Apartment Rent: Median vs. Mean

September 6, 2007

Sunnyvale Apartment Rents

Are you tired of hearing "spin" on the radio or in the newspaper and then finding they lacked, or even contradicted, actual data. View from Silicon Valley exists in large part to bridge this gap.

Readers occasionally send us publicly-available data. We appreciate it! Anybody with data they feel would be useful or interesting, please send it in!

Today's reader-submitted data is from the city of Sunnyvale's semi-annual survey of apartment rental rates and vacancies. Sunnyvale's plan is to use this data to:
"(1) to establish the rental rate for new Below Market Rate (BMR) rental units,
(2) to adjust rental rates on existing BMR rental units and
(3) as a factor in determining whether or not some rental properties may be converted to condominium ownership."

We'll save our rant about the government meddling in the rental markets and presuming to restrict private property usage for another day.

The "interesting" part is Sunnyvale contracts with Realfacts to conduct a semi-annual vacancy and rent survey of ~15,000 rental units which they define as "representative of all rental units in Sunnyvale." (Who knew there more than 15,000(!) apartments just in Sunnyvale?!?)

The July 2007 survey found 2.27% vacancies, down from 2.37% in January 2007 vacancy rate and 2.99% in July 2006.

Based on this data, the city authorized a 5% BMR increase.

In a feat of government double-speak, the report references the low vacancy rate not to forbid condo conversions and instead said:
"On May 9, 2006, the City Council amended Municipal Code Section 19.70 to eliminate the restriction on conversion of apartment units to condominiums when the vacancy rate was 3% or less. In lieu of this restriction, a 175-unit limit was placed on conversions in any given 12-month period."

In a victory for the free market, Sunnyvale seems to be conceeding they are unable to stop condo conversions. Even larger complexes can be approved if the owner "proves" the conversion will not adversely affect the rental market. Given, the government can "prove" there is no inflation and that we have full employment, private enterpise should have no trouble with this low hurdle.

From our perspective, the really "interesting" data was in the graphs included with the report.

First up was "Median Rent Graph":

One difference between the "median" and "average" data is the average failed to show the same "pop" seen in the median data. Lacking this pop, the averages barely reached July 2002 levels.

Housing bulls and bears have a running argument about the significance (or insignificance) of median sales price data when sales volume declines. However, apartment leases roll over on an annual basis. Except for "losing" the apartments converted to condos, apartment data is culled from the same "volume" every year.

What does the out-performance of median apartment rents over average apartment rents actually mean?

We submit it's the preference of renters for "nicer" apartments in lieu of buying a house. It's a sign more people who might normally buy are instead renting.

If so, this trend could be expected to continue until real estate sales volume picks up. Any re-convergence of apartment rental median and mean prices might be a leading indicator for when housing sales volume is about to recover...

Time will tell...