Originally published on Forbes.com Oct 31st, 2013
Jeffrey J Furnish just became my hero. I have been complaining for some time about taxpayers losing to the IRS when they receive mysterious 1099s from insurance companies. Not only did Mr. Furnish win the case ,he did it pro se. Mr. Furnish is an actuary and he won by insisting that the insurance company should show back-up for its computations of income. The company was unable to do that, which shifted the burden of proof to the IRS, a burden the Service could not bear. Practitioners who have clients with these surprise 1099s should study this case closely.
Mr. Furnish purchased a policy from Northwestern Mutual Life (NML) in 1972. The policy called for an annual payment of $340 and provided a $20,000 death benefit. In 1974 he signed up for an additional $50,000 for $462. (I’m ignoring some bells and whistles.) Mr. Furnish recalls paying four of the first seven annual premiums and electing to pay the remaining annual payments with premium loans. There was no evidence that Mr. Furnish ever took any money out of the policy.
In 2009, NML informed Mr. Furnish that the policies would expire if he did not put more money into them. He chose not to. I know that NML was not being spiteful. It just seems that way. They issued him a 1099-R showing a gross distribution of $78,414.14 of which, they said, $49,255.24 was taxable.
First An Argument That Will Go Nowhere
On or about August 16, 2010, petitioner submitted to the Internal Revenue Service (IRS) two Forms 1040, U.S. Individual Income Tax Return, for 2009, along with a written statement. Return “A” did not include the income reported by NML on Form 1099-R, whereas petitioner reported the income on return “B”. Petitioner’s written statement described the events leading to the lapse of the insurance policies and the issuance of Form 1099-R and set forth his claim that it would be unfair to impose income tax on what he considered an artificial distribution.
Thirty plus years of tax work will teach you that “It’s not fair” and “It does not make any sense” are not compelling arguments.
The IRS accepted and filed petitioner’s return “B” and assessed the tax reported on that return.
Mr. Furnish was not done.
How Did NML Compute His Taxable Distribution ?
NML responded to Mr. Furnish’s questions about his policies, but their answers were not clear. He requested clarification:
NML responded to petitioner by letter dated November 8, 2010, repeating verbatim the statements contained in the November 1, 2010, letter but adding the following statement:
At the time the policy lapsed to Extended Term Insurance in 1999, any remaining cash value in the policy provided insurance coverage until March 19, 2010, for *** , and June 4, 2010 for *** . There were no cash values paid to the policyowner in 2009 because loans and unpaid loan interest depleted cash values causing the lapse to Extended Term Insurance.
Dealing With The IRS
On December 1, 2010, petitioner wrote to an IRS office in Fresno, California (presumably an examination unit), stating that he questioned NML’s calculations underlying the Form 1099-R and that NML had declined to provide him with a “history of distributions” under his insurance policies because its electronic records went back only to the mid-1990s and the company was unwilling to do the research necessary to fully respond to his request.
During the next several months petitioner engaged in discussions with various IRS personnel in an effort to resolve the matter in his favor. In September 2011 petitioner attempted to obtain the assistance of the Taxpayer Advocate Service.
That went nowhere. The Service issued a deficiency notice for $22,444.52 in tax and an accuracy-related penalty of $4,488.90. Given Mr. Furnish’s efforts up to this point, the latter seems misguided to me. He had apparently not been given a thirty day letter, but he took a shot at appeals anyway along with his Tax Court petition.
Petitioner nevertheless wrote a letter to the IRS Office of Appeals dated August 13, 2012, asserting that he did not receive a distribution from NML, summarizing his efforts to obtain from NML detailed calculations underlying the Form 1099-R, and expressing doubts about the accuracy of NML’s calculations supporting its determination that his policies had lapsed during 2009.
The IRS continued to rely on NML’s numbers.
