Saturday, September 8, 2018

Ree-diculous Ree-na - Part II


Ridiculous RHNA ("Ree-na")

Part II

In Part I we saw how ridiculous is the current application of RHNA (Regional Housing Needs Allocation) numbers to housing in terms of Scott Wiener's SB-35 law.  Part I is here:

In this, Part II, we see what RHNA numbers are actually for.

"Not Meeting RHNA Numbers"

(I.e., "A city isn't building enough")

The California State Dept. of Housing and Community Development (HCD) is the agency in charge of ensuring that there are "housing elements" of the General Plan for each jurisdiction.  This is to ensure that zoning and guidelines are in place so that housing for different incomes can be built should a builder want to.

More about HCD and RHNA here:

The narrative of "housing activists" (some adopt the name "YIMBY") is that cities, by refusing to build housing or allowing others to build housing, are creating an artificial shortage of housing thereby driving up the price of what housing there is.  There are two problems with that narrative.

Problem One... that cities can't build housing.  Only builders can build housing.
Cities don't do this.
Builders do this.
If you want more housing, go build some.  Buy the land, hire architects and workers and build it.  As long as you follow all the various fire codes, construction codes, and zoning rules then you can build what you want.  You will go before a planning commission to make sure you follow the rules and then you start construction.  The city council will look over your plans only if you want a waiver from some rule.

Holding cities and counties responsible when they have no control over who builds or doesn't build housing - is irresponsible.  In particular, holding cities responsible for not building a specified quantity of housing for a particular income level makes no sense because cities don't build housing.  There is an issue if cities stand in the way of builders building housing but that has nothing to do with RHNA.  We will hold that discussion off to Part III.

Problem Two... that the RHNA numbers make no sense in relation to the "housing crisis".  As we saw in Part I, some cities and counties are very poor choices for housing development, and contribute nothing to the current "housing crisis".  Yet, because "they didn't make their RHNA numbers" those cities and counties lose local control over planning and zoning.  Of the 58 counties in California, 47 "didn't make their RHNA numbers" for reasons that have nothing to do with resident objections.  Nonetheless, they are subject to what are becoming increasingly draconian state laws over-riding local control.

Not a Goal!

RHNA is a planning tool.  It is not a "goal" - though you will hear that word used frequently.
You can get the complete HCD list of RHNA spreadsheets or PDFs here:

If no one builds the housing in the RHNA spreadsheet, all it means is that no one built housing there.  It isn't a failure to meet a "goal".  RHNA numbers aren't "goals" - they are requirements for planning.

But "goals" gets repeated over and over again implying cities or counties aren't trying hard enough.  If only cities would just get on with it and do something!  But what?  If builders decide not to construct apartments or houses in the quantity desired, no city or county can force them to.  Here are some newspapers reporting as "goals" what are really more like "zoning allotments".

Misuse of RHNA Numbers

"San Jose Mercury News"
CA State Senator Scott Wiener (author of SB-35) said “When 97 percent of cities are failing to meet their housing goals, it’s clear we need to change how we approach housing in California.”

To be fair to the SJ Mercury News, they are not calling them "goals", they are merely quoting Senator Scott Wiener who is calling them "goals".  Even though they are not "goals".

"San Mateo Daily Journal"
"City exceeds housing goals". Except they aren't goals.
Above from:

"The Bay City Beacon"
"According to the Department of Housing and Community Development (HCD), 97.6% of California cities did not meet their full RHNA goals."  (Like, Biggs, Alpine County, etc., looked at in Part I.)  "Only 13 cities (2.4%) met their goals in full, and would be exempt from SB 35 streamlining."from:

This is the fault of the Bay City Beacon reporter who fell for the "goals" concept.  Nowhere can I find HCD calling them "goals".

"Los Altos Politico"
"A close look the City of Los Altos Housing Element report suggests, there is not enough land zoned such that we can meet the goals for lower-income housing.  The only good news is that we do produce enough market-rate housing.  And that’s because to the State’s way of thinking, a tear down and rebuild of a single family home counts as a “new” house!"

The amount of housing stays the same, but tearing one down and rebuilding it counts as an additional housing unit?  Now the housing numbers make even less sense.

Official Description of RHNA Objectives
(Which Are NOT Goals!)

Interestingly, official documents about RHNA numbers never use the word "goal".  That is a word only used by people like Senator Scott Wiener and repeated by some newspapers - possibly merely quoting Senator Wiener.

What does California's Housing and Community Development (HCD) Agency say about RHNA?  The main information web page for HCD is here:

Under "Background" they say that because funding for housing usually requires "a compliant housing element" there needs to be a well specified description of what is zoned for.
Fairfax, Virginia's zoning map.  Every city has to do this.
Cities aren't responsible for building anything.
The number of housing units that must be planned for (including zoning but other aspects are considered, too) is allocated on a regional basis by HCD.  The regional Council of Governments (like ABAG in the SF Bay area) allocates it within the region.  They have formulas they use but they are allowed a certain amount of flexibility so it isn't a hard and fast set of rules.

