WRITTEN STATEMENT OF

STATEMENT OF

THE HONORABLE J. RUSSELL GEORGE

TREASURY INSPECTOR GENERAL FOR TAX ADMINISTRATION
before the
U.S. SENATE COMMITTEE ON FINANCE

SUBCOMMITTEE ON TAXATION AND IRS OVERSIGHT

“A Closer Look at the Size and Sources of the Tax Gap”

July 26, 2006

Introduction

Chairman Kyl, Ranking Member Jeffords, and Members of the Subcommittee, I appreciate the opportunity to appear before you today to discuss the size and sources of the tax gap, as well as some opportunities for closing the tax gap.

The objective of our tax system is to fund the cost of government operations. The Internal Revenue Service (IRS) attempts to meet this objective by implementing a tax system that provides adequate funding for the Federal Government and is fairly applied to all taxpayers. But, as we know, the system has failed to capture a significant amount of the tax revenue that is owed, which we call the tax gap. The IRS defines the tax gap as “the difference between what taxpayers are supposed to pay and what is actually paid.”[1]

It is worth noting, that if we were to capture the estimated annual tax gap, it would offset the projected fiscal year (FY) 2006 budget deficit of $296 billion and give us a surplus of approximately $50 billion. Because the tax gap poses a significant threat to the integrity of our voluntary tax system, one of my top priorities for TIGTA is to identify opportunities for improvements to the IRS’ tax compliance initiatives.

Similar to nearly all other Federal agencies, the IRS has limited resources to apply to the objectives it seeks to achieve. Nevertheless, the IRS must face the challenge of trying to increase voluntary compliance and reduce the tax gap.

When I testified on the tax gap last year, I reported that some of the most challenging barriers to closing the tax gap are tax law complexity, incomplete information on the tax gap and its components, and reduced IRS enforcement resources. These same barriers exist today. To an extent, a portion of the tax gap can be closed through more effective enforcement and a commitment of additional enforcement resources. A significant portion of the gap, however, may not be amenable to traditional examinations and audits. Other means might better address that portion, such as tax law simplification and increased third-party reporting. Some of TIGTA’s more significant findings and recommendations to improve tax administration and help the IRS reduce the tax gap are presented later in my testimony.

The Tax Gap: Its Size and Sources

The IRS describes the tax gap as having three primary components — unfiled tax returns, taxes associated with underreported income on filed returns, and underpaid taxes on filed returns.[2] Within the underreported income component, the IRS has further delineated specific categories of taxes, such as individual, corporate, employment, estate, and excise taxes.

In 2006, the IRS updated its estimate of the tax gap, which had been based on data for tax year (TY) 1988. The new estimate was based on data obtained from the National Research Program (NRP) for TY 2001 individual income tax returns. [3] Data from the NRP were used to update the 2001 tax gap figures. The IRS developed a chart called the tax gap map to graphically depict the gross tax gap for TY 2001, its components, and their relative sizes. The first iteration of the map was completed prior to the first phase of the NRP.

Figure 1: IRS Tax Gap Map Before the National Research Program

Source: IRS

The map attributes various certainties to the tax gap estimates representing the IRS’ confidence in the figures based on the quality and age of the estimates. Subsequent to the NRP, the map was updated with preliminary figures in March 2005. After refinement of the NRP data, a third iteration was issued in February 2006. Figure 2 shows the most recent version of the tax gap map.

Figure 2: IRS Tax Gap Map After the National Research Program

Source: IRS

As shown in the preceding tax gap maps, the IRS’ preliminary TY 2001 gross tax gap estimate for individual, employment, corporate, and other taxes was $310.6 billion with a voluntary compliance rate (VCR) of 85.1 percent. However, the most recent gross tax gap estimate is $345 billion with a VCR of 83.7 percent based on updates using the NRP data.

