A Tax Crimes Fable - Plea Dilemmas for the Innocent Defendant (3/9/09)

Once upon a time a major accounting firm looking for revenue and thus profits began a tax shelter operation. The tax shelter operation involved accounting firm personnel at the center (the designers and the decision makers that the strategy should be promoted) and personnel at the periphery in the regional offices of the firm who would market the shelter designed and approved by those at the center. Persons outside the firm were engaged to design and implement a trading strategy that would, at least facially, support the shelter. One such person was a true market expert who designed a trading program that, while independent of the tax rules play giving rise to the shelter, would be wrapped into it to supply the claimed profit motive. A lawyer with an independent law firm worked with this team to refine the program in a way that he felt would give him a basis for rendering a more likely than not tax opinion to the taxpayers buying into the shelter. The tax shelter was successfully marketed. Billions of dollars of tax liability escaped the fisc. The Government was not happy and spied a conspiracy with a cast of hundreds. In a first wave, the Government convinced the grand jury to indict the promoters. The promoters indicted included 12 of those at the center (including accounting firm personnel, the outside lawyer, and the market expert) and three of those in the field whose only role was to sell the shelter. The indictment alleged an overall conspiracy (both offense and defraud / Klein conspiracy) and 39 substantive counts of tax evasion (with perhaps overlapping Pinkerton and accomplice aiding and abetting liability for those not directly involved in some or all of the substantive counts) with respect to the shelters sold.

The Government then tried to pick off some of the vulnerable defendants to offer a plea. The plea negotiation went like this: If we are successful in prosecution, you (the defendant being solicited for the plea) are at risk of a sentence that the Guidelines (then mandatory) pegged at a low range of 25 years. The Government’s first offer to shake this defendant down was for 2 counts of tax evasion (maximum 10 year sentence). Getting a no, the Government then sweetened the pot with a 1 tax evasion count offer (maximum 5 year sentence). The defendant who has always maintained his innocence is a rational actor and is tempted by the new offer. Being a rational actor and knowing the risks that he might be tarred by being included in a conspiracy where there are some persons with less compelling objective facts than his, he perceived that a certain cap at 5 years, even with the loss of some civil liberties, would be better than the risks of going to trial – to wit up to 25 years. He wanted the 1 proffered count but he was unwilling to profess guilt in order to get the plea. The fable ends there with his dilemma.

Comments:

1. The concept of permitting a truly innocent defendant to plea to and be convicted of a crime that he did not commit is disturbing but it does illustrate a truth. In our system, the innocent are convicted because those involved in prosecuting, determining guilt and imposing punishment draw the wrong conclusion (even when that may be a – or even the – logical conclusion from the incomplete evidence).


2 I posited this fable to illustrate this truth and the tools that are available to deal with this truth. From the defendant’s perspective, the risk after indictment is that the trier of fact – usually the jury in a criminal case, but sometimes the judge – draws the conclusion that the truly innocent is truly guilty. That is the risk the defendant in this fable faced. How real was that risk? Obviously, that is fact and case specific, but there are certain general systemic ways of analyzing the risk.

3 In order to convict, the trier must find guilt beyond a reasonable doubt. You may think that this standard prevents conviction of the innocent in all except the truly aberrational case. Any system that required certainty could not work. So, societally, we are willing to accept conviction of the innocent in the aberrational cases which we think are minuscule because of the high burden of proof before conviction.

4. What does that burden of proof – beyond a reasonable doubt – mean? The following is a traditional formulation for the jury (from Fifth Circuit Pattern Jury Instructions):

While the government's burden of proof is a strict or heavy burden, it is not necessary that the defendant's guilt be proved beyond all possible doubt. It is only required that the government's proof exclude any “reasonable doubt” concerning the defendant's guilt. A “reasonable doubt” is a doubt based upon reason and common sense after careful and impartial consideration of all the evidence in the case. Proof beyond a reasonable doubt, therefore, is proof of such a convincing character that you would be willing to rely and act upon it without hesitation in the most important of your own affairs.

5. The standard presumes that there can be errors. Using the metaphor of important decisions in our own affairs, we know that we make mistakes on important matters. Many authors note that this type of definition of the standard is not particularly helpful but it is perhaps the best we have.

6. Trial lawyers often state burdens in percentages. For example, in the normal civil case, the burden of proof for a proponent of a fact is “more likely than not.” In percentage terms, this means that the trier of fact must believe that the fact is 51% more likely true than false. 50% won’t do; it must be 51% or higher to find the fact for the proponent. Using this percentage methodology, how do we describe – hopefully in some meaningful way – the burden imposed by the “beyond a reasonable doubt” criminal burden of proof?

7. In various polls of judges (who after all have the best opportunity to judge and a real incentive to be able to describe the process), the ranges are from 75% to as high as 95%, with the critical range being from 80% to 90%. (I threw out as aberrational one response from a federal judge who said it was 50%.) For further information about this and my sources, click here.

8. If we can state these percentages in terms of finding an innocent person guilty, then the critical range is inferentially 10% to 20%. This inferential extrapolation is not a perfect exercise, but no lesser light than Judge Posner has engaged in this type of inference. Still whatever the appropriate range inference inherent in the beyond a reasonable doubt standard, the point is inescapable that innocent people are convicted. Innocent people are convicted, and the systemic chance that that might happen is very high.

9. The function of the nolo contendere plea and the Alford plea in this context is to the permit the rational actor innocent defendant to assess the chances that injustice will result from the trial and reach a plea resulting in conviction while maintaining his innocence. Sure, they also permit the truly guilt to continue the charade of innocence, but I am not particularly concerned about that type since I only represent innocent defendants. (Maybe those of you who represent guilty defendants desiring to maintain their innocence can send me your experiences in trying to assist them in doing so.)

11. And, in the fable, that was precisely what the defendant tried to do. But, in the rest of the fable (which I left off), he faced off with DOJ policies that stymy the rational decision by denying prosecutors the latitude to reach such a plea. These plea policies are set forth in the DOJ CTM and may be reviewed here. (There is one infamous exception where the DOJ Tax apparently agreed to a nolo plea – Former Vice President Spiro Agnew entered a nolo plea to a charge including tax evasion.)

12. In sum, the innocent defendant has the following three choices. (1) go to trial where he believes there is a better than even chance he will be found guilty despite his actual innocence and suffer a draconian sentence as well as depleting his assets in a futile attempt to avoid that freight train; (2) re-imagine his guilt or innocence and come to the Eureka moment that he is really guilty after all; or (3) lie to the judge in the plea colloquy in asserting his guilt. I know of no empirical study as to the incidence of (2) or (3), but one such case which is hard to imagine in either category is too many. None of the choices are, however, good ones.

13. For more reading on this, see the relevant section of my Federal Tax Crimes Book by clicking here.