Wednesday, January 22, 2014

Sunnyvale, CA Population Density - II

Permanent Link:

This is part II, see also part I here:

In examining population density it may be helpful to compare it to the surrounding area to see if any conclusions can be drawn based on density.  Current Mayor Jim Griffith mentioned in the League of Women Voters' Oct. 2013 candidate's forum that he thought Sunnyvale should have 200,000 people to accommodate all the office space here.  I put that in the table and charts as Sunnyvale+50%.  Since so much more office space has been or is about to be approved since the election, I also included a doubling of Sunnyvale's population to accommodate all the new office jobs as Sunnyvale+100%. Here are selected cities around San Francisco Bay with their populations and density/square mile:

City or Urban Area Land Area in Sq. Miles Population Density per sq mi
San Francisco 47       825,111   17,620
Sunnyvale + 100% 22       280,162   12,400
Berkeley 10       112,580   10,752
Sunnyvale + 50% 22       210,122     9,300
Oakland 58       400,740     7,181
Alameda 11         73,812     6,956
Campbell 6         39,349     6,700
San Leandro 13         84,950     6,367
Santa Clara 18       116,468     6,300
Sunnyvale (2010)  22       140,081     6,200
Mountain View 12         74,066     6,172
San Jose (City) 177       984,299     5,576
Milpitas 14         66,790     4,900
San Jose Metro Area 448    1,894,388     4,228
Concord 31       122,067     4,000
Redwood City 35         76,815     3,956
Union City 19         71,763     3,700
Menlo Park 10         32,026     3,271
Walnut Creek 20         64,173     3,200
Newark 14         42,573     3,068
Palo Alto 24         64,403     2,500
Fremont 77       214,089     2,400
Saratoga 12         29,926     2,400
Hayward 45       144,186     2,300
SF Bay Area 6,984    7,150,000     1,023

The data is graphed below:
(Note Berkeley is denser than most but that may be because students live in very compact quarters.)

It is hard to see any correlation between population density and any urban characteristic such as safety, shopping availability, or public transport.  

BART serves Concord, Fremont and other cities in the East Bay which have much less density than Sunnyvale has now (2010 census).  It is not density but political will that is lacking to get BART to serve San Jose, Sunnyvale, and other nearby communities.

Does anyone want to seriously argue that increased density implies increased safety?  Oakland is denser than Palo Alto.

Is shopping better in dense Oakland compared to not-dense Palo Alto?

One can argue whether increased density is desirable or not but arguing that increasing density necessarily leads to better public transport, lower crime, or better or more convenient shopping should come up with some data confirming their hypothesis.  Nothing in the above data confirms a correlation between density and much of anything.

Tuesday, January 21, 2014

Sunnyvale CA Population Density - I

Permanent Link:

This is part I, See also part II:

In all the discussion about Sunnyvale's many new offices and growing apartment density no one has compared where we are with other cities near and far.  To have an informed discussion, it might be helpful to look at comparably densely populated cities.  I have included projections of Sunnyvale's growth per statements of Sunnyvale's Mayor Jim Griffith and the recent approvals of additional office space.  (C.f., discussion after the graph.)  See table below:
City Land Area in Sq. Miles Population Pop. Density per sq mi
Shanghai (Puxi) [3] 111      11,380,800     102,530
NYC - Manhattan 23        1,619,090       70,518
Paris (City) [2] 41        2,243,833       55,131
Hong Kong Island [4] 31        1,289,500       42,450
New York City 303        8,336,697       27,550
San Francisco 47           825,111       17,620
Tokyo  845      13,185,502       15,604
Mexico City 573        8,851,080       15,446
London  607        8,308,369       13,690
Boston (City) 48           636,479       12,900
Sunnyvale + 100% 22           280,162       12,400
Moscow [1] 970      11,503,501       11,865
Chicago  227        2,714,856       11,864
Washington, D.C. 68           646,449       10,528
Paris (Urban) [2] 1,098      10,413,386         9,484
Sunnyvale + 50% 22           210,122         9,300
Los Angeles 469        3,857,799         8,225
NYC - Staten Island 58           468,730         8,045
Seattle  84           634,535         8,045
Santa Clara 18           116,468         6,300
Sunnyvale (2010) 22           140,081         6,200
Mountain View 12             74,066         6,172
San Jose 177           984,299         5,576
San Diego 325        1,307,402         4,003
Dallas 341        1,241,162         3,518