In any discussion about whether a specific VCR goal can be met, the logical starting point would be an assessment of the reliability of the measurement data. In April 2004, Senator Baucus called for a 90 percent VCR by the end of the decade. Based upon the best information the IRS had available as of February 2006, the gross tax gap for TY 2001 was approximately $345 billion and the VCR was approximately 83.7 percent. Assuming the current IRS tax gap and VCR were complete and accurate, the 90 percent compliance target would present major challenges to tax administration. For example, assuming that in TY 2010 the total tax liability is the same as it was in TY 2001, to reach a level of 90 percent voluntary compliance, noncompliant taxpayers would have to timely and voluntarily pay an additional $134 billion.[4]

In April 2006, my staff reported results of a review to determine whether the IRS’ compliance efforts and strategies will enable it to achieve a 90 percent VCR by 2010.[5] In all three compliance areas across the major tax gap segments, TIGTA has concerns about whether the tax gap projections are complete and accurate. TIGTA’s primary concerns are described in the areas of nonfiling, underreporting, and estimated payments that result in the difference between the gross and net tax gaps.[6]

Nonfiling

Prior to the NRP, the IRS estimated the nonfiling gap to be $30.1 billion, which was composed of $28.1 billion of individual income tax and $2 billion of estate taxes. In February 2006, the IRS updated this estimate to $25 billion for individual income tax. The individual estimate was based on data provided by the U.S. Census Bureau. However, there are supplemental data that suggest substantial amounts are not included in the tax gap estimates. For example, the tax gap maps describe the nonfiling estimate as reasonable despite the missing segments of corporate income, employment, and excise taxes. The IRS does not have definite plans to update the estate tax segment[7] or to estimate the corporate, employment, and excise tax nonfiler segments, suggesting that the nonfiler estimate is incomplete and likely inaccurate.

In July 2004, researchers in the Small Business/Self-Employed (SB/SE) Division issued a report on business nonfilers recommending implementation of an enhanced system for creating and selecting inventory. Subsequently, the SB/SE Division Research office developed a prototype that matched $4.6 trillion in transactions to over one-half of the business nonfilers for TY 2002, detecting approximately $1 trillion of apparent taxable income. This fact alone brings the $27 billion individual and estate nonfiling estimate into question and demonstrates the need for more research to better estimate nonfiling for all tax segments.

Underreporting

The tax gap attributed to underreporting is by far the largest identified portion of the tax gap at an estimated $285 billion. Yet, TIGTA concluded that this estimate may not be complete since there are at least four areas that suggest substantial amounts are not included in the tax gap map projections.

· First, the business income portion of the individual underreporting tax gap estimate is incomplete because it lacks information from another NRP study that the IRS is undertaking on flowthrough returns[8] of Subchapter S corporations.[9] The study, which began in October 2005, will take 2 years to 3 years. Thus, the information from these audits was not available for the February 2006 updated TY 2001 tax gap estimates. Over 2.9 million Subchapter S Corporation returns were filed in TY 2001 with more than 5.3 million shareholders reporting $187.7 billion in net income.

· Second, the revised tax gap map lists the underreporting gap at $5 billion for small corporations and $25 billion for large corporations. These amounts are essentially carryovers from the previous estimate, including their weaker certainty status, since no new information was developed. For small corporations, the estimate is based on the 1980 TCMP survey. For large corporate underreporting, the previous estimates were not based on random TCMP audits but on operational audit coverage from the mid-1980s. These projections assume constant VCRs, yet current experience suggests compliance may not be constant. For example, in 2003, an IRS contractor estimated that the yearly tax gap arising from abusive corporate tax shelters alone was between $11.6 billion and $15.1 billion.[10]

· Third, the revised map similarly categorized as reasonable a $4 billion figure for estate taxes and provided no estimate for excise taxes, yet the estate tax estimate was not updated during the current NRP. In addition, there are no firm plans for further studies or updates of these components.

· Fourth, for the employment tax component, the combined $15 billion Federal Insurance Contributions Act and unemployment tax gap figure was also carried over and will not be further studied. Most of the employment tax component consists of self-employment tax. Yet, similar to the business income portion of the individual income tax gap, this too is incomplete without the flowthrough data.

