An "!"Cost ME $10,329.44? Decoding Real Estate Prose

In Chapter 2 of Freakonomics (which is salciously titled How Is the Ku Klux Klan Like a Group of Real-Estate Agents?), Steven Levitt takes on real estate agent word choice, asserting that, sometimes, real estate agents obscure negative attributes with vague terms when describing a listing. He offers that these ambigious terms and punctuation are actually predictive of price and provide signals to buyers of the property's quality.

Below is his list of words that he identified as predictive of the sold price of a home:

Negatively predictive Positively predicitive
Fantastic Corian
Great neighborhood Granite
!!! Maple
Spacious State-of-the-art
Charming Gourmet

Levitt separates between these lists by explaining:

Three of the five terms correlated with a higher sales price are physical descriptions of the house itself: granite, Corian, and maple. As information goes, such terms are specific and straightforward — and therefore pretty useful. If you like granite, you might like the house; but even if you don’t, “granite” certainly doesn’t connote a fixer-upper. Nor does “gourmet” or “state-of-the-art,” both of which seem to tell a buyer that a house is, on some level, truly fantastic. “Fantastic,” meanwhile, is a dangerously ambiguous adjective, as is “charming.” Both these words seem to be real-estate agent code for a house that doesn’t have many specific attributes worth describing. “Spacious” homes, meanwhile, are often decrepit or impractical. “Great neighborhood” signals a buyer that, well, this house isn’t very nice but others nearby may be. And an exclamation point in a real estate ad is bad news for sure, a bid to paper over real shortcomings with false enthusiasm."

As a real estate brokerage, it is important to us to see which of these words resonate in the Abilene market.  Too, it is instructive to recognize terms like “cozy” and “Corian” as subtle signals of quality that, no matter what you call it, implicitly reveal value.

To regionalize Levitt’s analysis, I looked at his same terms to gauge how predictive each were regarding the sold price of the home.  I used listings for 4,551 home sales for in Abilene and 3,786 home sales in Dallas to assess the information provided in agent-drafted property descriptions.  To control for obvious factors that predict sold price, square footage and age were included.  A graph of the relationship between age and sold price was assumed to be U-shaped – a new house sells for its best price and rapidly loses value until it hits a point when the landscaping matures and the house becomes historic, causing price to climb once it crosses that threshold.

Using this data, the results of an ordinary least-squares regression revealed value by the use of these terms in the property’s description.  Terms like maple and "state-of-the-art" (or, just as often "state of the art" (sic)) had no statistical contribution to the story and were dropped.  A few additional terms like "cute" and "cozy" were added just for my curiousity.  

The terms in blue are those terms that are statistically significant and reveal something about the value of the sold home in Abilene and Dallas markets.

Conclusions to draw from these results include:

  • Square footage drives price – one more square foot adds $246.04 to price in Dallas and $92.94 to price in Abilene
  • Age, itself, drives down price by $1,674.42 per year in Dallas and faster in Abilene where houses lose $2,233.69 per year.  If enough time passes, the squared age of the property shows Dallas properties recover historic value, roughly, 4 times faster than in Abilene. One could view this two ways: Abilenians like things new and shiny or we just don’t have enough quality, aged properties to where it matters when it comes to pricing older real estate.
  • In Abilene, attributes like hardwood floors (+$39,325.44), gourmet kitchens (+$19,731.83) and granite countertops (+$11,572.83) matter. Together, they add, roughly, $70,000 to the price of a house.
  • In Dallas, attributes like hardwood floors (+$50,568.40), gourmet kitchens (+58,029.47) and stainless appliances (+$28,880.38) matter. Together, they add, roughly, $137,478.26 to the price of a house.
  • Countertop choices provide insight into tastes in these markets and dates Levitt’s work. Evidently, when he first wrote Freakonomics, Corian countertops had meaning, but time appears to have blunted or reversed their contribution to price. In Abilene, Corian countertops are statistically meaningless with regard to price. In Dallas, Corian countertops translate into a $98,559.50 drop in price which approximates the cost of gutting a dated, 1990’s kitchen. In Abilene, granite adds $11,572.83 to the price of the house, but in Dallas granite is statistically meaningless to price. It would appear that Dallas buyers have an appetitive for new, more exotic solid surfaces in their kitchens.
  • Terms about design like “cozy” or “spacious” negatively impact price in both markets. “Spacious” homes are less desirable, indicating by a drop in price by $10,036.16 in Abilene and $25,652.72 in Dallas. On the other hand, Abilene homes described as “cozy” lose $7,244.87 in value. This suggests that a savvy Abilene home builder should strive for something that strikes the balance between these terms.
  • Hands-down, my favorite agent superlative is the use of “!” in the listing description. In Abilene, “!” is not statistically price-punitive, however in Dallas, the liberal use of exclamation points by the listing agent should ready the owner for a lower price. Each ! employed indicates a drop in price by $10,329.44 – now that’s pricy punctuation (that acknowledged, your author feels the extravagance is warranted in closing this sentence with the aforementioned punctuation mark)!
  • The fact that ! has no statistical bearing on Abilene prices, but does in Dallas brings me to my last point: Abilene buyers have a taste for homes that can be described using vague superlatives; Dallas buyers do not. In Abilene, a house that can be described as “cute” (+$8,249.20) and “charming” (+$10,297.22) stands to sell for $18,546.43 more. In Dallas, homes that meet this description realize no gain in price.

