Printerr-friendly PDF file

Why SFMTA released its Evaluation Report nearly four months prematurely analyzing its Commuter Shuttle Pilot Program on October 5, 2015 — when the trial period is set to run through the end of January 2016 — isn’t known, but it’s thought the premature analysis was rushed into print so SFMTA’s Board could vote on November 17 to make the program permanent, even before the end of the 18-month pilot trial period.

Sadly, SFMTA’s October 5 report is riddled with flaws.

Flawed Data Collection Methodology

On page 5 of the report, Muni (“Muni” is used interchangeably in this secondary analysis to refer to the SFMTA) claims that the Pilot Program allowed it to collect data regarding the movement and usage of commuter shuttle buses; further claimed that based on the data collected, the October 5 report evaluated how the Pilot Program had performed on its objectives; and also claimed the Evaluation Report would be used as a basis to make recommendations as to whether the program should be continued, making it permanent.

One of the key objectives of the Pilot Program was to gather data regarding shuttle activity in the City. Given the paucity of data collected from shuttle bus operators, the objective to adequately collect reliable data appears to have failed, miserably.

And given that the data that was collected was wholly incomplete, the premature Evaluation Report should not be used as the basis for making the shuttle bus program permanent.

It is not until page 34 of the Evaluation Report that Muni admits the severity of gaps in the data collected from shuttle operators.

The report notes on page 34 that it had seemed to be a straightforward requirement at the outset of the program to require shuttle operators to provide real-time data on shuttle stop-events and vehicle movements, but it proved to be “more complicated that originally contemplated.” Jeesh! Collecting GPS data is now “complicated”? In 2015? Really?

Maybe GPS tech whizzes at Google could volunteer their skills and jump in to help the Google-bus shuttle operators uncomplicate the collection and submission of GPS data.

For that matter, the City has seventy (70) 1042 “IS Engineer-Journey” employees on the City’s payroll, with combined total pay of $7.8 million, many of whom are GPS experts. Why can’t they be pressed into service to get the commuter shuttle bus GPS flowing accurately to SFMTA?

Shuttle Operators Fail to Provide Data

Despite a requirement to provide GPS data regarding real-time location and stop-events of each shuttle bus in order to receive a permit, damning admissions in the report noted on page 34 indicate the failure of the shuttle operators to provide the required GPS data, even though shuttle providers were required to reapply for all permits by February 2015 six months into the project, but may have received renewed permits, despite being GPS non-compliant:

• In the section titled “Compliance with permit terms,” Muni acknowledges it had relied on the Pilot Program shuttle operator “partners” to abide by the permitting rules of the program. Then Muni acknowledges that “due to limited enforcement resources described [earlier in the report], relying solely on issuance of citations to keep shuttles out of Muni and other no-stopping zones appears to have limited effectiveness.”
  
  What this appears to say is that shuttle operators used Muni stops inappropriately and even issuing them citations appears not to have changed operator’s behavior. If they won’t comply with permit terms during a trial period, why should the program be made permanent and who believes they’ll comply going forward?
  
  
• “While all shuttle operators have made at least some effort to provide this data, some have provided the data without interruption or issue, while others have failed to provide data regularly and accurately.” Muni didn’t quantify how many of the 16 shuttle operators failed to provide data regularly, so readers of the Evaluation Report have no way of evaluating the scope of shuttle operator’s failure to provide data.
  
  
• “Some operators who have failed to send data have worked closely with SFMTA staff to resolve data delivery issues, while others have been slow to respond in inquiries from staff and do not appear concerned about ensuring the proper delivery of data.” Again, Muni failed to quantify a precise number of shuttle operators who were slow to respond, whether all of the shuttle operators eventually responded, or how many of them aren’t concerned about providing the mandated data.
  
  
• “Issues with SFMTA’s data vendor have complicated the process even further, such that, more than a year into the Pilot Program, the real-time vehicle data is still not flowing completely or accurately from all operators.” Another damning failure of Muni to quantify the number of shuttle operators who are not providing real-time data, as required to participate in the program. Muni’s failure to name the operators by name doesn’t help to weed out the bad actors who shouldn’t be allowed to participate in the program if it is made permanent.
  

