Editor's note: The following article was rekeyed and reprinted on August 24, 1997 in Resource Library with permission of the author. Prepared for the column Dunbier on Fine Art Valuation, Dr. Dunbier writes on issues regarding the valuation of fine art. If you have questions or comments regarding the article, please contact Lonnie Pierson Dunbier in Scottsdale, AZ, at ldunbier@mac.com.


Fine Art Comparables - Part One

By Roger Dunbier, PhD


In the real estate profession,comparables are at the heart of understanding values.In original one-of-a kind painting, they have never truly existed or have been used only rarely and in a very rudimentary fashion.

Realizing this omission, we at ENCompass Fine Art are changing this circumstance by bringing our large well-organized data base into utilization. Spending the last fifteen years creating a database of American artists linked to their retail and secondary market prices and to the literature in which they are referenced, we--that is the computer--can now accurately position an artist in the marketplace. We can also tell you the relative position to other artists. These determinations are called COMPARABLES.

A little background in the way of analogy....In the appraisal of a house, it is essential to know the prices paid in recent transactions involving properties in the vicinity and which at the same time approximate the size, etc. of the residence under scrutiny.

Traditionally in the appraisal of paintings, what could be termed "internal" circumstances of size, medium, framing, condition, etc. are only taken into careful consideration so that the value of a small watercolor by a given painter can be related to a larger oil by the same hand.

The real estate analogy here would be the comparison of values of a small house, two bedrooms, totally modern on a large lot with a larger house, four bedrooms in need of repair on a smaller lot. The comparative value of bedrooms versus lot size can be calculated "internally"---only one house needed---provided the appraiser has a background in the market. The other half, call it external, is finding a recent sale nearby. Depending on circumstance, this may be difficult, but in real estate it is always attempted, usually accomplished.

In the field of fine art, this process is analogous to finding another artist who is comparable. This is not done, usually not attempted. In appraisal of a Winslow Homer watercolor, comparisons with Mary Cassatt are not made, but they could be. In comparison of hypothetical artists, Smith and Jones, to these above luminaries, having markedly lesser values in the order of one-twentieth to one-hundredth, comparables should be utilized, but they are not.

The PROBLEM has always been:

1. Not Enough Data

2. Too Many Variables

3. Too Many Opinions

4. No Mechanism

We at ENCompass have always proceeded on the basis that the ANSWER to:

1. Not Enough Data is to Collect Data and then collect some more where nobody has previously looked.

2. The Variable Problem is to Isolate, Structure and Integrate.

3. The Opinions Problem is to Investigate, Ignore or Incorporate.

4. No Mechanism is Build One. Use the computer. Write programs.

Following this agenda, we assert that legitimate comparables can and should find a place in the appraisal of painting, drawings and illustration.

The question could be asked: "Why do we need them?"

The best answer is that determining comparables allows the evaluator to fit the work of a subject artist into a logical position between comparable individuals who are both more valuable and less valuable in today's marketplace. This positioning will facilitate a valid comparison with one or more artists who are better known to the investigator and to a wider audience. This positioning is important because so many good artists' average to best works so rarely find their way into result-published auctions, the principal methodology presently used by evaluators.Too many collectible works are by artists that seldom show up at auction or have not done so regularly or in reasonably recent times.

It is also well known that auctions have their problems as sources of valid information, a matter outside the purvue of this correspondence. I should advise, however, that relatively isolated estimates and prices realized at auction, particularly when unaccompanied by pictorial or other clear evidence of exact subject matter, can mislead -- disastrously providing estimates totally outside any reality.

Let me reemphasize that non-existent or isolated auction results are the rule rather than the exception for the works of artists that are quite valuable in collections. Recent and numerous sales at auction are the exception. To illustrate the fact that many top reputation artists do not have auction results to validate their position, let us use a slice of the database.

Using the ENCompass database, we isolate 382 painters who were active in the state of New Mexico prior to 1940. Since acquiring the works of early New Mexico artists is a growth activity, this group is in its entirety collectible. Over the last 9 years, these 382 had 4762 works show up at auction. 3413 were sold. Of the 382 total, 162 artists had nothing put up for sale at auction over the period, and 180 had nothing sold. About half of the 202 selling artists had only one or two works exchange hands, and these transactions were in a period of nine years! Taking this information alone, the cynic would say that for any of these individuals, this is not data; it's anecdote. For this reason alone a little demonstration of the importance of comparable data is in order here.

Using as examples ENCompass-listed painters who worked in New Mexico before 1940, one finds the following twenty names. All have had works which have sold at auction or other venues for over $15,000. and all are comparable in collectibility. They rank roughly between 900 and 1300 on our total price list of American artists. However, in the numbers of pictures having sold at auction over the last 9 years, they vary all the way from 0 to 83.



New York Birge (Lovell Birge) Harrison 13

Connecticut Eric Sloane 83

New Mexico Warren Rollins 16

New Mexico Gene (Alice Geneva) Kloss 7

California James Swinnerton 20

California George Otis 47

California Zavier Martinez 19

Arizona Lon Megargee 3


Arizona LEW DAVIS 1


New Mexico Arnold Blanch 9

California Thaddeus (Tad/Thad) Welch 28

Connecticut Vincent Colyer 2

Oregon Louis B Akin 0

New Mexico Julius Rolshoven 10

New Mexico Alfred Morang 3

New Mexico Joseph A Imhof 13

New Mexico Russell Vernon Hunter 0

Virginia Eliot Candee Clark 27

New Mexico Alfred Morang 3


So the question remains, why with minimal auction results, are these artists considered collectible?

