Date: Sample Cross-Examination
April 1, 1999
To: Mr. Attorney
From: Spectrum Economics
Re: Cross Examination of Dr. Expert
The following are sample questions that have to do with Dr. Expert's qualifications and
methodological assumptions. The questions are designed to reveal errors that lead to inflated loss
Dr. Expert is an Assistant Professor of Economics at Regional University, an entry level title for new
Ph.D. professors. Prior to becoming an Assistant Professor in 1985, he was an Instructor in
Economics for fifteen years, the title many graduate students have in Ph.D. programs. He has a M.A.
from State University. His career activities consist of teaching at Regional University and
consulting. He has not published in any peer review periodicals.
Dr. Expert uses a very simplistic technique for forecasting future values that is not only bad
economics, but is incorrect and misleading. Anyone can do what Dr. Expert does with a copy of The
Economic Report of the President and a calculator. By choosing a historical period that goes back
to 1953 as the basis for future predictions, Dr. Expert is able to include a time when wage growth
was much higher than it has been for the last twenty years. This enables him to calculate a low net
discount rate which results in a higher present value of economic loss.
The point here is that Dr. Expert's methodology is not credible. Any economist, academic or
business, would be considered hopelessly inept if he tried to seriously use such a technique. And
Dr. Expert is of course not consistent with top economic forecasters. His method leads to estimates
of compensation growth that are 2% higher than consensus estimates by top forecasters in the U.S.
(Blue Chip and DRI).
Dr. Expert wrote a paper in 1983 on forecasting college enrollments. It should be easy to show that
applying his current technique to forecast college enrollments would be ludicrous.
To determine future earnings growth, you use a method for forecasting that's based on historical
averages. Is that correct?
Why do you do that? (He'll probably say something about it being necessary to go as far back in the
past as he's projecting into the future in order to get an accurate forecast.)
And the period you use for your average is 1953 to 1993. Correct?
And you use the same method for all your projections: wage growth, interest rates, medical costs,
You once wrote a paper called "An Economic Forecast of College Enrollments." Did you use the
same method for forecasting in that paper? Did you base future college enrollment on a simple
historic average of past growth in college enrollment?
Did you take into account anything other than the average growth in the number of college enrollees
for the past thirty years or so? How about the number of students graduating from high school?
How about the costs of attending college? What about the number of jobs that require a college
education, or the difference in earnings for college vs. non-college attendees?
You wrote that paper in 1983. I looked up college enrollments in the 1994 Statistical Abstracts. (1-
1) Do you know what the average annual growth rate in college enrollments was from 1955 to
1980? (6.24%). It was about 6% per year. In 1983, is that about what you predicted college
enrollments to increase by every year? Would there be a problem with doing it that way? Actually
there would be, wouldn't there? Do you know what the average annual growth in college
enrollments was from 1980 to 1990? (1.34%) It was just a little over 1% per year.
If Dr. Expert says that historical averaging is inappropriate for forecasting college enrollment
because it doesn't pick up changes over time that affect college attendance, you can ask him
if there aren't changes in national and global economic conditions that affect wage growth
and interest rates.
So if you used the average annual growth in college enrollment in 1983 to project 1995 enrollment
you'd be off by quite a bit. In fact, at 6% per year from 1980 on, college enrollment in 1995 would
have hit about 30 million. In fact, current projections for next year are about 15 million.
Now when you forecast wage growth, interest rates, and inflation, you base future growth on historic
data from 1953 to 1993. Is that right? And your source for this historic data is The 1994 Economic
Report of the President?
To be sure we understand what you've done, let's take wage growth as an example. If I were to use
Dr. Expert page 2
your method to forecast future wage growth, I'd go to the 1994 Economic Report of the President
and add up the change in compensation every year from 1953 to 1993 and divide by 40 to get the
And then I would have a forecast of future compensation growth for every year into the future
And would this forecast be just as accurate as the forecasts done by economic forecasting groups
such as DRI or the Blue Chip panel of forecasters?
Are you aware that the blue chip forecasts are the consensus forecasts of a panel of about 50 of the
top economists in the country including economists from Chase Manhattan, Kemper Financial
Services, Merrill Lynch, Nationsbank, Standard & Poor's, and Wells Fargo, just to name a few?
Do you think that they use the Economic Report of the President and a calculator for their forecasts?
Can you explain why their forecasts of compensation growth are so much lower than yours?
What if I'd used your method ten years ago to forecast compensation growth today? Suppose I take
average annual compensation growth from 1953 to 1983 and then increase actual 1983 compensation
by the growth rate I'd calculated and compare the result I get to the actual data. Let me show you a
graph of how far off I'd be. Now this is in inflation-adjusted dollars and the data is taken straight
from the 1994 Economic Report of the President. (Exhibit 2)
8. Life Expectancy
Dr. Expert cites as his source for age at "Normal Life Expectancy" as Vernon's Annotated Statutes,
a 1983 individual annuitant mortality table. According to Dr. Expert's testimony, this life table was
commissioned by the insurance industry or some actuarial firm for the industry. He states that it is
in the Missouri Statutes.
Generally, his life expectancies are above those of the U.S. Department of Health and Human
Services, National Center for Health Statistics. The following table is a comparison of his life
expectancies and those of the Department of Health (males, all races), for various male claimants
he has worked with.
Dr. Expert page 3
Dr. Expert's U.S. Dept
Dr. Expert's Life Health
Age Expected Life Health Life
Expectancy Expected Life
41 39 80 35 76
33 47 80 41 74
The Vernon's data is not a reliable source for life expectancy for the general population. It is an
annuity table and healthier (wealthier) people tend to buy annuities. The life expectancy of an
annuity purchaser, then, is statistically greater than that of the general population. In addition, most
annuity tables contain a "reserve" component. This is a method of adjusting mortalities so they
reflect a larger than statistically expected required payment. (Insurance companies are profit
seekers.) This reserve component increases the life expectancy used in annuity mortality tables.
Clearly, Dr. Expert's source skews life expectancies to a life span of about 80 years. This is
inconsistent with U.S. Department of Health life tables that show life span estimates do not converge
for those of different ages.(7)
Dr. Expert, on what do you base your life expectancy assumption?
Do you have a copy of that source with you?
For whom was the study prepared? Who produced it?
Does the study take into account sex and/or race?
Why do you use this source as opposed to others? What makes this source more reliable? Is the
study you use biased in any way?
Is your study used by the insurance industry? If it were prepared by the insurance industry, wouldn't
one conclude that the insurance industry would use it?
What do insurance companies mean by the term "reserves"? How are reserves incorporated into
Dr. Expert page 4
mortality tables? What impact do reserves have on life expectancies as presented in annuity tables?
What would insurers use mortality tables for?
How do you explain the difference between life expectancies between the Vernon's table and those
of the U.S. Department of Health?
Dr. Expert page 5