Muni appears to have allowed shuttle operators to continue in the Pilot Program even though they weren’t complying with the GPS requirement to obtain a permit, and when the operators had to reapply six months into the program, Muni appears to have allowed operators to obtain renewed permits, despite not complying with the GPS requirement.

Muni claims on page 34 that it is now undertaking processes to bring the data collection and reporting in-house to eliminate vendor issues, but didn’t indicate how long this will take, or when ti might be fully implemented.

But Muni offered no information on why Muni buses — the majority of which are thought to already have GPS capabilities — hadn’t been submitting reports of conflicts in Muni bus zones themselves through Muni’s on-board camera systems that are capable of recording video of traffic in front of them violating double-parking regulations.

Brazenly, public records show that on August 8, 2014, LinkedIn Corporation’s Global Transportation Program Manager, Michael Alba, e-mailed SFMTA Shuttle Program manager Carli Paine indicating LinkedIn wanted to implement a new Caltrain shuttle between the Financial District and a new site in SoMa within a week, but LinkedIn’s GPS provider had a two-month backlog.

When Alba asked Paine if he could submit “a formal request for leniency on the GPS requirement” to obtain a permit, Paine clearly and thoughtfully replied saying the GPS requirement “is not a negotiable element of the program,” and suggested LinkedIn “look into other GPS providers so that you can meet the permit terms.” But it’s not known whether LinkedIn quickly found a new GPS provider, or whether LinkedIn simply began using the shuttle without GPS capability, or without a permit.

Flawed Data Collection Methodology:  Multiple Sampling Biases

The Evaluation Report appears to have several problems with “sampling bias.” First, on page 18 of the report, Muni admits it conducted field data collection only in June 2014 (before the pilot Program began in August 2014), and June 2015 to examine the impact of the program on traffic conflicts and potential safety issues.

Nowhere in the report does it indicate on how many days in June 2014 and June 2015 field data collected actually occurred on a daily basis. And the report fails to mention why field data collection occurred in just one (June 2015) of the first 11 months (between August 2014 and June 2015) of the 18-month study. Why wasn’t data collected between July and October 2015, or was it collected and not reported?

First, of the 124 bus zones in the Pilot Program, Muni surveyed just 20 locations (16.1%), 10 each during the morning and evening commute periods, which drops to just 8.1% of locations during each commute period. This may not have been a very representative sampling of the 124 bus zones.

Second, of the 20 stops used for data collection, 7 of them — 35% — were in “shuttle-only” zones that Muni buses weren’t allowed to use, skewing how many blocked Muni buses there were per hour. Theoretically, you shouldn’t expect any blocked Muni buses in shuttle-only zones (although that happened at one of the 7 shuttle-only zones), but instead of sampling 20 Muni zones, SFMTA allowed the co-mingling of data from 35% (7) of the 20 zones reserved for shuttles only.

Third, Muni noted on page 28 that the Pilot Program included funding for a l0-person morning and evening enforcement team known as the “shuttle detail,” and then acknowledged on the next page that there was limited staffing for the shuttle detail.

This is ludicrous: Extrapolating from the 2,978 daily stop-events made by the shuttle buses times $3.67 per bus per stop-event, Muni received $2.94 million on an annualized basis from fees for stop-events alone (based on about 269 week days annually). And it could only fund a 10-person enforcement team from the stop-event fees?

And what about the 1,200 citations issued to shuttle buses by all of Muni’s enforcement officers (not just by the 10-person shuttle detail assigned to enforcement) between August 2014 and the end of May 2015 — ostensibly at only the 20 zones monitored during the pilot period for the 124 zones? Didn’t those citations generate any revenue in fines from the citations to hire additional shuttle detail enforcement staff? For that matter, how many additional citations — and resulting increase in revenue from fines — would a much larger shuttle-detail enforcement team have issued had the other 104 zones not even surveyed been routinely monitored for enforcement?

Footnote 13 on page 30 of the Evaluation Report noted that SFMTA can shift enforcement staffing “by using shuttle bus GPS data to determine where enforcement is needed most.” An obvious, but unanswered, question is why Muni can’t issue more citations just from the GPS data it is collecting. Since FasTrak mails citations to drivers crossing Bay Area Bridges when transponders fail to collect the bridge tolls, why can’t MTA issue citations based on GPS data?