Payments collected

The IRS’ tax gap maps, both before and after the NRP, list $55 billion as recoveries or enforced collections and other late payments.[11] This figure does not represent an actual amount but is an estimate projected from historical information and formulas based on what is known about the amount of collections on accounts over time. However, TIGTA found the actual basis of these formulas to be very limited, as well as dated. According to IRS officials, these formulas were developed “quite some time ago.” Thus, these formulas most likely do not take into account changes in the IRS’ ability to collect revenue.

To determine the validity of the potential $55 billion in collections, TIGTA requested data from the IRS on actual collections for TY 2001 by year of collection. These collections have two basic components: voluntary payments received by the IRS after the due date and payments received by the IRS as a result of some type of IRS intervention. The IRS, however, does not currently correlate either type of payment to the applicable tax year. Consequently, the IRS has no means of determining whether the $55 billion is ever collected. While the IRS is currently developing a way to associate collections resulting from enforcement actions to the related tax years, no similar data are being developed for voluntary late payments. Unless the latter data are similarly correlated, the IRS will be unable to determine actual collections or an accurate net tax gap.

Methodology

Beyond the concerns about the completeness and accuracy of the current tax gap estimates, TIGTA has additional concerns about the samples and adjustments made to the estimates. Both sampling and adjustments can have a significant effect on the confidence that can be placed in the estimates.

A significant concern about sampling for the NRP studies is whether the sample sizes would affect the usefulness or, perhaps, the accuracy of the data. Performing the studies depends on limited audit resources. IRS research officials described the problem as acute for the current flowthrough study. The most recent TCMP survey of partnerships, for example, involved a sample size that was significantly larger than the present normal, or operational, audit coverage of partnership returns. That sample size was sustainable at the time of the TCMP survey when the operational audit program was much larger, but it is not feasible now. The same was true of the last TCMP survey of Subchapter S corporation returns. The NRP office proposed much smaller sample sizes for its current surveys of flowthroughs, not because the same sample sizes are no longer needed but because the resources necessary to sustain the required number of random audits are not available.

Although IRS Research officials told TIGTA that the sample size was not an issue for the individual income tax study, the size of the calibration sample[12] for the individual study was constrained by fewer resources than what would have been preferred. An IRS consultant’s evaluation of the NRP’s sample design expressed concern that using the calibration audits to adjust the NRP could introduce considerable error into the adjusted estimates. My staff consulted with one researcher who believed the appropriate size of the original NRP sample was an open question that needs addressing for future work. Furthermore, as a result of software issues and anomalies in the data, only about 37,000 of the original 46,000 cases in the original sample were used in the most recent tax gap estimate.

TIGTA is also concerned about the estimate’s use of multipliers. Multipliers are essentially methods used to estimate undetected additional taxable income. Using multipliers complicates the confidence in, and precision of, the estimates.

In summary, much of the information remains dated, the new information is incomplete in several respects, and methodology differences create challenges. Considering this, a somewhat different picture of the tax gap map emerges. My staff concluded that, despite the significant efforts undertaken in conducting the individual taxpayer NRP, the IRS still does not have sufficient information to completely and accurately assess the overall tax gap and the VCR. Although having new information about TY 2001 individual taxpayers is better when compared to the much older TY 1988 information from the last TCMP survey, some important individual compliance information remains unknown. Additionally, although individuals comprise the largest segment of taxpayers and were justifiably studied first, no new information about employment, small corporate, large corporate, and other compliance segments is available. With no firm plans for further studies or updates in many areas of the tax gap, both the underreporting tax gap and the nonfiling gap will indefinitely leave an unfinished picture of the overall tax gap and compliance.