HCD's website states under "Housing Elements" that "Since 1969, California has required that all local governments (cities and counties) adequately plan to meet the housing needs of everyone in the community. California’s local governments meet this requirement by adopting housing plans as part of their “general plan” (also required by the state)."

Got that?  A plan is needed.  Once a plan is in place, HCD is doneThe RHNA requirement has been satisfied once a plan is in place.  If no builder decides to build, well - too bad - or not.

Here are some local California Council of Governments (CoGs) explaining what RHNA numbers mean.

Tulare Council of Governments
Tulare, CA
"Allocation targets are intended to assure that adequate sites and zoning is made available to address anticipated housing demand during the planning period and that market forces are not inhibited in addressing the housing needs of all economic segments of a community."

In other words it is a planning tool.  Market forces determine what gets built and in what quantity.

Santa Barbara County
Santa Barbara, CA
"The housing targets are intended to assure that adequate sites and zoning exist to address anticipated housing demand during the planning period."  It is not a requirement or goal that the housing gets built, only that there be space allocated and zoned for housing.

San Diego Association of Governments
2601 Juan St., San Diego, CA
"The Draft RHNA Methodology and Allocation distributes housing in accordance with the four RHNA objectives in state law: by reflecting the region’s commitment to planning for housing for all income levels in all jurisdictions, balancing jobs and housing, focusing development in our urban areas, and protecting our rural areas, open space, and habitat lands."

I highlighted the part about protecting open space and habitat lands because some people seem to think RHNA is about housing, more housing, and only housing.  The goals are a balance between housing and many other things which make places nice to live in.

Southern California Association of Governments
Los Angeles
"Because RHNA is a representation of housing need for the eight year planning period, it does not necessarily address existing housing need. ... State housing law requires that jurisdictions plan for all types of housing based on the needs identified through RHNA.  In addition, local jurisdictions are also responsible for ensuring there are no unnecessary barriers to the housing approval process."

Note the phrase "plan for", not build.

You can get the complete HCD list of RHNA spreadsheets or PDFs here:

The RHNA numbers were only intended to indicate what zoning should be incorporated into publicly available city and county general plans.  This enables builders to know what and where they could build.  It was not intended as a "goal" for a city or county.  This is because of the very simple fact that cities and counties don't build housing.

More than anything, this shows Scott Wiener's SB-35 is a grotesque misuse of a general planning tool.

But what about places that refuse to allow certain housing to be built?  Aren't they violating RHNA?  No.  The cases that get the most prominence - like Brisbane - have nothing to do with RHNA.  Those that get the most publicity are just some builder complaining a city won't let them build somewhere.  In the case of Brisbane it is on an unregulated toxic waste dump that really should be an EPA "Superfund" site.  RHNA does not enter into it in the slightest.

Builders build housing and if they decide for perfectly legitimate reasons that there is no market for new housing - if, for example, the area is not growing in population - they will not ask for a permit to build.  Even if they do build, but not to the numbers specified in the RHNA spreadsheet, for the income levels planned and zoned for, that city or county "didn't make their RHNA numbers".

This concludes Part II.
Part I is here:

Ree-diculous Ree-na - Part I


Ridiculous RHNA ("Ree-na")

Part I

Part II is here:

California has something called the "Regional Housing Needs Allocation", usually pronounced "ree-na".  It is intended to indicate how many housing units for different income levels a city or county needs to zone for.  See screen shot below (click to enlarge):

Housing and Community Development (HCD) web page for RHNA
RHNA numbers are being misinterpreted in order to justify California laws overriding local control of development.  Proposed laws like Senator Scott Wiener's SB-827 would allow high rise apartment buildings in the middle of single family neighborhoods.  Under existing law (SB-35) and proposed laws (SB-827, SB-828), builders need not consider residents' legitimate concerns about traffic, schools, parking or much else.  No public hearings, no parking requirements no environmental impact requirements.  All this because of gross misinterpretations of what RHNA numbers really mean.

In this part we look at RHNA numbers as applied to various cities and counties and see that if understood to mean "goals" they make no sense.  In Part II we will look at what RHNA numbers actually mean.

Cities that "don't make their RHNA numbers" are subject to SB-35 which is State Senator Scott Wiener's law to "streamline" approval of building companies' housing construction plans.

