The use of litigation risk assessments to help make informed case management decisions.

Litigation risk assessments (LRA) use decision analysis tools to enhance the traditional approach to the evaluation and management of a tax controversy or other legal dispute. The end product of an LRA is a detailed analysis of the dispute in which all of the issues, their relationships, and the potential outcomes are identified. Percentage probabilities and present dollar values are used to describe the potential outcomes. The basics concerning LRAs and their usefulness for the evaluation and management of tax controversies are described in an article I published a little more than a year ago in The Tax Executive. "A Systematic Approach to Tax Controversy Management," 49 Tax Executive 223 (May-June 1997).

This article follows up on a second article entitled "Further Thoughts on the Usefulness of Litigation Risk Assessments for the Management of Tax Controversies," which appeared in the May-June 1998 issue of The Tax Executive (50 Tax Executive 187). That article discussed the different ways in which an LRA can add value for the settlement and management of a tax controversy, focusing on the use of an LRA to facilitate settlement. This article addresses the use of LRA data and tools for the tactical and resource allocation decisions that are at the core of litigation management.

Discussion

Through the preparation of an LRA, company management and counsel develop an understanding of the overall dispute and of the details. Those details, which can drive you to abstraction, are critical to success or failure. The key is paying careful attention to the details without losing sight of the overall strategy. Decision trees and the other litigation risk assessment tools assist you in doing just that.

The LRA tools offer the means to make informed tactical decisions involving the wide variety of factual and legal issues that can arise. Those decisions are important for both tactical and resource management purposes. For example, in a matter for which we recently provided assistance, the client had to decide whether to retain an expert to provide testimony on a complex, technical issue. As a result of preliminary discussions, the client was confident that the expert's testimony would be supportive of its case. The question was whether the support was sufficiently important to justify paying the projected six-figure fee of the expert. The LRA provided the client with the necessary tools to evaluate the importance of the expert's testimony to its case. With the testimony, the client had slightly more than a 60-percent chance of avoiding any liability. Without the testimony, the probability dropped to less than 50 percent. Many millions of tax and interest dollars were at issue. Not surprisingly, the client retained the expert. The more than 10-percent probability swing in the predicted outcome of the large dollar dispute made retention of the expert prudent.

That decision was consistent with the instinct of the management officials of most companies when faced with this type of choice. They are inclined to take every step possible in order to protect their position, with little regard for cost. The difference here is that the decision to retain the expert was premised on more than a gut feeling or instinct. The client was able to make an informed decision, based on a cost-benefit analysis.

Further, not all such analyses lead to the retention of the expert or pursuit of other case preparation activities. In several instances in which we have been involved, the decision was made to forgo expert testimony. Each decision was grounded on a cost-benefit analysis performed using data from the LRA.

One decision concerned the retention of an expert accounting witness. There was no question that "some benefit" would be derived from the testimony. The decision faced by the company was whether the benefit was sufficient to justify paying the large fee that the expert proposed charging for the testimony and extensive preparation that would precede it. The LRA data indicated that favorable testimony by the expert was unlikely to have a large effect on the outcome. The company's probability of prevailing was just a few percentage points greater with the testimony than without it. Taking that small effect into account, along with the size of the potential litigation loss that it was facing, the company decided not to retain the expert for testimony. The company did use the expert as a consultant, but at a significantly reduced cost.

I believe that the company would have made the opposite decision and retained the expert for testimony if it had not had the benefit of the LRA data. That belief is largely based on comments by the tax manager and company counsel. The usual instinct in such circumstances is to cover all of the bases because that approach reduces the opportunity for later criticism if the litigation does not turn out well. Counsel and other responsible company management are able to say that they did everything they could to achieve a favorable result.

The instinct to cover all of the bases also extends to discovery, which is another area that can generate significant litigation costs. Typically in tax cases, the taxpayer does not file many discovery requests. But there are instances in which discovery opportunities for the taxpayer do exist. In one case in which we were involved, the company had to decide whether to seek discovery of documents located in a foreign country. It was estimated that the cost of a foreign proceeding to obtain the documents and other costs related to the effort would be more than $100,000.

We evaluated the potential effect of the documents on the client's chances of success. Based on information that the client had obtained concerning the documents, we assumed for purposes of our analysis that the documents would be favorable to the client's case. We also assumed that the client would be able to obtain the documents through the foreign proceeding. The uncertainties we looked at were whether the client would be able to get the documents admitted into evidence and what effect they would have on the client's case if that occurred. The analysis that we performed using the LRA data told us that the documents would only cause a three to five percent improvement in the client's chances of prevailing. Based on that modest effect and uncertainty whether the attempt to obtain the documents through the foreign proceeding would be successful, the client decided to forgo the effort and save the $100,000.

In some cases, the law related to a critical issue is either unsettled or its application to the facts of your dispute is unclear. Argument by analogy is frequently tempting. It may be essential in some instances. The development of the factual and legal components of such analogies can be time and resource intensive.

An LRA can assist you make the decision whether the investment should be made. We were involved in one matter related to a company's use of the completed contract method of tax accounting. The primary issue was whether the contracts that the client had for the development and initial production of "prototypes" constituted a manufacturing contract that qualified for use of the method. There was very little in the way of directly applicable authority, but the company's tax manager and outside counsel were very imaginative in developing potential analogies. They estimated that researching all of the potential analogies would take three to four months and cost several hundred thousand dollars.

