numerical estimates other than percentages have margins of error of plus or minus 5 percent or less of the value of those numerical estimates, unless otherwise noted. [12] GAO, Tax Deductions: Further Estimates of Taxpayers Who May Have Overpaid Federal Taxes by Not Itemizing, [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-02-509] (Washington, D.C.: Mar. 29, 2002). [13] GAO, Student Aid and Postsecondary Tax Preferences: Limited Research Exists on Effectiveness of Tools to Assist Students and Families through Title IV Student Aid and Tax Preferences, [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-05-684] (Washington, D.C.: July 29, 2005). [14] GAO, Paid Tax Return Preparers: In a Limited Study, Chain Preparers Made Serious Errors, [hyperlink, http://www.gao.gov/cgi- bin/getrpt?GAO-06-563T] (Washington, D.C.: Apr. 4, 2006). [15] Department of Justice, U.S. Government Sues Jackson Hewitt Tax Preparation Franchises in Four States, Alleging Pervasive Fraud (Apr. 3, 2007), available at [hyperlink, http://www.usdoj.gov/tax/txdv07215.htm]. [16] Department of Justice, Corporations That Owned Jackson Hewitt Franchises in Three States Agree to be Barred from Tax Return Preparation (Sept. 28, 2007), available at [hyperlink, http://www.usdoj.gov/tax/txdv07779.htm]. [17] GAO, Internal Revenue Service: Fiscal Year 2009 Budget Request and Interim Performance Results of IRS's 2008 Tax Filing Season, [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-567] (Washington, D.C.: Mar. 13, 2008). [18] The National Taxpayer Advocate, National Taxpayer Advocate--FY 2002 Annual Report to Congress (Washington, D.C.: Dec. 31, 2002) and National Taxpayer Advocate--2003 Annual Report to Congress (Washington, D.C.: Dec. 31, 2003). [19] The 2 years of experience can be time spent preparing tax returns in another state or while working for an attorney, CPA, or enrolled agent. It may not include time preparing tax returns in violation of the registration requirement. [20] Trust company and financial institution employees functioning within the scope of their employment are also exempt from the registration requirements. [21] The California FTB is responsible for administering the state's personal income and corporate tax. [22] Of the 77 individuals identified, 56 registered within the 90-day period and were not fined. The other 21 were fined $2,500. Of the 21 who were fined, 11 registered in the next year and were not subject to any additional penalties. Six of the 21 did not register and were issued the $5,000 penalty. The remaining 4 were no longer preparing returns. [23] Public accountants and their employees, employees of businesses who prepare only their businesses' tax returns, fiduciaries and their employees while acting on behalf of estates, and employees of governmental agencies while performing official duties are also exempt from Oregon's licensing requirements. [24] Tax preparation businesses operating in Oregon must also register with OBTP. As of February 1, 2008, the annual tax preparation business registration fee was $110. [25] The laws applicable to paid preparers do not apply to attorneys, CPAs, and the employees of CPAs. However, an LTP working under the supervision of a CPA or attorney must still follow the applicable paid preparer laws because the LTP has chosen to be licensed by the OBTP. [26] Continuing education may be accepted for up to 260 hours of work experience at the rate of 1 hour of education for 5 hours of work experience provided the course is tax related, taken within 1 year of applying to become an LTC, and credit for the course is not claimed to fulfill continuing education requirements for a license renewal. [27] All passing rate figures are for the 12-month period beginning March 1, 2006. The 30 percent passing rate for the LTC examination is an overall figure for both the full examination and the state-law-only portion of the examination given to enrolled agents. The passing rate for the full LTC examination is 25 percent and the state-law-only portion is 71 percent. [28] We limited our search for enacted laws and pending legislation to those concerning paid preparer qualifications and did not search for pending or enacted legislation concerning paid preparer conduct. Also, our search may not have identified all recent activity in states aside from the states we found. [29] The United States territory of Guam also has a tax preparer program that requires all paid preparers to pass an examination, register with Guam's Department of Revenue and Taxation, and maintain a surety bond. [30] The bounds of our estimates for how Oregon compared to the rest of the country are relatively wide. For example, the 95 percent confidence interval for our model among paid preparer returns suggests that the odds of accuracy among Oregon returns are higher than those for returns in the rest of the country by somewhere between 5 percent to 181 percent. [31] The odds of accuracy are defined as the percentage of returns that are accurate over the percentage that are inaccurate in each category. The ratio of odds for one group (e.g., Oregon) to another group (the rest of the country) helps to illustrate the relative likelihood of accuracy. For an illustration of how odds ratios are calculated, see appendix I. [32] We also tested alternative dependent variables such as liability changes over $10 in value and whether the net value of line item adjustments exceeded $99. The results were largely consistent with our model using the $100 liability threshold change. [33] Computing accuracy for all Oregon returns takes into account that if the Oregon paid preparer regime decreases the likelihood of noncompliance for paid prepared returns, those wishing to be noncompliant might switch to preparing their own returns. Because Oregon self-prepared returns were no less accurate than returns elsewhere in the country, even if this switching occurred it likely would not completely offset the increased accuracy of paid prepared returns. [34] States besides Oregon with a