Achieving Targeted Voluntary Compliance Goals

While TIGTA has concerns about the overall reliability of the tax gap projections, the annual amounts collected that reduce the net tax gap, and the VCR, my staff determined that it was instructive to analyze what additional amounts the IRS would have had to collect to reach 90 percent voluntary compliance at different estimated intervals for TY 2001. Figure 3 shows the range for TY 2001 based upon the total tax liability for TY 2001 as estimated in February 2006. The IRS has proposed in the FY 2007 budget that the VCR will be raised from 83.7 percent to 85 percent by 2009. Accordingly, if the total tax liability remained constant, the IRS would have to collect, on a voluntary and timely basis, $28 billion more in TY 2009, thus reducing the gross tax gap to $317 billion. To reach 90 percent voluntary compliance by TY 2010,[13] the amount voluntarily and timely collected for TY 2010 would be an additional $134 billion, thus reducing the gross tax gap to $211 billion if the total tax liability remained constant.

Figure 3: Additional Voluntary and Timely Payments
Required to Reach Specified VCR Levels[14]

Source: Treasury Inspector General for Tax Administration

Barriers Hampering Compliance Research

Performing a compliance measurement program is expensive and time consuming. The estimated cost for performing the TY 2001 individual taxpayer NRP was approximately
$150 million. According to IRS officials, resource constraints are a major factor in NRP studies and affect how often the NRP is updated. Operational priorities must be balanced against research needs. From FY 1995 through FY 2004, the revenue agent workforce declined by nearly 30 percent while the number of returns filed grew by over 9 percent. This shortfall in examiner resources makes conducting large-scale research studies problematic.

The IRS’ budget submission to the Department of the Treasury (the Department) for
FY 2007 requests funding to support ongoing NRP reporting compliance studies. The IRS Oversight Board[15] supports ongoing dedicated funding for compliance research. Unfortunately, funding for those resources in previous fiscal years did not materialize. Without a resource commitment for continual updating of the studies, the information will continue to be stale and less useful in measuring voluntary compliance.

Learning From Previous Attempts to Reduce the Tax Gap

In June 1993, IRS executives met with Treasury and Office of Management and Budget (OMB) officials to discuss key issues for the FY 1995 budget. The issues facing the Federal Government at that time were similar to the current issues: severe fiscal constraints and the desire for good tax administration. Consequently, both the Department and OMB agreed to work with the IRS on a comprehensive plan to reduce the tax gap.[16] The IRS formed a task group that performed an extensive review of the tax gap.[17] The resulting task force report addressed the major areas of the tax gap and provided recommendations. The report concluded that:

· Enforcement is the most costly option and delivers only limited revenue.

· Methods to increase voluntary compliance are less costly but more burdensome to taxpayers.

· Legislative changes are needed as the primary means to increase compliance levels.

· The TCMP surveys can be used to identify the types of noncompliance but not the causes.

· The IRS needs to reevaluate its media and taxpayer education efforts.

· The tax gap needs to be treated as a multibillion dollar market, and efforts need to be made to capture as much of that market as possible.

· The IRS needs to consider making a high-level official responsible for overseeing efforts to close all components of the tax gap.

· The IRS Strategic Plan needs to be modified to more closely align with the tax gap components.

TIGTA was unable to determine if the report’s recommendations were implemented.

Opportunities for Closing the Tax Gap

Although better data will help the IRS identify noncompliant segments of the population, broader strategies and better research are also needed to determine what actions are most effective in addressing noncompliance. The IRS must continue to seek accurate measures of the various components of the tax gap and the effectiveness of actions taken to reduce it. This information is critical to the IRS for strategic direction, budgeting and staff allocation. The Department also needs these measures for tax policy purposes. Additionally, Congress needs this information to develop legislation that improves the effectiveness of the tax system.

Recommendations on how to address closing the tax gap have been circulating for many years. Some of those recommendations, made over 15 years ago, are still relevant today. I would like to focus on the following opportunities that TIGTA, other oversight groups, and interested stakeholders have identified to address the tax gap:

· Reduce the Complexity of the Tax Code.

· Gather Better Compliance Data.

· Refine Compliance Strategies.

o High-Income Taxpayers.

o Abusive Tax Shelters.

o Information Reporting on Sales of Investments.

o Withholding on Non-employee Compensation.

o Document Matching.

o Late Filed Returns.

o Coordinated National Nonfiler Strategy.