[1] Moscow recently doubled it's area while only adding 233K people so it's former density is roughly 22K/sq.mi.
[2] For historical reasons, Paris proper is confined to a small area so I added the metropolitan area as more representative.  (Populations and land areas are from Wikipedia.)
[3] Shanghai has a huge area - roughly twice the land area of the counties of San Mateo, Santa Cruz and San Francisco combined.  To compare it to a US city area I look at "Puxi" which is the historic old center.
[4] Hong Kong includes Kowloon and is comparable in area to Santa Cruz County.  Most people thinking of Hong Kong are thinking of Hong Kong Island which is much denser than the entire administrative region.

The current Sunnyvale density of 6,200 per square mile (only counting land area) is comparable to immediately adjacent communities. 

Current Mayor Jim Griffith stated in the League of Women Voter's Candidate forum in October, 2013 that he felt given the amount of office space in Sunnyvale we should have about 200,000 people vs. the current 140,000 (2010 census).  I show that as Sunnyvale + 50% (210,000) in ORANGE.  That would put Sunnyvale's density above LA's and very close to the density of Paris's urban area.

A tremendous amount of office space was approved recently which will be built along Mathilda between the CalTrain tracks and Highways 101-237.  If apts and condos go up to accomodate population growth commensurate with office space, we may be looking at Sunnyvale having a population twice what it currently has,or around 280,000 - shown in RED above.  That would put us well above the density of Chicago, Washington, DC, and just a little below that of Boston and London.

A couple of interesting items:

1.  San Francisco is one of the most densely populated cities in the world.  Denser than Tokyo, Mexico City, London, etc., etc. (not including the water area of any city).

2.  Sunnyvale has almost exactly the land area of Manhattan.  So Sunnyvale could hold 1.6 Million people.  Would that make rents go down - i.e., is Manhattan known for low rents and readily available housing?  And where is our "Central Park"?

Michael Goldman

Thursday, January 16, 2014

Economic Danger Signs - 2014

Permanent Link:

Joke: a big investor asks his friends "How's business".  They smile and say "Couldn't be better!"  So, the investor sells all his stocks because if it couldn't be better - it can only get worse!

The current economic recover is pretty anemic nationwide relative to other historical recoveries and yet has some signs of (yet another) bubble.  Lost in all the discussion of increased development in Sunnyvale is whether all these new buildings will be filled.  It is quite a boom at the moment and after a boom you get an inevitable bust.  Some charts and discussion of where we are now below:

Standard 12-month Stock P/E (Price/Earnings ratio) going back to 1900 is in the upper range.
Chart of the Day

Yale economics professor Dr. Shiller is famous for his work on the Case-Shiller housing price index which is also back up:

(Begin update 1/22/2014) Are home prices in the SF Bay Area at unsustainable bubble prices?  I drew some crude trend lines on the above chart to determine very roughly the long term lower and upper bounds.  See below:

The lines diverge because the vertical axis is in normalized absolute values so a 20% drop in 1990-1995 doesn't look as dramatic as it does in 2008-2009.  The lines are crude but suggest that we may be at, or a little beyond, the upper bound for housing prices, though as 2005-2006 shows, they can certainly go higher. In the event of a decline, single family housing prices in south Sunnyvale will likely not drop as much as the overall SF Bay Area but they could take a while to rebound to their current value (the chart is NOT adjusted for inflation).  Note that someone buying an average house in the SF Bay Area at the peak in 1990 had to wait about 10 years for the price of their home to get back to what they paid for it - again NOT adjusted for inflation.  (end update 1/22/2014)

The most recent version of this chart for SF and 20 other cities is here:

Stock earnings adjusted for inflation are at 2007 credit bubble highs which were the highest since 1900:
Chart of the Day
Available at:

Yale economic professor Dr. Robert Shiller - best known for his book "Irrational Exuberance" predicting the 2007-2008 stock and housing collapse - won the 2013 Nobel Prize in Economics for his work in pricing assets (housing and stocks).  He said recently (December, 2013): "I am not yet sounding the alarm. But in many countries stock exchanges are at a high level and prices have risen sharply in some property markets," Shiller told Sunday's Der Spiegel magazine. "That could end badly," he said.  "I am most worried about the boom in the U.S. stock market. Also because our economy is still weak and vulnerable," he said, describing the financial and technology sectors as overvalued.