Attempting to answer that question, let us take three artists from the middle of the list: Joseph Fleck, Lew Davis and Charles Craig. Take my word for it, these three are highly collectible, particularly in their home states of New Mexico, Arizona and Colorado. However, if you were to try to establish their values from auction data taken alone, eleven total sales over the nine year period is "slim pickings" indeed. Chances are that most appraisers would incorrectly evaluate work by these artists because otherwise knowledgeable east or west coast dealers would know little about these three New Mexico painters and those "experts" from their respective home states would over-emphasize their values.

The comparable list, having no truck with opinions, places these three artists between seventeen other artists whose names are familiar to experts and collectors on both coasts and in-between. To price any work of the three, either higher or lower, than the price parameters represented by the seventeen others would it seems necessitate considerable explanation directed toward the quality or condition, etc. of the subject work. In this context, we have created a new evaluation tool, not necessarily as a replacement for existing appraisal techniques but as an additional methodology in the arsenal.

More importantly in the category of values at time of sale where formal appraisals are not generally obtained, this knowledge of relative position or ranking, derived from our evaluation methods, provides a much needed buttress against critical mistakes in judgment of artist name value, plus inoculation against buyer's (or seller's) remorse.

For more understanding about our innovative evaluation methods, see the next section titled Fine Art Comparables - Part Two.

© 1997 Roger Dunbier



About the Author:

From 1982, Roger Dunbier (1934-1998) applied his training as a professional economist to the fascination he had with American art and to his conviction that something needed to be done to correct the fact that marketing hype rather than factual information was creating artists' reputations and driving the art market.  Dunbier's answer was the development of a comprehensive computerized art management system based on economic principles of 'arbitrage' and 'comparables'.   In other words with access to a shared pool of information, people from all regions of the country could make decisions based on the same data.  Dunbier dedicated the remaining years of his life to realizing this vision, and by the time of his death in 1998, his system had nearly 17,000 artist records linked to their respective literature and auction-price results.  Facilitating the technology was Charles Lefebvre who devoted thousands of volunteer hours to programming increasingly sophisticated computers. With this aggregation, as Dunbier predicted, interesting results began to emerge as to 'who was really who' in American Art.  Today his work lives on as the database underlying AskART.com, the internet site that now offers information on several hundred thousand artists. (right: Roger Dunbier, photo courtesy Lonnie Pierson Dunbier, derived from a larger image at http://tfaoi.org/am/16am/16am17.jpg)

In those days before the internet, Dunbier and his wife, Lonnie, published the growing information in regularly updated books.  The basic directory was The Artist Bluebook.  The  name was suggested by his friend, John Hazeltine of Traditional Fine Arts Online, who perceived that Dunbier had created a buyers' and sellers' gauge for American Art similar to the Kelly Bluebook, the 'standard reference' of car marketing. 

Dunbier's innovation of computer systems began in 1963, when he pioneered computer mapping on what were then relatively primitive computers.  In 1967, he utilized the concepts of arbitrage and comparables in designing the first real estate Multiple Listing System.  Its direct descendent remains in use across the United States. 

He was born Omaha, Nebraska in 1934, and grew up in that city.  His interest in American art was linked to his formative environment of immersion in fine art viewing and conversation.  His father, Augustus Dunbier, (1888-1977) was a prominent landscape, still life and portrait painter and art teacher, whose studio and classroom were in the family home.   Although Roger picked up a general interest in art, it was obvious (to his father's frustration) that he had no art talent,  But he did have other talents.  Graduating first in his class and Summa Cum Laude from the University of Omaha in 1955 with majors in economics and history, he received a Marshall Scholarship, which led to enrollment at Oxford University in England from 1955 to 1959.  During that time, he was also on the Oxford University basketball and track teams, and was a member of the British National Basketball Team.  In 1961, he received a Doctorate of Philosophy, Economic Geography from Oxford.  His dissertation, The Sonoran Desert, Its Geography, Economy, and People, was published by the University of Arizona Press in 1960, and subsequently used as a text book for college geography courses.

After his formal education, Dunbier held full-time professorial positions for several years at the University of Omaha and at the University of California-Irvine.  He lived most of the remainder of his life in Phoenix and Scottsdale, Arizona.  He had economic geography related jobs including CEO of his management consulting firm that prepared demographic and locational studies; and President of Metro Press, Inc., publisher of over 100 computer generated area directories for Metro Phoenix.

In 1991, he married Lonnie Pierson of Lincoln, Nebraska, and she joined him full-time, developing the American artist database.  Two years after Roger Dunbier died,  Lonnie and Charles Lefebvre partnered with George Collins of San Francisco to create AskART. com.   

-- By Lonnie Pierson Dunbier, 2008

Editor's note: readers may also enjoy:


return to Fine Art Valuation by Roger Dunbier

Visit the Table of Contents for Resource Library for thousands of articles and essays on American art.

© Copyright 2007 Traditional Fine Arts Organization, Inc., an Arizona nonprofit corporation. All rights reserved.