· Increase Resources in the IRS Enforcement Functions.

Reduce the Complexity of the Tax Code

The topic of tax law complexity generally evokes calls for tax law simplification. Government, academic and technical studies suggest a strong correlation between tax law complexity and tax law compliance. The greatest case for the correlation is that complexity allows legal tax avoidance, which at times can evolve into illegal tax evasion. The argument continues that because of tax law complexity, it is often difficult to ascertain whether a taxpayer has intentionally evaded taxes, or whether there was an honest misunderstanding. Therefore, the IRS use of punitive penalties must be tempered to ensure taxpayers are not penalized for honest misunderstandings.

In 2001, the Joint Committee on Taxation conducted a study on the complexity of the tax law. The Committee found that, at that time, the tax code consisted of nearly 1.4 million words. There were 693 sections of the code applicable to individuals, 1,501 sections applicable to businesses, and 445 sections applicable to tax exempt organizations, employee plans, and governments. At that time, a taxpayer filing an individual income tax return (Form 1040) could be confronted with a 79 line return, 144 pages of instructions, 11 schedules totaling 443 lines (including instructions), 19 separate worksheets embedded in the instructions, and the possibility of having to file numerous other forms.[18]

The President’s Advisory Panel on Federal Tax Reform also cited tax code complexity as a significant problem.[19] Among others, sources of complexity include duplicative and overlapping provisions, phase-outs, and expiring provisions. In addition, the panel cited the instability of the tax code as another source of complexity. Since 1986, there have been more than 14,400 changes to the code. This complexity is costing the U.S. economy $140 billion each year, with taxpayers spending over 3.5 billion hours preparing tax returns, and more than 60 percent of them now rely on a tax practitioner to prepare their tax returns.

One of the major effects attributed to tax law complexity is that it causes lower voluntary compliance. According to the American Institute of Certified Public Accountants, tax law complexity:

· Increases perceptions that the tax system is unfair.

· Increases costs for tax administration and tax compliance.

· Decreases the quality of tax administration and tax assistance.

· Increases the number of inefficient economic decisions.

Although it is believed that tax law simplification would increase voluntary compliance, there are significant factors that suggest simplification will be difficult to achieve throughout the Internal Revenue Code. The Joint Committee on Taxation identified various sources of complexity, and no single source was primarily responsible. The sources identified were:

· A lack of clarity and readability of the law.

· The use of the Federal tax system to advance social and economic policies.

· Increased complexity in the economy.

· The interaction of Federal tax laws with State laws, other Federal laws and standards (such as Federal securities laws, Federal labor laws and generally accepted accounting principles), the laws of foreign countries, and tax treaties. [20]

The lack of clarity and readability of the law results from:

· Statutory language that is, in some cases, overly technical and, in other cases, overly vague.

· Too much or too little guidance with respect to certain issues.

· The use of temporary provisions.

· Frequent changes in the law.

· Broad grants of regulatory authority.

· Judicial interpretation of statutory and regulatory language.

· The effects of the congressional budget process.

A major cause of tax law complexity is tax expenditures. The Congressional Budget and Impoundment Control Act of 1974 defines tax expenditures as “revenue losses attributable to provisions of the Federal tax laws which allow a special exclusion, exemption, or deduction from gross income or which provide a special credit, a preferential rate of tax, or a deferral of tax liability.”[21] While there is disagreement over whether the tax expenditure concept accurately measures “tax subsidies,”[22] the methods used to target provisions to specific taxpayer groups does cause tax law complexity.

Experts in tax policy maintain that any tax system will have complexity. Therefore, even though many people believe that tax simplification could provide the impetus for increasing voluntary compliance, a simple tax system could be a very difficult goal to achieve given the complexities of our society and multiple uses of the Internal Revenue Code. Thus, closing the tax gap through tax simplification and eliminating tax expenditures could prove to be challenging. But, to the extent that the tax law can be simplified, most experts believe that voluntary compliance would improve.

