Download PDF

It is comforting to know that as assessors and appraisers, we speak the same language. We had a great learning experience last week in IAAO 331, Mass Appraisal Practices and Procedures. Our instructor was David Cornell, CAE, MAI. David is from New Hampshire and brought fantastic discussions to our group of 23 North Carolinians. The discussions and examples we experienced can be used for improving appraisal equity and uniformity in individual jurisdictions throughout our state. One of the items that we discussed was a worthy repeat from other mass appraisal courses: The importance of data in the assessor’s office. Not only do we need to collect the right data for model specification, but we have to collect it accurately.

At the upcoming NCAAO Fall Conference, the NCDOR will be conducting sessions on their new reappraisal standards, to be published later this year. Various parts of the IAAO Standard on Mass Appraisal of Real Property, IAAO Standard on Ratio Studies, and the IAAO Standard on Property Tax Policy will be referred to in the NCDOR reappraisal standards. It will probably help in these discussions if we’re speaking the same language. I want to share some information about data descriptions in this blog and then later I hope to make a connection to how the understanding can help the assessor effectively communicate needs in the budget.

I often hear real property guys ask each other, “How’s your data”? The appropriate answer apparently is, “Our data’s fine”. We can talk about how to make sure it is fine in another post, but the data that they’re talking about refers to the many codes, words, numbers, and symbols used in a mass appraisal system to estimate the value of a property.  Once a data element, square footage or quality for example, is identified as needed, its coding – or how it should be described in their mass appraisal system – is important to determine. Data element descriptions may be quantitative or qualitative. Quantitative descriptions, such as square footage, are based on counts and measures. This leads quantitative variables to be objective and more verifiable. If you and I both measure the square footage of a room, we should both get close to the actual area. How close, you may ask? That’s where the IAAO and NCDOR standards are useful to help us speak the same language. The IAAO Standard on Mass Appraisal of Real Property sets a standard within 1 foot (rounded to the nearest foot) of the true dimensions or within 5 percent of the area. For objective data other than measurements, the standard is that 95% of the coded entries should be accurate.

Alternatively, data descriptions can be qualitative, which tend to require judgement. Qualitative variables establish categories. Determining the correct category, which is usually done by a data collector or appraiser, tends to be more subjective. The goal in determining the best coding method is to make even the qualitative data as objective (without judgement) as possible. For example, building style is qualitative and could include categories such as ranch, split-level, 1.5 story, 2+ story, and others. So to make this data as objective as possible, an effective data collection manual is imperative. Does your county use an effective data collection manual? If not, work towards it. Among other things, a data collection manual should describe as objectively as possible all the characteristics of each building style so that when judgement is needed, most of the data collected will be accurate. How accurate? IAAO’s standard is that qualitative data should be coded correctly at least 90 percent of the time. Correctness may not be difficult to judge with variables such as heating type or building style, but becomes more difficult to keep accuracy for quality grade or effective age. So when determining codes for grade, for example, grades should be broad enough to not cause subjectivity errors between grades like B+++ and B++. I have heard discussions with statisticians and expert modelers include the opinion that at most, 8 categories are needed in a model to be effective for the quality grade variable. Regardless whether 8 or 48 grades are allowed in your models, each grade should include objective characteristics in the data collection manual descriptions, to enable multiple data collectors to choose the same category for the same property.

So that’s a brief review of the difference between quantitative and qualitative, objective and subjective data. How can that information translate to your budget request? In a later blog post, I’ll share how using objective measures can be much more effective than using subjective judgements in requests for staff positions, or other resources. The goal is the same that we’ve reviewed here. That is, turn qualitative data, such as grading your ability to get a job done with the staff you have (is the ability a B-or an A?), into as much of an objective decision as possible.