Drawbacks to the Use of “Preliminary” Estimates

by Thomas R. Ireland

This article was originally published in the autumn 1998 issue of the Expert Witness.

In a recent posting to an e-mail service for economists who act as expert witnesses, Thomas Ireland, a well-known American economist, wrote about an issue that usually receives little attention — the dangers of relying on preliminary estimates. Although we do not agree with all of Dr. Ireland’s conclusions, we feel that many of the points he raises are of sufficient interest to be reported here. Accordingly, we reprint the following letter, with Dr. Ireland’s permission.

Fairly recently, I sent a letter to an attorney who had requested that I avoid producing any report except the final report of my opinions. This involved a situation in which I was asked to prepare preliminary estimates, but might not be asked to finalize my reports until several months later. It occurred to me that the text of this letter might be of interest to fellow practitioners.

Dear Attorney:

The purpose of this letter is to explain why I feel that I need to develop preliminary reports. I write my reports in the form of a letter addressed to an employing attorney unless asked to use a different format. In a very real sense, however, I write them to myself as well. This is for three reasons. First, in developing most reports, I must make a number of very small decisions that do not have a large impact on my damage assessments, but are things I need to remember at depositions and at trials. In my tables, is “age” shown as the plaintiff’s age at the start of the year, or on his birthday during the year? What issue of USFinancial Data was the source of my discount rate? How did I annualize the earnings figure in the year of injury? And so forth. If I am preparing a preliminary table, I need this kind of information if I pick up the file three months later and am trying to figure out exactly what I did to produce the exact numbers in my tables.

Second, my reports are not full of fluff that is designed to make it look like I did more work than I did (like many other economist reports that I see). They are simple narratives explaining what I was asked to do, what assumptions I was asked to make, what materials I was given to prepare, what I took from those materials, what additional assumptions I made, what my opinions are and how my tables work to produce those opinions. In effect, as I am writing the narrative parts of my report, I am subjecting my calculations to a step by step logical consideration of whether I have performed those steps correctly. I find mistakes much more easily when writing my narrative than by staring at tables or spreadsheets.

In other words, I need to develop a preliminary report to remember what I did and to check my calculations at the time I create them. Attorneys hate to see what they regard as unnecessary documents that might come back to haunt them. But the cost of not creating documents may be serious and much more embarrassing errors or failures of memory. For me to prepare extensive notes rather than narratives would take more of my time and be more likely to result in errors. Further, the notes themselves then become a document that could cause more trouble than a narrative. As I write my narratives, even in a preliminary report, every word is considered from the standpoint of whether I would want to be cross examined about the meaning of that word. I try to be very precise in my narratives, but am much less careful in notes I write to myself. I now typically type all of my handwritten notes as a part of deposition preparation to make sure that I have no handwritten notes that I do not understand.

Third, I want attorneys to read my narratives to see whether I may have misinterpreted any of the information they have provided to me. Sometimes important facts are given to me by telephone and I may have written them down incorrectly. Sometimes documents that I have been given create misleading impressions. I have found that attorneys are very good at picking up errors that relate to demographic facts in my reports. That is a very important part of the process of checking the validity of my analysis.

Perhaps a fourth reason is that right after I write a report, I see what I intended to write, but not necessarily what I did write. Several weeks later, I will see what I did write and be able to catch things that I could not have caught at the time I wrote a report.

The bottom line here is that if you ask me not to produce preliminary reports, you are asking me to take a greater chance of making an important mistake or being unable to explain how I arrived at one of the values in one of my tables.


Dr. Thomas Ireland is a professor of economics at the University of Missouri at St. Louis.