Another effective method to increase voluntary compliance might be through greater transparency of transactions. A study by senior IRS researcher Kim M. Bloomquist suggests that beyond the tax law complexity/tax law compliance correlation there may be “trends in the environment that account for the rising tax noncompliance.”[23] According to the study, the presumed rise in tax noncompliance “may be due, at least in part, to a shift in taxpayer income away from more visible to less visible sources.” According to the study, income that is not subject to third-party reporting is highest among taxpayers with the highest incomes. For the top 5 percent of taxpayers, unmatchable income as a percentage of Adjusted Gross Income increased by over 98 percent between 1980 and 2000.

The IRS has shown that there is a high correlation between tax compliance and third-party information reporting. The difference in compliance rates between individual wage-earning taxpayers and those operating businesses is striking. The IRS has estimated that individuals whose wages are subject to withholding report 99 percent of their wages for tax purposes.[24] In contrast, self-employed individuals who formally operate non-farm businesses[25] are estimated to report only about 68 percent of their income for tax purposes. Even more alarming, self-employed individuals operating businesses on a cash basis[26] report just 19 percent of their income to the IRS.

TIGTA believes that a combination of efforts will be required to increase voluntary compliance and reduce the tax gap. Tax simplification and increased transparency through third party reporting are significant contributing factors toward achieving these goals.

Gather Better Compliance Data

The IRS’ National Research Program (NRP) is designed to measure taxpayers’ voluntary compliance, better approximate the tax gap, and develop updated formulas to select noncompliant returns for examination. The first phase of this program addressed reporting compliance for individual taxpayers, and data from this phase were used to produce the updated estimates of this portion of the tax gap. These initial findings should enable the IRS to develop and implement strategies to address areas of noncompliance among individual taxpayers.

The next phase of the NRP, which has begun, focuses on Subchapter S corporations (Forms 1120S). These initiatives allow the IRS to update return-selection models for more effective return selection for its compliance efforts. In 2005, TIGTA reported that the return-selection formulas, developed in the 1980s, only accounted for the selection of 22 percent of the corporate returns selected for examination in FY 2004.[27] Updated selection models should contribute to more effective use of the IRS’ compliance resources.

In April 2006, TIGTA recommended that the IRS Commissioner continue to conduct NRPs on a regular cycle for the major segments of the tax gap. TIGTA also recommended that the IRS augment the direct measurement approach, and devise indirect measurement methods to assist in quantifying the tax gap. The IRS agreed with these recommendations, subject to available resources. In addition, TIGTA recommended that the IRS Commissioner consider establishing a tax gap advisory panel that includes tax and economic experts to help identify ways to better measure voluntary compliance. The IRS agreed to look into establishing such an advisory group with the intent of using it to validate and improve estimation methods.

Refine Existing and Develop New Compliance Strategies

The IRS conducts various compliance activities in an effort to reduce the tax gap. However, the IRS needs to develop a comprehensive strategy to reduce the tax gap. Almost 27 years ago, the GAO testified that “…it is clear that the Service [IRS] needs a comprehensive compliance strategy. To develop this, the IRS needs to determine the extent to which it is presently detecting unreported income from the various pockets of noncompliance. It then needs to consider reallocating its resources based on that determination and assess the need for additional resources to close the tax gap for each source of unreported income.”[28]

High-Income Taxpayers

Since FY 2000, the IRS’ Small Business/Self-Employed (SB/SE) Division has increased examinations of potentially noncompliant high-income taxpayers. In FY 2005, examinations of high-income taxpayers were at their highest level since FY 1996. As previously noted, the IRS considers high-income taxpayers to be those who file a U.S. Individual Income Tax Return (Form 1040) with Total Positive Income (TPI)[29] of $100,000 or more and those business taxpayers who file a Form 1040 with Total Gross Receipts of $100,000 or more on an Profit or Loss From Business (Schedule C) or on an attached Profit or Loss From Farming (Schedule F).

TIGTA recently reported the results of its review of the IRS’ increased examination coverage rate[30] of high-income taxpayers.