Tax Court Vacates Stipulated Decisions in Tax Shelter Case

By James Beavers, J.D., LL.M., CPA

In the latest decision in one of its longest running series of cases, the Tax Court granted the motion of a group of taxpayers to vacate stipulated decisions in their cases that were part of the IRS’s Kersting tax shelter project. The court also ordered that the accounts of all Kersting project taxpayers with stipulated decisions be adjusted on terms equivalent to those given by the Ninth Circuit to the Kersting project taxpayers who had not settled their cases. It granted the taxpayers’ motion and issued the order because it found that all the Kersting project taxpayers, including those who had settled with the IRS and executed stipulated decisions, had been damaged by the IRS’s fraud on the court in the trial of the Kersting project test cases. The court ruled that the IRS had not rectified the harm by disclosing the fraud to the Tax Court and had made low value settlement offers to Kersting project taxpayers after the fraud had been committed.


In response to the large volume of cases generated by tax shelter examinations during the late 1970s and early 1980s, the IRS and the Tax Court developed procedures (test case procedures) intended to expedite tax shelter litigation and to reduce the amount of IRS and judicial resources used to resolve them. Under the test case procedures, when a large number of taxpayers are contesting deficiencies related to the same tax shelter, several representative cases are chosen to be tried, and agreements to be bound by the results of these cases (piggyback agreements) are obtained from as many taxpayers as possible, thus greatly reducing the number of cases that are actually tried.

In the early 1980s, the IRS established the Kersting tax shelter project to investigate taxpayers who had claimed interest deductions from tax shelter programs promoted by Henry F. K. Kersting. The IRS issued deficiency notices disallowing these deductions to taxpayers that it identified as having participated in the programs. In response, Kersting hired an attorney to represent the taxpayers, which led to the filing of a large number of petitions in the Tax Court contesting the deficiency notices. From 1982 to 1988, to reduce the number of taxpayers litigating the deficiency notices, the IRS offered the Kersting project taxpayers a settlement agreement similar to those offered in other tax shelter projects that reduced the taxpayers’ deficiency amount by approximately 7%.

The majority of the Kersting project taxpayers did not accept the settlement offer. In order to try the large number of cases efficiently, the IRS, the attorney hired by Kersting, and the Tax Court agreed to use the test case procedures and chose to try eight test cases. A large number of Kersting project taxpayers signed piggyback agreements.

Before the test cases were tried, in clear violation of IRS and Tax Court rules, the IRS lead attorney and the supervising attorney of the local IRS office entered into secret settlement agreements with two of the test case parties, the Thompsons and the Cravens. The more advantageous of the agreements, the Thompsons’, reduced the Thompsons’ deficiency by more than two-thirds. Under these agreements, the petitioners agreed to remain test case parties. In the actual conduct of the test case trials, the IRS lead attorney actively misled the court to cover up the existence of the secret agreements. In the test case trial, the Tax Court held that the Kersting tax shelters were invalid and that the Service had correctly disallowed the interest deductions related to the shelters (Dixon, TC Memo 1991-614).

After the end of the test case trial, the IRS National Office found out about the secret agreements. The IRS immediately entered motions to vacate the decisions of the two test case petitioners who had entered into the secret agreements and of a third test case petitioner who did not have an agreement and who notified the court of the misconduct of the lead attorney and his supervisor. The Service did not, however, directly inform the nontest case Kersting project taxpayers of the full details of the misconduct or the secret settlement agreements.

While the rest of the test case decisions were on appeal, the IRS renewed its pre-trial settlement offer to the nontest case Kersting project taxpayers. In the offer letter, the Service disclosed that its attorneys had entered into secret settlements during the test case trials, but it did not reveal the terms of the settlements or the identity of the taxpayers involved. It also indicated that the test case decisions were being appealed but that they were likely to be affirmed. Some of the taxpayers accepted the renewed offer and executed stipulated decisions.

