June 26, 2007
A wrong use of statistics – California Housing Affordability Index
I don’t mean for this to become a real estate blog – there’s plenty enough already, but if there’s one thing that bugs me, it’s the incorrect use of statistics.
Recently I received a well intention e-mail that was a forward from Joe Brown, President of Coldwell Banker Residential Brokerage, Silicon Valley~Monterey Bay. A full version of the e-mail can be found here and other real estate websites, but here’s the part that really bugged me:
We are part of an area where jobs are very strong and the ability to buy a house is high. The housing affordability index is now at 25% for California (up from 14% no so long ago). This is greatly due to prices decreasing in outlying, less desirable areas of the state. More particular to our local areas would be folks who are deciding that it is time to invest their fortunes in real estate. When one gets a roof over his/her head, stability, and a tax write-off, sooner or later one realizes the excellent investment that real estate is.
I’ve bolded the part that I want to discuss. For those of you who aren’t familiar with why housing affordability for first time buyers is important, think of housing like a pyramid: first time buyers have to enter the market, so that people who own can sell their properties, and use that money to buy (presumably upgrade) to a bigger/better property. If the housing affordability index too low, it means that there aren’t enough potential first time buyers, which means the pyramid is endanger. Think of first time buyers as plankton (the graphic at the top btw.)
Ok, so now that we know why this index is important, let’s look at the numbers. It is claimed that affordability in California has gone up from 14% “no [sic] so long ago” to 25% now. Pretty dramatic, huh? This is excellent news right?
Well, let’s take a look at the press releases from where this data is derived:
| C.A.R. 2/9/2006 | C.A.R. 5/17/2007 | |
| Calculated: Affordability Index | 14% | 25% |
| Assumption: Downpayment % | 20% | 10% |
| Assumption: Purchase price | Median Price | 85% of Median |
| Assumption: Interest rate | 6.33% | 6.3% |
That’s right – in the two time periods, the assumed downpayments went down, the purchase price went down, and interest rates went down as well. Hence, affordability went up dramatically.
To explain why that is, here’s the verbatim text from the press release for 2006:
The minimum household income needed to purchase a median-priced home at $548,430 in California in December was $134,200, based on an average effective mortgage interest rate of 6.33 percent and assuming a 20 percent downpayment.
…and for 2007:
The minimum household income needed to purchase an entry-level home at $480,670 in California in the first quarter of 2007 was $96,910, based on an adjustable interest rate of 6.3 percent and assuming a 10 percent down payment.
Ah hah. You see, in 2006, it was about purchasing a “median-priced home”, but in 2007, it is now about a “entry-level home”.
Technically there’s nothing wrong with changing the assumptions. That’s fine.
But is it really right to make this comparison?
The housing affordability index is now at 25% for California (up from 14% no so long ago).
3 Comments to “A wrong use of statistics – California Housing Affordability Index”
June 26th, 2007 at 10:46 am
You are more correct than you give yourself credit for! Men like the one who sent you that letter either should know better or should not be in the position that they currently occupy. It is blatant manipulation and dishonest sales policy to prop up the market outlook with an apples to oranges comparison.
Not only is it NOT right to make that comparison, but it is the grade A, # 1 reason why a good portion of our population detests my colleagues in the Real Estate profession. Our practitioners are advised by officially licensed REALTOR newsletters to make changes in the presentation of facts. Changes like moving away from presenting month to month statistics and to presenting yearly statistics (we never should be using monthly statistics for such a long term investment… unless we LIKE fueling speculation). And changes like stressing the tax benefits over the appreciation rate. We are all selling something, and that is OK in my book. But when facts are deliberately misrepresented, some people might even call that a lie…
June 27th, 2007 at 7:25 pm
The famous saw: “There are three kinds of lies: lies, damned lies, and statistics.”
I’m impressed you bothered to track down why the claim felt wrong…
Also, reminded me of recent reflections on the mindset of a person who wants to get things right, and one who wants to win. (Was recently comparing notes on this with an engineer who had gone into patent law, and it seemed that the way he approached the work was pretty different than most i.p. attorneys I’d dealt with)
Matt
January 25th, 2008 at 9:56 pm
I never know if the agents are idiots or liars… and I’m not sure which is worse.
In either case, while you are right to point out an apparent “wrong use of statistics,” maybe the title out to be “wrong use of language.” (Both really.)
Go beyond the surface of the word choices and look at the numbers. There are two things to notice.
First, I assume we can all agree that the median home price has come down between 2006 and 2007 when the two reports were issued. Therefore you would expect the home price for 2007 to be less – especially if this “entry-level” home is 85% of the median. However, 85% of the median home price of $548,430 (2006) is $466,165.50. Clearly, that value is less than the supposed $480,670 home price for 2007. The words do not match the numbers.
Second, if we can assume for the moment that anything in the email is meaningful, then let’s suppose they got the 20% down (2006) and 10% down (2007) correct. That means we have loans of 80% and 90% respectively.
$548,430 * 80% = $438,744
$480,670 * 90% = $432,603
The loan amounts are practically identical. So, for affordability to have jumped from 14% to 25% interest rates must have dropped substantially or everyone got a huge raise.
Code should be self documenting. The numbers should be able to speak for themselves and any language is just fluff for presentation purposes. Language is tremendously important, but you’ve got to look beyond that to see if the data makes sense. And in this case, nothing in the email makes sense as presented.
SanDiegosFinestRealEstate.blogspot.com
http://www.SanDiegosFinestRealEstate.com
Leave a Reply