Dr. Shiller is famous among economists for his Cyclically Adjusted Price Earnings (CAPE) charts going back to the latter part of the 19th century.  His chart shows us very near bubble territory.  For more about Shiller, a graphic of his CAPE P/E chart, and a few other warning signs see here:

Shiller PE Ratio Chart
The most up to date Shiller P/E Chart:

Mid-term election years tend to be below average for stocks:
Chart of the Day

On the plus side, all the governments are on the lookout for any bubbles, and ready to act.  Whether they act sensibly or in time is another issue.  We could avoid a crash and just meander on in the current slow and anemic (some say jobless) recovery.

Monday, January 13, 2014

Why Rents Don't Decrease with Higher Density

The idea has been advanced that the way to lower rental prices is to allow much more building of housing (increasing density) to satisfy the demand.  This is not true and most economists acknowledge that rent prices are "sticky" - resistant to decrease. The technical term is "nominal rigidity".  Elementary economics texts talk about the law of supply and demand in general.  It clearly works for stock prices and food prices but the real world demonstrates this does not always work for housing prices and wages except under extreme conditions such as general economic collapse after a wild boom - a cure far worse than the disease.  Steve Hoffman pointed this out last city election when he noted that if increased density decreased rents then Manhattan, Hong Kong, and Shanghai would have the lowest rents around.  Most seemed to ignore his point as in "whom are you going to believe - me or your lying eyes?".

You can do a web search on "sticky rents" and "nominal rigidity" and get 9M hits.  One starting point is here:

David Ricardo in 1809 ("Law of Rent") showed how increased supply could increase costs here:

Translating this to modern Sunnyvale, a builder constructs apts. and notes the average $/sq. ft.  In the case of Sunnyvale he notes that rents are high and apparently going higher so he adds a little extra to his costs with granite counters and marble bathrooms and charges a premium for newness and luxury.  He has no incentive to build low rent apts since land and labor costs are the same as for luxury units.  If rents start to decrease *before* he builds, he will decide not to build since he has no way of knowing how low rents will go and will not risk losing money so supply will *not* increase indefinitely.  If he cannot fill his building *after* building (say only 60% rent out) he will not want to decrease rents because his other tenants will ask for a rent decrease themselves costing more than the cost of the unused space.  He will instead offer "move in incentives" such as 3 mos. free rent for a 2 year lease or something with the hope that by the time the 2 years are up, rents will have rebounded.  He can afford to wait until a recovery because he has "deep pockets" and can take temporary losses.  If he does not have "deep pockets" and needs money he will have to sell to someone with deep pockets so eventually all landlords left will have "deep pockets" and can wait out a year or two of recession.  In a very prolonged local slump, rents *will* decrease but so will job opportunities and wages so you may not be able to afford rents anyway - e.g., you may not want to live in Detroit.

Here are some fully accessible economic papers acknowledging "sticky rents" (most others require a Journal subscription):

"Sticky Rents and the Stability of Housing Cycles"

"Sticky rents and the CPI for owner-occupied housing"

Nobel Laureate Paul Krugman wrote a popular intro economics text that covers the standard law of supply and demand. But, in his NY Times blog he continually talks about "sticky wages" in arguing for moderate German inflation.  The idea is to effectively decrease "relative" real wages in depressed low-productivity economies in Europe such as Greece and Spain to get people there working again.

(added 1/16/2014)
We did have a general economic collapse in 2007-2008 and what happened to housing costs was instructive.  The price of individual homes dropped by 50% on the very low end as a great many went on sale - condos and small townhouses in less desirable neighborhoods - as people who were barely able to afford housing lost their ability to meet balloon mortgages.  Banks actually didn't always take over houses in default because increasing the supply of defaulted housing would lower prices even further accelerating a downward spiral costing them more than taking a monthly loss on mortgage payments - until a recovery.  Housing prices in the upper middle class areas dropped about 20% but there was very little supply.  People in those houses could mostly choose to wait out the downturn.  Rents actually increased(!) because people have to live somewhere and if they can't afford a house, they will rent so the demand for rental housing actually increased and with them, rental prices.