statistically significant likelihood of having paid preparer returns that were more accurate than the national average, controlling for other factors, were Colorado, Iowa, New Mexico, Ohio, Pennsylvania, West Virginia, and Wisconsin. [35] CTEC's fiscal year operates from July 1 to June 30. [36] OBTP's biennial fiscal years 2005 through 2007 budget was about $980,000. [37] This includes about $29,000 in legal fees billed to OBTP by the Oregon Department of Justice. According to OBTP, these were mostly associated with enforcement actions, but also included some non- enforcement-related matters. [38] GAO, Tax Compliance: Multiple Approaches Are Needed to Reduce the Tax Gap, [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-07-488T] (Washington, D.C.: Feb. 16, 2007). [39] At $6 million in total cost, the Oregon regulatory regime would have to account for only about 6 percent ($23.4 million) of the $390 million in higher federal tax return accuracy to compare favorably to IRS's estimated overall 4:1 return on investment. [40] An alternative to finding an LTC to supervise LTPs is to hire a CPA. Any individual employed by a CPA in Oregon may prepare tax returns, whether an LTP or not. [41] CRTP clients in California can make claims against the surety bonds that CRTPs are required to obtain. LTP and LTC clients can make complaints to OBTP, and OBTP can order restitution along with fines and penalties. [42] Morris M. Kleiner, Licensing Occupations: Ensuring Quality or Restricting Competition, W.E. Upjohn Institute (Kalamazoo, Michigan, 2006) summarizes the results of several studies on the effects of licensing on quality of service, prices, and earnings for workers in different service markets, including teachers, dentists, lawyers, and optometrists. [43] An enrolled agent applicant who is requesting enrollment based on former employment with IRS must have had (1) a minimum of 5 years continuous employment with IRS during which the applicant must have been regularly engaged in applying and interpreting the provisions of the Internal Revenue Code and the regulations relating to income, estate, gift, employment, or excise taxes, or (2) an aggregate of 10 or more years of employment in positions involving the application and interpretation of the provisions of the Internal Revenue Code, at least 3 of which occurred within the 5 years preceding the date of application. [44] Between May 2007 and April 2008, 1,856 of 4,844 attempts (38 percent) at Part 1 of the examination were successful, as were 1,558 of 3,438 attempts (45 percent) at Part 2, and 1,777 of 2,591 attempts at Part 3 (69 percent). [45] IRS data do not permit comparison of return accuracy by type of paid preparer. [46] OPR establishes and enforces standards of competence, integrity, and conduct for enrolled agents, attorneys, CPAs, and other individuals and groups covered by IRS Circular 230. [47] More accurate returns that result in higher revenues collected than less accurate returns are the measure of societal benefit that we considered for purposes of this report. [48] We define tax liability as taxes owed after accounting for the Earned Income Credit and the additional child tax credit. [49] Our measure of complexity is a three-point scale based on research presented by John Guyton, Karen Masken, and Mark Mazur at the 2007 National Tax Association Conference on Taxation. [50] The examination class is defined by the income reported on the return and, for sole proprietors or farm owners, the gross receipts of the return. [51] All models were calculated using sampling weights and robust estimation to account for potential correlation between returns in the same sampling stratum. We used likelihood ratio tests and Aikake's Information Criterion when deciding on a final model specification. [52] The design effect, which compares the effect of a complicated sample design compared to a simple random sample, helps to illustrate the impact of the NRP sampling design on state-level estimates. These design effects indicate that standard errors for estimates of the effect of being in California and Oregon were more than 2 ½ times what we would expect to see from a random sample. Large standard errors make it more difficult to detect statistical significance. [53] Controlling for other factors, paid preparer returns claiming the EIC had odds of accuracy 76 percent lower than that of non-EIC paid preparer returns, whereas self prepared returns claiming the EIC had odds 68 percent lower than those of non-EIC self-prepared returns. [54] We could not find written IRS guidance on how to interpret the flag for computer-generated returns. Although IRS staff confirmed that the flag was distinct from the e-filing code, we found some overlap in the NRP data. The variable did not consistently improve model fit when added to the model described above. [End of section] 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 Web site [hyperlink, http://www.gao.gov]. Each weekday, GAO posts newly released reports, testimony, and correspondence on its Web site. To have GAO e-mail you a list of newly posted products every afternoon, go to [hyperlink, http://www.gao.gov] and select "E-mail Updates." Order by Mail or Phone: The first copy of each printed report is free. Additional copies are $2 each. A check or money order should be made out to the Superintendent of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a single address are discounted 25 percent. Orders should be sent to: U.S. Government Accountability Office: 441 G Street NW, Room LM: Washington, D.C. 20548: To order by Phone: Voice: (202) 512-6000: TDD: (202) 512-2537: Fax: (202) 512-6061: To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: E-mail: fraudnet@gao.gov: Automated answering system: (800) 424-5454 or (202) 512-7470: Congressional Relations: Ralph Dawn, Managing Director, dawnr@gao.gov: (202) 512-4400: U.S. Government Accountability Office: 441 G Street NW, Room 7125: Washington, D.C. 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, D.C. 20548: