Does assessed value influence valuation judgments?

This was the research question and title of a paper by Mathew L. Cypher and J. Andrew Hansz that was published in the Journal of Property Research (Volume 20 Issue 4, December 2003, pp. 305-318).  In a prior post, we discussed the anchoring and adjustment heuristic and we learned that certain reference points influence human judgment.

This study was setup to observe if an assessor’s ad valorem value estimate influenced the market value opinions of others, both novices and expert real estate decision makers.  Novices were defined as undergraduate students taking an introductory real estate course.  Experts were operationalized as experienced real estate appraisers.  The 53 novices volunteering were randomly assigned to a control (or no assessed value), a low assessed value treatment group, and a high-assessed value treatment group.  The 44 experts volunteering were also randomly assigned to three groups: control, low treatment, and high treatment.

A valuation problem, based on an actual property and market, was designed and provided to all study participants, in the form of a written case, all the information necessary to make a valuation judgment: subject property information, neighborhood and area descriptions, maps, comparable sale information, and photographs of the subject property and comparable sales.  The subject property was a vacant piece of land, which is typically valued based on comparable sales data exclusively.

The novice and expert control groups received the valuation case exclusively and were asked to make individual judgments of market value.  The four treatment groups received the exact same valuation case as the control groups with one additional piece of information, the property’s current ad valorem and assessed values.  The assessed value was configured at two levels in relation to the comparable sale data: ‘high’ and ‘low.’

Would the assessed value reference points influence the treatment groups?  In other words, would the average property value estimates for control and treatment groups be approximately the same or would the ‘high’ and/or ‘low’ assessed value information cause different results among the groups? 

Because the participants were randomly assigned to groups, we should not expect any large differences between control and treatment groups if the assessed value information had no influence on valuation judgments.  However, if assessed value treatment did influence valuation judgment then differences in the groups would be observed.  What do you think was found?  Would the assessed value of a property effect your valuation judgment?  Prospective buyers often check the property’s assessed value.

Below is a summary of the study results.  Each cell is the average appraised value per acre.

	‘low’ assessed	Control	‘high’ assessed
novice groups	$82,242  (a)	$85,338  (b)	$93,261  (c )
expert groups	$91,054  (d)	$90,464  (e)	$89,753  (f)

Now, there are different average values in all cells, however, the only statistical (or significant) differences between these group means was between cells labeled a and c ($11,019) and b and c ($7,923).  Note that the significant differences are only in the novice groups.  Statistically speaking, there were no significant differences among the three expert groups.

These results appear to indicate that assessed value influence novice judgment and not expert judgment.  Novices (students), probably not very confident in their valuation abilities, look toward assessed value as a valid indicator of a property’s worth.  Using assessed value as an initial reference point (anchor), the assessment value did appear to influence their judgments.

Experts (experience appraisers) know about the limited content validity of assessed values and could very well be a flawed market value indicator.  Experts appeared to disregard this unsanctioned reference point and rely more on the comparable sales data.

Although we often use assessed value as indications of market values we should be careful in doing this.  As you can see, experts who make a living selling their own valuation opinions give very little credence in assessed values as an indication of market values.  Should you?