Terrible, No Good, Very Bad, Just Awful Places That
"Didn't Make their RHNA numbers!"
What is wrong with these places, don't they know there's a "housing crisis"?
A list of over 500 cities and counties that "didn't make their RHNA numbers" is available from the State Housing and Community Development Agency (HCD).  Cities and counties on the list are subject to Senator Scott Wiener's SB-35.  One of the cities is Biggs, in Butte County.  The list looks like this:

There's Biggs at #31 in Scott Wiener's Hall of Infamy!
The full document is available here:

Biggs, CA - Butte County:  Population 1,707 (2010) down by 90 from 1,797 in year 2000.  Biggs "didn't make their RHNA numbers" so is subject to Scott Wiener's SB-35, over-riding local controls on housing construction.  
Biggs, CA (Butte County) in it's entirety.
Biggs "didn't make their RHNA numbers"
Busy Biggs.  Build, Biggs! Build!
We need a Bigger Biggs!  A Biggsier Biggs!  
High Density Transit-Oriented-Development goes here!
Get with the program Biggs!  Senator Wiener is coming after you!

A three hour commute for Biggsians!
This is why we need High Speed Rail!
Biggs Version of Housing Crisis
3 BR house for $717/month
We'll see in the next census if they have reversed their population decline.  In any event, "housing crisis" hardly describes Biggs' situation.

City of Amador - Amador County: Population 186 (2010) - down from 190 in 2000 - "didn't make their RHNA numbers" and is on Scott Wiener's "list of infamy" that we saw earlier.

The picture below of Amador is from the "Encyclopedia of Forlorn Places" which notes that "The town mostly burned down in 1878 but some buildings survived that fire. It's a great town to walk around in."  No one built housing recently - maybe they are afraid of another fire like the one in 1878.  Sure, it's been 140 years, but you never know, right?  "Once burned, twice shy."
Amador: "Forlorn" but not forgotten!
Here is a screen shot of the Amador County housing spreadsheet with Amador City at the very top  (click to enlarge):
You can find this spreadsheet and many others at:
It shows that the town of Amador's RHNA numbers were for 2 housing units.  One for "Very-Low" and one for "Low".  No one built those two units so Amador "didn't make their RHNA numbers."  And for that, Amador City is on the list of jurisdictions subject to SB-35.  Too bad! 😢

Housing Crisis Extends All the Way to Here!

Alpine County: Population 1,154 (2018) down from 1,175 (2010).  Area = 738 sq. miles.  

Alpine County was formed in 1864 and peaked in population that year at 11,000.  A year later it was down to 1,200, 10% greater than now. Population declined by 21 in the 9 years from January, 2010 to January, 2018.  The US federal government owns 94% of the land in the county.

There are ski resorts and lots of up-scale vacation homes - probably vacant most of the year.

Nonetheless, the California State HCD lists them among jurisdictions (cities and counties) that "..have insufficient progress toward their Above Moderate income RHNA ...subject to SB 35."  Their "RHNA numbers" are seen in the spreadsheet screen shot below (click to enlarge):

Alpine County RHNA Numbers

Because no one built 11 housing units for Above-Moderate income people, Alpine County "didn't make their RHNA numbers".  (Maybe only 10 were built - but that wouldn't be enough.)  Therefore, should a builder want to, they can, to some degree, bypass local planning regulations.  Of course, it is highly unlikely anyone wants to build in a county where the population is declining and only about 6% of the land is not owned by the US Govt.  Doesn't matter.  Alpine County "didn't make their RHNA numbers" so local control in terms of considerations of traffic and schools is over-ridden by Scott Wiener's SB-35.

"Housing Crisis" goes to the Nevada border!

Sierra County: Population 3,207 (2018) down from 3,240 (2010).  Area = 953 sq. miles.
Downieville - Sierra County Seat
Sierra County was formed in 1852 and peaked in population in 1860 at 12,000.  Population declined by 33 in the 9 years from January, 2010 to January, 2018.

Like Alpine County, Sierra County "didn't make their RHNA numbers" and is subject to SB-35 to alleviate the "housing crisis".  Let's all hope someone decides to build 11 units of "Above Moderate income housing" in Sierra County so the "Housing Crisis" will be resolved!
"Help us!  Builder-Wan, you're our only hope!"

Modoc County: Population 9,612 (2018) down 74 from 9,686 (2010).  Area = 4,203 sq. miles.

Original home of the Modoc People.  Most of the land is owned by the US Federal Govt.  The National Park Service and National Forest Service employ a large fraction of the population.  The county is on the border with Oregon and Nevada.  It reached a population of 8,000 in 1930. Population has never reached 10,000 inhabitants.

Like Alpine and Sierra Counties, Modoc County "didn't make their RHNA numbers" and is subject to Scott Wiener's SB-35 to alleviate the "housing crisis".
Modoc County "Housing Crisis"
5 BR, 2,332 Sq. Ft., Mortgage = $712 per month.

Other Counties Here is a list of 15 counties in California most of which lost population over the 9 years from 2010 to 2018.  I added Amador and Kings which gained a whopping 3 and 31 residents, respectively, in those 9 years.  All of them are on Scott Wiener's list of infamy.