In order to set priorities and make this process manageable, the potential analogies were factored into the LRA that had been prepared for the dispute. Based on the facts and law known at the time, predictions were made as to the potential effect of each argument by analogy on the outcome of the dispute. Those predictions were factored into the LRA. This enabled the client to identify the analogies that had the greatest potential upside. The client pursued those analogies first. Ultimately, the client stopped the research after just three of the ten potential analogies that had been developed. Based on the information developed during the research, which was used to update the predictions in the LRA, the client decided that further effort was not warranted. The potential return did not justify the cost.

It should be noted that an LRA was not an essential predicate for this type of focused development taking place. Judgments could have been made and priorities set without the LRA. In my experience, however, it is unlikely that would have occurred or that the research effort would have been terminated after the development of just three of the ten arguments.

I say this for two reasons. The first is the probability that the type of hard thinking concerning the importance of the analogies that preceded the research would not have occurred. Human nature supports my belief. It would have been easier to simply do the research without making the hard calls. The performance of an LRA ensures that the hard thinking occurs early, rather than later (if at all).

The second reason is the ability that the LRA data gave the tax manager and outside counsel to make meaningful comparisons in deciding which to pursue and which ones to defer. They were able to compare the potential effect of each analogy on the ultimate outcome using percentage probabilities. For example, the LRA could tell the company that analogy A had a potential effect of 15 percent on the outcome, while the potential effect of analogy B was only 3 percent. That type of data is much more easily understood and used than the more traditional terms that are used to describe the effect of an argument. How do you compare an argument that is said to be "strong," with an argument that is said to be "substantial?" If you are trying to develop your case in the most effective, but cost-efficient manner possible, such descriptive language is not very useful in making hard resource allocation decisions. As a result, it is likely that the safe, but expensive route will be taken; all of the arguments will be developed. It is much easier to set priorities and make the hard decisions when you have data that show the effect, expressed as a percentage, of each alternative on the potential outcomes.

The use of an LRA to assist with the case management decisions is illustrated by analyses that can be performed using the LRA for the $12.5 million hypothetical tax dispute referred to in the article that was published in the May-June 1998 issue of The Tax Executive. The master decision tree for that dispute, which was included in the article, shows that there was almost a 62-percent chance of the taxpayer prevailing in that dispute.

That overall probability, as also shown by the master decision tree, is the product of the predicted outcomes of a number of issues. One of the major issues is whether the taxpayer's method of accounting is found to clearly reflect income. That issue is in turn dependent on several sub-issues that are diagramed in a sub-tree that feeds into the master tree. For example, the master tree shows an 81.7-percent probability that the taxpayer's method would be found to clearly reflect income. That percentage is the mathematical product of aggregating the probabilities assigned to each of the issues in the sub-tree. The assignment of each of those percentages is a subjective judgment that draws upon the collective knowledge and experience of the members of the assessment team. The sub-tree diagraming the various sub-issues is shown in Figure 1 on the following page.(1)

[Figure 1 ILLUSTRATION OMITTED]

For the hypothetical tax accounting dispute, we have assumed that the taxpayer engages an expert to support the accuracy of its accounting method. The LRA can be used to test the importance of that testimony to the chances of success. Without the expert testimony, the probability of the taxpayer prevailing on the accuracy issue dropped from 81.7 percent to 38 percent. When that lower percentage is fed into the master tree, the taxpayer's overall probability of success drops from 62 percent to 50.7 percent.

Assume that the projected expert fee was $200,000. Under the circumstances, it would seem likely that the taxpayer would retain the expert. The expert testimony increases by more than 11 percent the probability of the taxpayer prevailing, thereby avoiding a $12.5 million liability.

Another issue diagramed in the sub-tree is whether missing accounting data can be located. A probability of 60 percent has been assigned to the data being found. One case-management issue that the taxpayer would face is whether the potential benefit from searching for the data is sufficient to justify the cost. Assume that the search is expected to cost $50,000. The LRA gives the taxpayer data to evaluate whether that expenditure was warranted. If the search is not made, the overall probability of the taxpayer prevailing decreases by just a few percentage points from 62 percent to 59 percent. Given the potential tax liability of $12.5 million and the small shift in the taxpayer's likelihood of prevailing, the taxpayer might decide to forgo the search effort and save the $50,000. On the other hand, the taxpayer might decide to go ahead. The LRA would enable the taxpayer to make an informed decision, rather than one based solely on gut feeling and a cover all of the bases instinct.

Conclusion

In this article and the two others that I have had the opportunity to publish in The Tax Executive, I have discussed LRAs and the value that they can add for the management of a tax controversy. To paraphrase a favorite saying of a well-known federal judge, the benefit of an LRA is the ability to work smarter. Hard work is still required and mistakes will still be made. An LRA can assist you make that hard work more productive and reduce the number of mistakes.

(1) The computer program that I use for LRAs employs zeros and ones in the pay-off fields of a sub-tree in order that the calculated cumulative probability of the sub-tree can be treated in the master tree as the probability for the issue into which the sub-tree feeds. The zeros and ones in the taxpayer accuracy sub-tree allowed my computer program to treat the cumulative probability for the subtree as the probability of the taxpayer's method being found accurate in the master tree for the hypothetical tax dispute.

CRAIG D. MILLER is a principal in The Miller Group in Washington, D.C., and Savannah, Georgia. He is a former member of the Washington, D.C., law firm of Miller & Chevalier.