The Living-Area Adjustment and the Ratterman Method

The size adjustment is made on nearly every comparable, and it is commonly derived incorrectly, in a way that can overstate size differences by a wide margin. The method used to derive it, and the reason the method is built the way it is, are worth understanding in full.

A house that sold for $500,000 with 2,000 square feet of living area has an average price of $250 per square foot. That figure includes the land, the garage, the kitchen, the roof, the location, and everything else the buyer paid for, spread across the living area. It is not what the market pays for one more square foot of living area, because one more square foot does not come with more land, another garage, or a second kitchen. An extra square foot of a house does not add to the price at the same rate as the average cost per square foot of the whole property. Applying the average as though it were the adjustment produces a size adjustment several times too large: on a 200-square-foot difference at $250 per square foot the adjustment is $50,000, when the market’s payment for the additional space in that segment may be a quarter to a third of that.

Why size and price do not move together

A 1,800-square-foot colonial and a 2,200-square-foot colonial on the same street each have one lot, one kitchen, one set of mechanical systems, one driveway and one water and sewer connection. The additional 400 square feet in the larger house is mostly bedroom, family room and hall space, which costs less to build than a kitchen or a bath and comes with no additional land. Buyers pay for that space, and they pay for it at a rate well below the average of the whole property. The relationship also tends to flatten toward the top of a segment: a buyer who pays for 2,400 square feet over 2,000 does not necessarily pay at the same rate again for 2,800. Price rises with size, but not in proportion to it, and that is the reason an average price per square foot is not, by itself, a pricing method or an adjustment rate.

Establishing the difference first

The adjustment is the rate multiplied by the difference in living area, and the difference is only as reliable as the records behind it. Living area is the finished, above-grade area as defined by the ANSI Z765 standard. Finished basement space is excluded and compared in its own category. Gross living area for the subject and each comparable is generally taken from the best available record (the assessor’s card, the public record, the MLS listing, builder plans or a prior appraisal), with the source noted. Where sources disagree, the difference is noted, not guessed at. The ANSI standard is taught so the agent knows how the appraiser measures and why the appraiser’s figure may differ from the record. A 150-square-foot error in one comparable’s living area, at a rate of $70, moves that comparable’s indication by $10,500, a material amount on one sale.

The Ratterman Method. The method in general use for deriving the living-area adjustment from the sales at hand has three steps.

First, the average sale price per square foot is found across the comparable sales chosen for the analysis: each sale’s price divided by its living area, averaged.

Second, that average is multiplied by a factor of between 25 and 33 percent. Thirty-three percent is the traditional benchmark. The factor exists for the reason above: it strips out the share of the average that pays for land, site, improvements other than living area, and location, leaving an estimate of what the market pays for the living area itself.

Third, the result is the dollar-per-square-foot rate applied to the difference in living area between the subject and each comparable.

Worked: three colonials in the subject’s segment sold at $500,000 for 2,000 square feet ($250 per square foot), $520,000 for 2,100 ($248), and $472,500 for 1,890 ($250). The average is about $249. At the traditional factor of one-third, the living-area adjustment rate is about $83 per square foot; at 25 percent, about $62. A comparable 200 square feet larger than the subject is adjusted downward by roughly $12,400 to $16,600, not by $49,800.

The same three sales, measured against a 1,950-square-foot subject, show how much the choice within the range matters and how much the omission of the factor matters. The first sale is 50 square feet larger than the subject, the second 150 larger, and the third 60 smaller.

The third sale’s 60-square-foot difference is adjusted upward by $3,720 at 25 percent and $4,980 at one-third, because the segment’s sales show buyers paying for living area at that rate; differences between the subject and the comparable sales are generally adjusted, except when the difference is too small to quantify or the sales show no market reaction to it. Moving from 25 percent to one-third shifts the second sale by $3,150, the first by $1,050 and the third by $1,260; the factor matters, but it does not reorder the evidence. Applying the average whole moves the second sale by $37,350, takes it from the top of the set to the bottom, and turns the third sale’s adjustment of about $3,700 to $5,000 into $14,940.