Alpine County             (21)  Population declined by 21
Sierra County             (33)    "                  "          by 33 etc.
Calaveras County     (421)
Del Norte County   (1,389)
Kings County         (1,320)
Lassen County      (3,984)
Modoc County            (74)
Mono County            (380)
Mariposa County      (122)
Plumas County         (234)
Siskiyou County       (288)
Trinity County           (151)
Tuolumne County     (625)
Amador County            +3    Population increased by   3
Inyo County                +31    Population increased by 31

Population Decrease of 1,389 in last 8 years.
"Didn't make their RHNA numbers" so...

Scott Wiener's SB-35 applies
In the counties listed above, no one asked for as many permits as the RHNA numbers indicated had been planned for.  Probably because there is no demand for new housing in poor counties that are losing (or not gaining much) population.  As a result, those counties "didn't make their RHNA numbers" and are subject to Scott Wiener's SB-35.

There are nearly five hundred cities which "didn't make their RHNA numbers" but it serves no purpose to enumerate every one of them.  In fact, over 97% of California cities "didn't make their RHNA numbers".

It should be obvious by now that the phrase "didn't make their RHNA numbers" does not mean what you thought.  Don't use it.

Monday, September 3, 2018

Sunnyvale Going to District Elections


Wed., September 5th at 7 PM, there will be a special Sunnyvale City Council Meeting to discuss going to district elections. 
Location:  Council Chambers, City Hall, 456 W. Olive Ave., Sunnyvale, CA 94086

Currently all 7 council members are elected at large.  This makes it difficult for those without much money as they need to get their message to 60,000 voters.  Going to district elections would mean that a candidate will only need to reach about 8,500 voters or about 4,000 homes.  This is much more affordable.

The city basically has no choice in the matter.  If it does not do this it will likely get sued under the California Voting Rights Act as has happened to Santa Clara and several other cities.  The courts have consistently ruled against cities who tried to fight the issue.

The agenda has much more information.  It is available here:
Look for City Council Agenda for 9/05/2018.

I've copied some of the highlights below:

-------------  begin partial copy of Sunnyvale City Council Agenda for 9/5/2018 -------

Direction Regarding Public Outreach and Submitting a Charter Amendment to Voters Regarding Changing At-large with Numbered Seats to District-Based Elections

1. 18-0776 (link to detailed discussion)

Recommendation: Alternative 1: Direct staff to scope a public outreach and education plan for receiving public input on whether the Council should place a measure on the November 2020 ballot for voters to decide whether to amend the City Charter to change from at-large with numbered seats to district-based elections, and return to Council by November 2018 for approval of the outreach plan and resources necessary to implement that plan.


Direction Regarding Public Outreach and Submitting a Charter Amendment to Voters Regarding Changing At-large with Numbered Seats to District-Based Elections


Cities throughout the state have increasingly been facing legal challenges to “at-large” systems of electing city councilmembers. The California Voting Rights Act (“CVRA”) was adopted to address vote dilution caused by at-large election systems in the presence of racially polarized voting. Almost all cities challenged under the CVRA have settled claims out of court by voluntarily shifting to district-based elections. In the Bay Area, cities including, among others, Fremont, Menlo Park, and Morgan Hill, have recently switched to district-based elections.

This issue made headlines in our neighboring city of Santa Clara over the last several months. Santa Clara, which has a charter provision with an at-large, numbered-seat system nearly identical to Sunnyvale’s system, was sued by the South Asian Law Alliance claiming that the system violated the CVRA by diluting the vote of Asian voters. Santa Clara chose to litigate the issue and the case went to trial in April 2018. The Santa Clara County superior court judge agreed with plaintiffs, holding that Santa Clara’s system violated the CVRA. The second phase of the trial to determine remedies was held in July 2018, and the court ordered Santa Clara to shift to district-based elections (six districts and a separately-elected mayor) beginning in November 2018.

Although most cities that have changed their electoral systems have done so under the threat of CVRA litigation, staff is bringing this issue forward and recommending that the Council make the voluntary choice to submit the issue to Sunnyvale voters as a charter amendment to comply with the CVRA’s disfavor of at-large electoral systems and eliminate the City’s exposure to litigation. The fact that Santa Clara’s substantially similar system was recently struck down makes this issue timely for the Council’s consideration. As discussed in detail below, Sunnyvale would likely face a high burden if sued, CVRA litigation is tremendously costly, and the outcome of litigation would be highly uncertain. Voluntarily initiating the process to switch to by-district elections will give the Council greater flexibility to determine the process, and the community greater opportunity for input, than the City would have if CVRA litigation is threatened or commenced.