As an economist, the ease by which “!” can be tabulated makes it an appealing measure of the Panglossian spirit of the real estate agent. Using the data employed for this estimation, I calculated what I’ll call the effusivity index: the number of “!” in a listings by each brokerage divided by the number of sold properties by that brokerage.  The results are entabled as follows and ranked from most to least effusive:  

sold listings
effusivity index
21 7   3.00 Carlson Properties
1621 721   2.25 Coldwell Banker Panian and Mash, REALTORS
152 71   2.14 Red Apple, REALTORS
27 13   2.08 Fathom Realty
2 1   2.00 Geneva Dodson, REALTORS
2 1   2.00 Pennington Premiere Real Estate
192 110   1.75 The-Arnold Group, REALTORS
82 51   1.61 Sendero Properties LLC
14 10   1.40 Tommy Milliorn, REALTORS
179 136   1.32 Keller Williams Realty
18 14   1.29 Augusta, REALTORS
1137 938   1.21 RE/MAX OF ABILENE
46 45   1.02 Dalzell, REALTORS (S. Danville)
204 201   1.01 Stovall, REALTORS
5 5   1.00 Creekview Realty
3 3   1.00 Trinity Liveoak Real Estate, LLC
1 1   1.00 Legacy Realty
1 1   1.00 Texas Goldstar Real Estate
1 1   1.00 The Michael Group Real Estate
9 10   0.90 TriWest, REALTORS
11 13   0.85 Beckendorf Realty Group,LLC
305 382   0.80 Senter, REALTORS
5 7   0.71 Shade Real Estate Group, LLC
4 6   0.67 Alliance Realty Group
3 5   0.60 Excess Realty
12 22   0.55 Homefinders Plus
6 12   0.50 Jimmy Partin, REALTORS
5 10   0.50 Abilene Hallmark Real Estate Service LLC
11 24   0.46 TruStar Real Estate, LLC
26 59   0.44 Diamond Properties
3 7   0.43 Mission Real Estate Group
24 60   0.40 Dalzell, REALTORS (Radford Hills)
3 8   0.38 Key Property Management
1 3   0.33 Chris Barnett Real Estate
1 3   0.33 Peterson Real Estate
7 31   0.23 Beall & Company, REALTORS
1 5   0.20 Guardian Real Estate
3 20   0.15 Hatchett & Company Real Estate
4 34   0.12 The Real Estate Store LLC
0 21   0.00 Reunion Realty
0 5   0.00 ACE Property Solutions
0 3   0.00 Property Service Specialties
0 3   0.00 Tejas Properties
0 2   0.00 Barnett & Hill
0 2   0.00 Coates Real Estate
0 2   0.00 Foster & Company
0 2   0.00 Franklin Real Estate
0 2   0.00 Joe Lopez, REALTORS
0 2   0.00 Key City Realty
0 2   0.00 Rosanna's Realty
0 2   0.00 Willis Real Estate
0 1   0.00 Bill North Real Estate
0 1   0.00 Castereno, REALTORS
0 1   0.00 Gerard Real Estate, LLC
0 1   0.00 Jerry Taylor Real Estate
0 1   0.00 Moore Property Group Management, LLC
0 1   0.00 Railhead Realty, LLC
0 1   0.00 Regional Real Estate
0 1   0.00 Robert J. Russell Real Estate

An effusivity index of 1 indicates an “!” per listing and an effusivity index which is greater than 1 indicates a love of the exclamatory with multiple “!” per listing. Several brokerages scored 0 indicating no “!” used in any property description.  It would be unfair to suggest restraint, as almost all of these had 1 or 2 listings and this brings better odds when avoiding “!”. That said, it is worth recognizing that Reunion Realty seems the most steadfast in their resistance of the siren song of the “!” – they had 21 sold listings without using a single “!”.