The record includes a declaration executed by Carol A. Stilwell, NML’s director of policyowner services and a custodian of the company’s business records. Ms. Stilwell’s declaration states that petitioner’s 1972 and 1974 policies terminated on June 4 and March 19, 2010, respectively, with no value. Ms. Stilwell attached to her declaration separate exhibits which provide the same narrative statements and calculations in respect of petitioner’s insurance policies that appeared in NML’s letter to petitioner dated November 1, 2010.
The Lawyerly Stuff
The Tax Court has generally been quite unsympathetic to taxpayers in Mr. Furnish’s position.
When NML determined that petitioner’s insurance policies had lapsed, it applied the cash values of the policies to the outstanding balances on petitioner’s loans. As we have explained in numerous cases, the act of applying the cash value of a life insurance policy against an outstanding loan is not different from distributing the proceeds to the taxpayer (including the untaxed inside buildup) to permit the taxpayer to use the proceeds to pay off the loan.
So it looks like he is going to lose, just like all the others before him. Here is where his effort pays off.
A preliminary issue in this case, however, is whether NML correctly determined that petitioner’s insurance policies had lapsed in 2009. As a general rule, the taxpayer bears the burden of showing that the Commissioner’s determination is in error. Rule 142(a). As an exception to this general rule, if a taxpayer raises a reasonable dispute with respect to a third-party information return (such as the Form 1099-R in dispute) and has otherwise fully cooperated with the Commissioner, the burden of production may shift to the Commissioner to present reasonable and probative evidence to verify the information return.
That is what won the case for Mr. Furnish.
Contrary to respondent’s position, petitioner raised a reasonable dispute regarding the accuracy of the Form 1099-R. Although petitioner points to relatively minor discrepancies in NML’s records, we agree with petitioner that the discrepancies are of such a nature that their cumulative effect, compounded over the extended terms of the policies in question, would likely be significant and could very well alter the dates that the insurance policies lapsed.
It is a less than satisfying victory. If NML and the IRS could come up with accurate internally consistent numbers Mr. Furnish would still be nailed, unfairly so in my opinion, but they apparently cannot. This approach might not work on a larger policy, one that is not as old or a policy from a company that keeps better records than NML, but it is worth a try.
You Don’t Have To Be An Actuary – But It Probably Helps
I contacted Mr. Furnish for his comments on the case. Here is what he has to say.
In my view, taxpayers like me who bought policies before 1986 are indeed getting screwed. The problem is that the 1986 tax reform repealed the tax-deductibility of the interest paid on policy loans, but did not alter the tax consequences of policy surrender. This applied to policies already in force, and many of these were sold advising the policyholder to stop paying premiums after 7 years. The tax change affects the expected results when the policy was sold, and policyholders were not advised of the problems created for the premium payment schedule they were already following.
Be that as it may, for people caught in this situation, the unfairness of the tax code is not a winning argument with IRS. IRS is only interested in enforcing the tax law as written. My advice would be to be totally up front with the IRS and produce what records you have. In my case, my records showed that changes had occurred in the original terms of the policy, and the limited records disclosed by the insurer left many questions about the accuracy of the insurer’s calculation, as for example there had clearly been plan amendments along the way. Even small changes over the 35+ years these policies were in force would affect not only the amount disclosed on the 1099, but would also likely affect the lapse date, so that the taxable event might occur in a year other than the year in which the 1099 was issued. Fortunately the Tax Court agreed.
It has been so long that I actually forgot about interest paid for life insurance being deductible before 1986. It emphasizes how unfair the results often are on older policies, but, as Mr. Furnish notes, that will not get you anywhere with the IRS or the Tax Court. Challenging the IRS and industry bureaucracies to back up their numbers seems to be a more fruitful approach.
You can follow me on twitter @peterreillycpa.
Correction – In an earlier version of this post I indicated that Mr. Furnish was not an attorney at all. He informed me that he does have a law degree from George Washington University and is an inactive member of the bar in Virginia and DC. He has, however, always worked as an actuary.