-------------  end partial copy of Sunnyvale City Council Agenda for 9/5/2018 -------

Monday, July 30, 2018

Commute Times


Commute Times
Countries, Metro Areas, Cities

Let's Commute!!


This is part two of a series on commuting.  This part focuses on commute times.  Part one looks at commute distances.  That is here:

In looking at commute times around the world and particularly in the US, we find that most people try to arrange their lives so that the time to get to work is on average about 30 minutes.

Regardless of the transport mix or availability, the average time for each metro area in the US is very close to 30 minutes.  It is a little longer in other countries because more people there use public transit which is slower than driving.

This even carries across cities so that Chicago, LA, NY City, and San Francisco all have similar average commute times and distributions.  The latter two are compared in the graph below (click on graph to enlarge):

Second-by-Second Timing of Commute Times
NYC vs San Francisco

There is no point in broadly trying to shorten commute times because people everywhere adjust their lives so that commute times stay the same.


Everyone thinks their area has the worst commute times but in fact, most countries have longer commute times than the US as seen below (click image to enlarge):

Chart 1 - Commute Times in OECD Countries
OECD = Rich Countries
From OECD's "How's Life - Examining Well Being" 2011 edition
Red Box around USA and line at 30 minute mark.
Green Box around OECD and line at 38 minute mark.
In the above bar chart the average commute time for all the OECD Countries (developed economies) is just under 38 minutes (green boxed bar in the graph above).  This is probably because most other countries have better public transit than the US.  In dense older cities like Paris and London that developed before autos were common, public transit is the only viable way to get to work for many people.  Public transit is slower than automobile and lengthens the average commute time.
London Bus
We see in the following chart for the EU Capital Cities distinct differences with the US.  Far greater percentages use public transit, bike or walk to get to work. In 15 EU cities 50% or more use public transit to get to work.  Contrast this with the NY Metro Area where around 65% commute by car.  In the US, only 5% use public transit. 

In 17 of the major EU cities below, over 25% walk to work.  In the US as a whole, only about 3% walk to work.  The maximum combined walk/bike percentage in US cities is 17% in Boston and Washington, DC.

(Click on chart to enlarge).

Distribution of Commute Modes
European Capital Cities
Very few developed countries except the US were ever self-sufficient in oil.  Importing oil has a negative impact on their economy.  As a consequence, most European and Asian governments impose very high taxes on oil to lessen imports and consequent loss of foreign exchange.  With gas costing about $6 to $7 per gallon, commuting by car is far more costly in Europe than in the US.  The high gas taxes subsidize public transit.

Commuting in the US

From the US Census' American Community Survey we know that the average commute time in US is 26 minutes.  Here are the 20 longest commutes in US cities.  These average closer to 30 minutes.  California cities are 'boxed'.  (Click on chart to enlarge).

20 Cities With Longest Commute Times
Red boxes = Northern California, Blue boxes = Southern California
Average Commute time in US = 26 minutes
Majority use Transit in NYC, Jersey City, and Washington, DC
A couple of things stand out in the charts above:
  • The commute times are almost all around 30 minutes.  Something called "Marchetti's Constant" says just this - that people will arrange their lives so that on average they will commute around 30 minutes.
  • Note that the top three longest commutes are all in the NY Metro area.  The two longest commutes are the areas where most people use public transit.  The commutes take more time because public transit is slow.  Buses average about 12 mile per hour vs the average metro area commute driving speed of about 36 miles per hour.  The average commutes are a little above the 30 minutes in other cities where most people drive - Marchetti's Constant again.
  • Nine out of the top twenty long commute times are in California -  4 in the San Francisco Bay area, 4 in the LA area, and 1 in the San Diego Area.

US Major Metro Areas

The specific cities in the earlier chart on commute times are a small percentage of their respective metro areas.  Now we look at entire metro areas.  I (mostly) used the same 26 large and fast growing metro areas for commute time that I looked at for part 1 on commute distance (with a very few exceptions).  That represents 44 million workers and 100 million residents.

The commute times for these 26 metro areas is shown below (click graph below to enlarge).

Commute Times in 26 Largest Metro Areas
LA and SF Commute Times are the Same
28 Minutes by Car, 49 Minutes by Transit
Average Commute Times for 26 US Census Metro Areas by 3 Modes of Transit
Data taken from Governing.Com which derived data from the US Census Bureau.
CA cities are boxed and their data points are yellow
Above chart of 26 metro areas shows:
  1. The average commute times are almost all under 30 minutes by car and just under 50 minutes by public transit.  Since most people drive alone, the relatively few that take public transit raise the average to approximately 30 minutes.
  2. The average time by bus is just under 50 minutes.  Subtract 4 minutes to get to the bus stop and another 4 minutes to get from the bus stop to work (access time) gives 42 minutes actually on the bus.  At the 12 miles per hour avg. bus speed (see below) that would be 8 or 9 miles.
  3. Subways and streetcars are about the same as bus overall.  Part of this is because there are fewer rail lines per city than bus lines.  This means they are less accessible than buses so increased access time makes up for the higher speed of rail.
  4. The four biggest metro areas in California -  San Francisco, Los Angeles, San Jose-Sunnyvale, and San Diego - are right in line with other metro areas around the country.
  5. San Francisco and Los Angeles metro areas are identical in commute times at 28 minutes. (!?)
In the SF Bay Area there is a huge variation in commute times by community.  MTC (Metropolitan Transit Commission) has excellent data on each community and the Bay area as a whole as seen in the map below (click images to enlarge):

