In our industry, we are constantly barraged with information, statistics, data sets, and the like and it can be a chore to work through the seemingly unending torrent of information that comes at us. The "Information Age" was supposed to make us all smarter, better informed, and more educated in our decision making process.
However, the issue with this concept is that it assumed the information we were to have at our fingertips would be quality information. Unfortunately, with the 24/7 news cycle, the internet, blogs outnumbering people (OK, bit of an exaggeration there), wiki-this & wiki-that, and all of the sundry other sources of information out there, the quality component has suffered greatly. I have long felt that there has been a lack of critical evaluation of all of the data put forth (hence the incredible growth of sites such as Snopes.com and FactCheck.org).
For those of us in the real estate industry, the issue is just as relevant (if not more so). This was brought home once again by an innocuous article put forth recently in a Trulia blog where they introduced the "Trulia Rent vs. Buy Index." Always tempted to read about the latest index about our industry, I took a look. On the surface it appears to be a straightforward ranking of the rent vs buy outcomes for various US cities. But, as I read the article the statistical geek in me had serious questions about this report/index.
As stated by Trulia, the ratio was calculated "...using the average list price compared with average rent on 2 bedroom apartments, condos and townhomes listed on Trulia.com." Oh where to start with the concerns?
First using "average list price" as the basis for calculating the "buy" side of the equation? Wouldn’t the sale price be a more appropriate measure? Also, everyone knows using averages for a broad statistical group such as housing inventory is at best risky. Median prices would be a much better indication, especially when dealing with an entire metro area inventory (the same can be said for the rent component).
Second, the comparison is based on "…2 bedroom apartments, condos and townhomes listed on Trulia.com." So, this isn’t a market based inventory calculation but rather a subset based exclusively on Trulia’s data set. A couple of questions: a) what is the size, per metro area listed, of the data set used and b) how does that compare to the aggregate market inventory? Are we dealing with sample sizes of just 15 properties or 1,500? Is the Trulia data set equivalent to 2% of the aggregate market or 65%? These are critical data points in determining the value of this information.
Third, is the average list price only for "…2 bedroom apartments, condos and townhomes on Trulia.com" or is it for all listings on Trulia for that metro area. The sentence doesn't clearly articulate this, so the reader is forced to assume one of two potential scenarios: a) the average list price and rent figures are both based on 2 bedroom apartments, condos and townhomes, or b) the rent is for 2 bedroom apartments, condos and townhomes while the list price is for all properties. This simply gets to the question "are we comparing apples to apples or apples to steak?" And really, in either scenario are we really dealing with anything that can provide a substantive aggregate market assessment for these metro areas? Scenario A is such a small sub-section of a real estate market is it worthwhile to try to project it over the entire market? Scenario B is just not worth anything from a statistical point of view.
I incorrectly assumed that somewhere in this article we would see the disclaimers providing the definitions and characteristics of the sample sizes involved in making the calculations. What was the minimum sample size criteria to "make the list" (they simply state in the actual downloaded report that Fort Worth had an “insufficient” sample size – but they still reported a ratio for Forth Worth).
Now on to the "Interpretation Key" for the "Price-to-Rent Ratio." They explain the Price to Rent Ratio bands of 1-15, 16-20, and 21+. But there is no definition or explanation on how these ranges were established. Why is the split between 15 & 16 critical? Why not 11 or 19?
There were so many statistical holes in the data I wasn’t going to give it much of a thought, until I read the following comment by a reader: "…I expected to see San Diego on the Rent v Buy list, but like you am surprised that Omaha, Oklahoma City and Portland outprice our balmy environs. And Cleveland is just behind us??? Interesting numbers!" So, the information age has brought us information, but clearly lagging is our ability to critically assess the quality of the information put forward. It does highlight the concept that if it’s on the internet, it must be true!!!!
As a professional in this industry, I find it disturbing that all the data assessments I do with full regard to disclosure and proper statistical evaluation techniques can be easily discarded by a consumer as they read articles on the internet that may have a contrary perspective but no delineation of where that data originated.
It reminds me of a listing appointment a few years ago where the homeowner was offended by the estimated market price I had developed as he knew his home was worth far more than that because the newspaper said "average prices" for the metro area were up 16%. I guess he didn’t factor in the fact he was located on a large, arterial highway in the middle of a well established and well known drug trafficking area and whose property line was shared with an abandoned industrial warehouse. Clearly, the newspaper knew better……..
The information age has highlighted the importance of assessment skills related to the reams of data available. The question is, does the consumer value those assessment skills?
_________________________________________________________________________
Craig Frazer, Realtor, CRS, GRI, CLHMS
RE/MAX Metro
Cell & Text: (801)699-6046
Email: cfrazer@remax.net
Community Data Sets
My Blog
Just Data, Info & Advice -- No Sales Pitches