In 2003, more than 10 years after the conclusion of the test case trials, the Ninth Circuit held in Dixon, 316 F3d 1041 (9th Cir. 2003) (called by the Tax Court Dixon V), that the misconduct of the IRS attorneys in the test cases constituted a fraud on the Tax Court that violated the rights of the test case petitioners and the more than 1,300 other taxpayers who had agreed to be bound by the test case results. The court ordered the Tax Court to sanction the Service by entering judgments on the same terms as the Thompson settlement agreement for the remaining test case petitioners and the other taxpayers who had signed piggyback agreements and not settled with the IRS. The order did not explicitly give the Tax Court any instructions with regard to Kersting project taxpayers who had settled with the Service and executed stipulated decisions

Subsequent to the Ninth Circuit’s decision in Dixon V, the Lewises, taxpayers who had settled with the IRS when it renewed its original settlement agreement in 1992, filed a motion with the Tax Court seeking to have it vacate their stipulated decision and to reopen their cases so that they could receive the benefit of the Thompson settlement agreement. The Tax Court refused this motion, holding that the Lewises were not entitled to have their decision vacated because they had been aware of the IRS’s misconduct and the pending appeals of the test case petitioners when they agreed to their stipulated decision (Lewis, TC Memo 2005-205). The Lewises filed a motion for reconsideration, arguing that their settlement agreement did not encompass or foreclose imposing sanctions on the IRS on their behalf. The Lewises were joined in their motion by two other couples, the Hartmans and the Lius.

The Tax Court Reconsiders

Upon reconsideration, the Tax Court held that the Ninth Circuit’s decision in Dixon V required that all Kersting project taxpayers be given the benefits of the Thompson settlement. Therefore, it granted the petitioners’ motion to have their stipulated decisions vacated and held that after their cases became final, it would enter an order requiring the IRS to adjust the accounts of all the Kersting project taxpayers who had entered stipulated decisions in accordance with the Thompson settlement agreement, regardless of whether the taxpayers had filed a motion to vacate their stipulated decisions.

The Tax Court first explained that all the Kersting project taxpayers, whether they had agreed to be bound by the test case decision or had settled and entered a stipulated decision in their cases, were entitled to assume that the test cases would be “well and fairly tried.” Because they were not, all the Kersting project taxpayers were harmed by the IRS’s fraud on the court. The Tax Court further concluded that it had erred in the Lewis case in determining that the Ninth Circuit’s order in Dixon V did not apply to Kersting project taxpayers who had settled with the IRS because the court did not specifically state that the order applied to taxpayers that were not in front of the appeals court. The Tax Court found that the language used in Dixon V also referred to Kersting project taxpayers who had entered stipulated decisions and who had their cases reopened (by the Tax Court’s granting a motion to vacate their stipulated decision).

The Tax Court further found that it had erred in focusing on the Lewises’ acceptance of the settlement offer and the application of general contract law principles. Instead, it concluded that it should have focused on whether the Service had purged the fraud its employees committed on the court and rectified the harm the fraud caused through its disclosure of its misconduct and the settlement offer it made after the test case trials.

Looking at those issues, the Tax Court ruled that the IRS had failed on both counts. The court found that disclosure did not purge the fraud because the fraud had already been completed by the time it was disclosed. It found that the settlement offer did not rectify the harm caused by the Service because the terms of the offer were much less favorable than the terms of its secret settlement agreement with the Thompsons and that material facts were deliberately omitted from the offer, making it misleading in a number of ways. Therefore, the court held that all the Kersting project taxpayers bound by the results of the test cases who settled their test cases were entitled to the benefits of the Thompson settlement agreement, regardless of when they had settled their cases.


The Tax Court correctly concluded that all the Kersting project taxpayers were affected by the IRS’s fraud on the court and that the petitioning taxpayers (and the other Kersting project taxpayers who entered stipulated decisions) were entitled to the same relief provided to Kersting project taxpayers who did not settle their cases. However, in giving them the benefit of the overly generous Thompson settlement agreement terms, both the Ninth Circuit and the Tax Court seem to have forgotten that all the Kersting project taxpayers had unclean hands and deserved to pay the full amount of the deficiencies the IRS had assessed against them. While this case presents a difficult problem for the courts, there must be a better way to punish the IRS than to give taxpayers who have taken fraudulent deductions the benefit of these deductions.

Hartman, TC Memo 2008-124

Newsletter Articles


Directions in Individual Taxation

This article covers recent developments in the area of individual taxation, including the treatment of support payments and IRA and qualified plan distributions, the Sec. 469 material participation rules, and the taxability of state economic development credits.


2015 Tax Software Survey

See how nearly 5,000 paid CPA tax preparers rated the strengths and weakness of major tax preparation software products they used in 2015.