Drive Alone Commute Times - Residents
Sunnyvale residents average only 22 minutes driving time to work
Longest commutes are 35-40 minutes.  Shortest are 20 to 22 minutes.
High housing cost regions have shorter commutes.
Residents pay more to be close to work

Drive Alone Commute Time - Workers
Workers in Sunnyvale average nearly 40 minutes
Those who work in Sunnyvale and can't afford to live there commute.
Why do workers in Sunnyvale commute a lot more than residents of Sunnyvale?  Those with more money win the bidding process that determines rents and housing costs. Since Sunnyvale is in the center of Silicon Valley where the high paying tech jobs are, tech workers win the bidding wars to be closer to work.  Workers who can't afford to live in Sunnyvale commute further.  Sunnyvale residents may not work in Sunnyvale, but most don't travel far.

There is a certain amount of personal preference involved as well.  Some people just prefer a less dense "feel" for their home town and will commute from more rural areas even if they can afford a place closer to work.

Public Transit Usage

As a long time user of public transit of all sorts I can testify to the advantages.  It is usually less stressful than driving (not always).  With a set of noise-cancelling headphones and a book, the time flies by.  Still it takes more time and there is no getting around that.

Commute by Train and Enjoy the Views!
Commuting by train can be quite nice
(This is not CalTrain)
As seen in the chart below, US median and average bus speeds are below 13 miles per hour and going down.  This does not include access time, waiting for a bus, or any transfers.  In the San Jose, CA and San Francisco-Oakland metropolitan areas, buses are a little slower than average at 11 and 12 miles per hour, respectively: (click chart to enlarge)

Buses - 11 to 12 Miles per Hour
Select your favorite city here:
The much longer times for public transit (mostly buses) are due to a combination of time stopping to pick up and drop off riders, time getting to and from the nearest boarding point (access time), and the difficulty in maneuvering large buses on city streets.

That may be one of the reasons why public transit use is declining around the US, including NYC, as seen in the graph below:
Speed of transit is important in job access.  We found in the study of commute distances in the US that solo driving commutes average about 18 miles.  See chart below:
From "Commute Distances" at
We have seen here that commute times average around 30 minutes.  Going 18 miles (on average) in 30 minutes (on average) means that the average commute speed by car is about 36 mph.  Contrast this with the average bus speed in San Jose of about 12 mph.  If you can go 36 mph in a car but only 12 mph in a bus, then simple geometry says a car enables you access to 9 times the area for a job search.  [ Because of geometry:  Area of commute = (Pi) R(squared) ]

In fact, however, people only drive an average 18 miles to work and take the bus about 9 miles.  This still means a huge difference in the job possibilities. Consider Los Angeles as seen below with the range for an 18 mile, 30-minute commute by car vs. a 9 mile, 50-minute commute by bus. (click graph to enlarge).

Los Angeles Commute Ranges
9-mile 50-minute bus ride vs. 18-mile 30-minute drive
The 9-mile commute circle contains 254 sq. miles
The 18-mile commute circle contains 1,018 sq. miles
Someone driving a car (big circle) has access to 4 times more area and presumably 4 times more jobs than if they take transit.  This is a second reason (the first being time) that cars are preferred by most people for commuting and bus ridership is declining.

Santa Clara County VTA Bus Ridership vs. Employment
Usually ridership goes up with employment - not any more
From VTA 2018-2019 Budget at:
I cover the decline of transit usage in Santa Clara County and other areas here:

If commute times of around 30 minutes for most big US metro areas don't sound right to you, you are not thinking of averages, but of the two hours that comprise morning commute when almost half of all commuting to work is done.  This leads us on to the big issue of...

Commute Time Distributions

While most seem to think commuting in their particular area is much worse than other areas, it turns out to be pretty much the same for every city.

Pleasanton, CA
Pop. 70,000
Commute times similar to San Francisco

Travel Time Distributions Around SF Bay Area
Pleasanton = 32.7 Minutes, Central SF = 32.3 Minutes
Outer San Francisco = 35.5 Minutes, Sunnyvale = 24.2 Minutes
About the above charts.  The first chart on the upper left is for Pleasanton, CA - a suburban town of 70,000 people about an hour's drive from San Francisco.  It is worth noting that the majority of people there spend less than 30 minutes getting to work, although a significant number spend between 45 and 60 minutes and some spend over an hour.

