The IRS periodically releases estimates of the tax gap – the difference between taxes owed and taxes paid in a timely manner. The latest IRS estimates show a tax gap of about $350 billion for tax year 2001. The tax gap consists of three components – non-filing (taxes not paid by those with a filing requirement who fail to file), underreporting (taxes underpaid by those who file but underreport what they owe), and underpayment (taxes not paid by those who fail to remit reported amounts owed when due). IRS reports the tax gap for separate tax sources, including individual income taxes, corporate income taxes, employment taxes, and estate taxes.
The purpose of the Tax Gap Subgroup is to help the IRS improve its estimates of the tax gap. In prior years, the subgroup reviewed and supplied comments on the IRS research methodology for estimating underreporting of corporate income taxes, individual income taxes, income reported by flow-through enterprises (partnerships and subchapter S corporations), and estate and gift taxes. During 2010, the subgroup reviewed revised drafts of two white papers prepared by the IRS Office of Research that described improvements in the methodology used for estimating two components of the estate tax gap – the non-filing gap and the underreporting gap – and provided updated estimates. The group met with IRS research staff on June 8, 2010 in a meeting chaired by the Director of the Office of Research, Analysis and Statistics and provided additional feedback on these papers. Generally, the subgroup was supportive of the IRS analyses, but did offer some additional technical suggestions. At the meeting, the group also discussed the broader strategy for updating estimates of the total tax gap and all its components to a more recent year, based on new estimates of some components of the tax gap and revised projections of older estimates.
A. Updating Estimates of Estate Tax Filing Noncompliance
The subgroup continued a discussion initiated in 2009 of a revised methodology used by IRS to generate estimates of the estate tax non-filing gap, providing feedback on a revised draft paper. In response to earlier feedback, IRS staff had made several improvements in the estimating methodology. The latest IRS paper estimates a non-filing gap for estate taxes in tax year 2004 (at the midpoint of a confidence interval) of $2.5 billion, or about 9.2 percent of tax liability. The estimate is similar in magnitude to estimates for earlier years using a different methodology. The subgroup raised some additional technical issues about the estimates, including taking into account the division of assets within households, the interaction between the estate tax and the gift tax, and the effects of tax planning behavior.
B. Updating the Estate Tax Underreporting Gap for TY 2004
The subgroup also reviewed a white paper on the estate tax underreporting gap. The paper introduced a new methodology based on the use of IRS operational audit data, with statistical adjustments to remove sample selection bias.
The IRS research paper estimated a voluntary reporting rate of 92 percent (underreporting of 8 percent of tax liability), with a 95 percent confidence interval ranging from 87 to 96 percent. In dollar terms, this represents an underreporting gap of $1.9 billion in 2004, with a confidence interval ranging from $0.7 billion to $3.0 billion). Based on the statistical confidence intervals, IRS staff concluded that the results were not significantly different from the estimates previously released for tax year 2001. The subgroup provided IRS staff with some additional suggestions on presentation of the results.
C. Strategy for Updating the Tax Gap Estimates
IRS is committed to providing updated estimates of the tax gap by the end of 2011. IRS staff and the subgroup discussed how frequently the tax gap estimates should be updated in the future, taking account of public demands for updated information, the uncertainty surrounding the estimates, and issues of how the estimates might be misinterpreted.
Members of the subgroup emphasized the need to explain carefully the determinants of the tax gap estimates and the degree of confidence associated with various components. There was concern that changes that reflect improvements in methodology or random errors could be misinterpreted as reflecting real changes in compliance rates. The group discussed how best to communicate the results to avoid any misinterpretation.
The subgroup has received positive feedback from IRS research staff on the value of their input to the development of improved ways of estimating the tax gap and presenting the estimates. IRS research has offered to continue to consult with the subgroup on an ad-hoc basis as their estimates continue to evolve and new technical issues arise.