Choosing the factor. The factor is not fixed. It reflects how much of the average price is land and location rather than building, and it moves with the local land-to-value ratio. Where land is a large share of the price, as in a high-priced neighborhood or a waterfront market, the share of the average that belongs to the living area is smaller and the factor sits toward the low end or below it. Where land is inexpensive, as in an exurban subdivision, the building is a larger share of the price and the factor sits toward the high end. A broker who knows the segment’s land values knows which end of the range applies.

Reading the land-to-value ratio

Connecticut assessor’s field cards list the land and the building separately. Assessments are set at 70 percent of market value as of the town’s last revaluation, and that percentage applies to both figures, so the land’s share of the total assessment is a rough indicator of the land-to-value ratio in a neighborhood. Field cards for a dozen properties in the segment show whether land is typically a fifth of the total or half of it. Sales of buildable lots, and sales of houses bought to be torn down and replaced, show the same thing directly where they exist. Neither source is exact, since revaluations age and assessors’ land schedules are broad, and neither needs to be: the ratio places the factor toward one end of the range or the other, and the sales confirm it.

A shoreline segment. Three capes and colonials near the water sold for $780,000 at 1,850 square feet ($421.62), $835,000 at 2,050 ($407.32) and $760,000 at 1,780 ($426.97). The average is about $419. Field cards in the neighborhood show land at roughly half of the total assessment, which points to the low end of the range: at 25 percent the rate is about $105, against about $140 at one-third. Two capes on the same stretch, 1,800 and 2,000 square feet and otherwise similar, sold within two months of each other, after time correction, for $790,000 and $812,000. The pair indicates $22,000 for 200 square feet, or $110, a factor of about 26 percent. A comparable 150 square feet larger than the subject is adjusted downward by $16,500 at $110, not by $21,000 at the traditional one-third.

An exurban segment. Three colonials in a subdivision on former farmland sold for $395,000 at 1,900 square feet ($207.89), $415,000 at 2,050 ($202.44) and $380,000 at 1,800 ($211.11). The average is about $207. Field cards show land at about a fifth of the total, which points to the high end: at one-third the rate is about $69. A pair of 1,750- and 2,000-square-foot colonials in the subdivision sold for $372,000 and $389,000, indicating $17,000 for 250 square feet, or $68, which confirms the one-third factor for that segment.

The two segments produce rates of $110 and $68 from averages of $419 and $207. The difference in the factor, as well as the difference in the averages, reflects the market, and neither rate is carried into the other segment.

What the method is, and is not. The Ratterman Method is a starting point, and it is used as one. It is a rough figure derived from the sales already in hand, and it is confirmed or corrected by the market where the market permits: by paired sales that differ mainly in size, where the price difference divided by the size difference gives a direct reading of what the segment paid for the additional space; by testing the range at two or three plausible rates to see how much the conclusion depends on the choice; or, in an office that works with enough sales, by regression across the segment. Where a pair of sales differing mainly in size gives a rate well away from the Ratterman figure, the pair is the better evidence, and the factor was wrong for that segment.

Other property types and limits

In a condominium complex there is no individual lot, but the same logic holds: the average includes the unit’s share of the common elements, the location and the parking, none of which grows with the unit’s size. Complexes with several floor plans often supply pairs of units on the same floor that differ mainly in size, and those pairs are the best confirmation of the factor. In a new-construction development, closed sales of different models on similar lots are often the cleanest size pairs available, once the options in each sale are accounted for. Where the difference in living area between subject and comparable is large, beyond the size band within which the comparable was competing for the same buyer, the rate becomes less reliable at the edges, because the relationship between size and price is not a straight line; the answer is a closer sale rather than a more carefully chosen rate.

How it applies. A size adjustment in a grid is checked with one question: was the rate derived by the method above, with a factor, or was the average price per square foot applied whole? The second is recognizable by its size. A rate near the segment’s average price per square foot, in the hundreds of dollars in most Connecticut suburban markets, is generally an average that was applied without the factor, and a grid built on it has every size adjustment wrong by the same multiple. A rate in the tens of dollars, with the factor stated and a pair or two behind it, is a rate the office can rely on.