The surprise from the above charts is that the commute time for Pleasanton residents isn't a whole lot different overall than "Outer San Francisco" (the chart below Pleasanton's) at 32.7 minutes for Pleasanton, vs 32.3 minutes for Central San Francisco - a 24-second difference.  And the distribution of times is very similar as well.

However, the next chart shows that transport preferences by residents of Pleasanton are very different from those of residents in San Francisco.  Data is from the city-data webs sites listed above.  Click chart below to enlarge.

How the Residents Get to Work
Most common is car - even in San Francisco

The population density in Pleasanton is too low to support ass transit so people there mostly drive.  No surprise that San Francisco commuters use public transit and bike/walk to work much, much more than those in either Sunnyvale or Pleasanton. Even so, commute times are very similar both on average (around 30 minutes) and in their distribution.

Outer San Francisco
Ocean Breezes - some short and some long commutes

And even "Central San Francisco" - where many more walk or bike to work - is very similar to both "Outer San Francisco" and Pleasanton in time distribution.

Central San Francisco
High Rise Buildings but similar commute time distribution

Of the four areas plotted above, the one area that has the smallest percentage of commuters commuting over 45 minutes is Sunnyvale, CA.  Sunnyvale is a suburb in the "Heart of Silicon Valley" (tm).  It combines a pleasant place to live with a mix of single family housing and two to four story apartments with close proximity to many technology workplaces so there is less reason to drive long distances.

Notable is that in all four areas, driving solo is still the most popular mode of commuting.

Other cities...
    ...are no different in their commute patterns.

Here is the time distribution for Chicago which is very typical of big metro areas including the SF Bay Area. The green bars show the generally acceptable 34 minutes or less times, the yellow bar shows the longer 35 to 44 minute commute, and the pink and red bars show the more challenging longer commutes - with a dark red arrow indicating the average in the middle.  We see that most (57% of commuters) have a commute time of 10 to 35 minutes, and this doesn't even include those with less than 10 minute commutes: (click on image to enlarge)
Chicago Commute times

New York City vs. San Francisco  

The previous time distributions are from US Census data which comes from simply asking people how long they commute.  People tend to choose numbers ending in 0 and 5, perhaps distorted by a recent particularly bad commute.  The chart below from Chronos is derived from a second-by-second reading of cell phone GPS data so is about as accurate as you could wish for.  Discussion below (click chart to enlarge):

Commute Times SF vs. NYC
Using Cell Phone Info
Remarkably similar commute time distributions

As seen in the above chart, in both SF and NYC, the single most common commute time is about 25 minutes.  There are fewer in that range in NYC than in SF.  This is likely because NY City has nearly nine times the population, and over six times the land area as San Francisco so more people are traveling further.

The average for both cities is higher than 30 minutes because of the "fat tail" on the right of the chart stretching from 40 minutes to 1 hr 20 min.  Despite the vast differences in geography and population, the distributions are incredibly similar.  In both cases it appears the vast majority are keeping their commute time under 45 minutes, averaging about 30 minutes.

This suggests that people everywhere are making choices about where they live and work with regards to commute time with very similar outcomes.  This is important.  People make choices.  There are trade-offs with every choice.  For the same money you can...

1.  ... live in a big city in a small, expensive apartment and take transit to work ...

2.  ...or live in a large house in the suburbs and drive to work.  Same cost - personal preference as to which is better.

It is impossible to avoid that trade-off.  I discussed this in detail here:

and here:

To drive home the commonality of human behavior, we have the following very, very similar chart from cell phone data of commute distances (roughly equivalent to time) around the world - Ivory Coast, Portugal, Boston, Milan, and Saudi Arabia.  People are the same everywhere and make the same choices when faced with decisions about how far or long they are willing to commute.

Commute Distance Distributions
Very Similar Across Countries
Similar to US time distributions in NYC, SF, and Chicago
In probability and statistics this is known as a "Gamma Distribution".  It occurs naturally in modeling waiting times and the flow of items through a process - such as cars through a production distribution process or people shopping at a market.   See graph below.
Gamma Distribution with different parameters
Here is Tokyo commute distribution:

Again, a very similar distribution to Chicago's although the average is much higher at 62 minutes for a commute.  This is because, at 38 million people, Tokyo is the largest metropolitan area in the world, and is very spread out.  Japanese want a single family detached house as much as Americans and need to commute further to be able to afford it.

Here is the cumulative distribution chart of the above data showing the average time, and 2/3 cut-off

SF Bay Commute Times by County

The San Francisco Bay Area "Metropolitan Transit Commission" has some great data on their web site showing commute profiles.  The following profile by county of commute times (for driving alone) gives a variety of insights discussed below.
Data from MTC website:
The top bars in the bar chart above in the "<15 Min" category are clearly indicate that the less dense the county, the less the commute time.  People in rural areas don't have much public transit so they need to drive which is everywhere faster than public transit.

It also relates to the fact that density of people correlates with density of jobs.  The more jobs you have scattered around the more people commute from one place to another, so the greater the congestion for cars and the more people use (slooow) public transit.

On another graph from MTC's web site, I've highlighted some noteworthy points with boxes and arrows: (click image to enlarge):
In the above chart, we see...

...Under 30-Minute Commutes

If we look at the maroon line in the middle at the 50% mark, we see that for 8 of the 9 SF Bay counties, at least 50% of the population commutes less than 30 minutes.  The one exception is San Francisco with about 45% of commutes under 30 minutes (marked by green box with green-yellow arrow).  No surprise - more people in SF use public transit which is slow.

Looking at the two orange boxes (near orange-blue arrows), we find in the two rural counties of Napa and Sonoma about 70% of commuters commute less than 30-minutes.  There aren't many jobs there, nor many people, nor many towns.  If you live in one of the few towns there, you probably work there or very close, and have an easy commute along local streets and/or unimpeded highways.

...Over 60-Minute Commutes

The highest percentage of long commutes (over 60 minutes) in the SF Bay Area occurs in Contra Costa County as indicated in the above chart by the Blue Box (with blue and pink arrow) at around the 82% mark.  This is because Contra Costa County is a bedroom community remote from the job centers of Silicon Valley and San Francisco.  The 18% of commutes that are over 60 minutes long is in stark contrast to the 6% or 7% of commutes in Napa and San Mateo (Purple Boxes) that are over an hour.

We can see the disparity between the longest and shortest commutes and the difference between solo driving and public transit by state in this visualization (click to enlarge to readable size):

Average Commute to Work by State and City
Yellow bars are drive solo, brown bars are public transit, red lines are overall averages
Longest Avg. Commute = NY State = 33.4 Minutes
Shortest Avg. Commute = South Dakota = 16 Minutes

Red line "total averages" are almost all close to "solo driving" averages because (except for NY) only about 5% of commuting is done by public transit.
So what about all those horrible long distance commutes you always read about in the news media and on chat boards?  Some observations:
  1. Long commutes are the "man bites dog" aspect of news reporting.  That about half of all commutes are unexceptional and under 30 minutes is not newsworthy or likely to generate any conversation on social media.
  2. Most people don't "get" what an average is.  If the average commute time for two people is 30 minutes that could be that each has a 30 minute commute or it could be that 1 person has a 10-minute commute and the other has a 50-minute commute.  Same average but vastly different experiences.  One commute is a news item and one isn't.
  3. That it is "only" about 30 minutes by car in LA and San Francisco does not express the tension and frustration of stop-and-go freeway traffic on people.  Auto accidents are expensive and often deadly - this increases the tension as people try to drive.  Avoiding an accident on curving freeways with people changing lanes to save a minute or two is nerve-wracking and makes 28 minutes "feel like" 2 hours.
Driving in LA can be Dangerous

Driver assist technologies such as automated lane-keeping, "dynamic cruise control", and collision avoidance can alleviate a lot of the tension from commuting.  With less stress, there should be fewer people engaging in dangerous (and rude!) behavior which would make commuting nicer and far more tolerable.

But if half of all commutes are under 30 minutes, then what is all the concern about traffic?  What most people are thinking of when they talk about bad traffic is...

Rush Hour Congestion.  
The Nerve-Wracking Aspect of Commuting

I was going to cover congestion and long distance commuting (mega-commutes, stretch-commutes, etc.) but those topics turn out to be really complicated with a lot of variations.  They would easily double the length of this post so I will cover them in another post.

As a small preview of that post - one item is that around 40% or more people work what are called "anti-social hours".  That doesn't mean they don't like people but that they are working at times most people are sleeping.  Examples would be the midnight to 8 AM shift in a hospital, or coming at 5 AM to open a coffee shop at 6 AM then going home at 2 PM.  This means half the population never faces congestion.  They cruise to work on deserted streets and highways at (or above) the speed limit.

LA Freeway - Without Traffic


The cost of commuting in terms of stress and money means some are willing to pay more to live closer to work.  Others, by preference and/or trade-offs between housing costs vs. commute time choose longer commutes.  You can't change people's preferences - everyone is different.

"You say To-MAY-to, I say To-MAH-to.."

We have looked at every state in the US, every country in the OECD (developed countries), some developing countries and various parts of each metro area.  Looking at the averages and distributions of times we found over and over and Over that most people arrange their lives so that commute times are somewhere under 30 to 40 minutes.