For that matter, Muni mentioned not one word in its Evaluation Report about why Muni Transit Operators (bus drivers) weren’t enlisted to collect and report bus-zone conflicts using their on-board cameras, or merely using a clipboard to record the number of times drivers had to lower wheelchair mechanical lifts directly into the roadway for disabled passengers when shuttle buses blocked Muni bus access to red zones daily between August 2014 and May 2015.

Almost comically, Muni reported on page 29 that the primary goal of the shuttle detail was not to issue citations, but to keep zones safe and traffic flowing smoothly. Then the report says the number of citations issued “is not necessarily instructive of whether the Pilot Program’s goals were met through enforcement actions.”

The goals of the pilot program included reducing conflicts with Muni and other vehicles, and improving safety interactions with other users. It sounds to this observer that an average of 103 citations monthly points to a glaring problem with both safety issues, Muni bus conflicts, and other traffic issues.

Did Muni think readers wouldn’t spot skewing of data from its data collection sampling biases?

Garbage In, Garbage Out

Given the biases Muni introduced gathering data to assess the effectiveness of the Commuter Shuttle Pilot Project, is it any wonder that much of the data presented in the report is suspect, when not apparent lies?

For instance:

• On page 6 of the Evaluation Report, Muni indicates that there was a 29% increase is shuttle daily stop-events, but instances of shuttles blocking Muni buses decreased by 35% from the June 2014 pre-pilot period and the August 2014 to May 2015 during-pilot data collection periods.
  
  On the next page the report notes that an average of 2.7% of shuttle stop-events resulted in blocking Muni bus access to red zones, apparently each day. So of 2,978 daily stop-events, 2.7% represents at least 80 blocked Muni buses daily, and may annualize to about 21,629 occurrences per year. If so, this is completely unacceptable.
  
  But it’s unclear whether those 80 blocked buses occurred only at the 10 shuttle zones surveyed in each of the morning and evening commutes, and it’s also unclear whether the 80 blocked Muni buses was an “educated guess” based on extrapolating to all 2,978 daily stop-events from data actually observed at the 10 locations.
  
  
• Page 7 of the Evaluation Report notes that shuttle buses “block other drivers’ views of pedestrians, or block crosswalks, less than 2% of the time they stop.” Of the 2,978 daily stop-events, 2% translates to approximately 60 blocked crosswalks and blocked views of pedestrians daily, which annualizes to about 16,022 occurrences per year. This also is completely unacceptable.
  
  
• Page 7 of the Evaluation Report also notes that “shuttles block travel and bike lanes about 35% of the time that they stop.” Out of 2,978 daily stop-events, 35% translates to 1,042 blocked travel lanes and bike lanes — apparently daily — a very significant number that may well impact various flows of traffic in San Francisco, because if this volume is occurring daily, it translates to a staggering 280,379 occurrences annually!
  
  
• On page 11 of the Evaluation Report, Muni noted that the pilot network of zones grew from 101 zones in June 2014 to 124 zones in July 2015, for a 23% net increase. The report doesn’t indicate how many zones may have been added between July 2015 and when the Report was issued on October 3.
  
  Notably, the agenda for the SFMTA’s Board of Directors meeting on November 17, 2015 indicates that the Board will consider making the program permanent at 129 zones. And a Planning Department Certificate of Determination: Exemption from Environmental Review for converting the Shuttle Pilot Program into a permanent program dated October 22, 2015 without a full Environmental Impact Review analysis indicates the Planning Department reports it believes the permanent program may increase by 41%.†
  
  A 41% increase in the number of zones portends that the zone network will increase by 51 zones to a total of 175 zones, if the program is made permanent. And all of the data — from total number of shuttle bus passengers, total shuttle buses, total daily stop-events, total conflicts with Muni bus access to red zones, total number of blocked travel and bicycle lanes, and total blocked views of pedestrians and blocked crosswalks — may also increase by 41%, which if you’re a pedestrian or a disabled rider relying on Muni portends a totally scary picture of what’s to come in terms of passenger and pedestrian safety issues if the program is made permanent.
  
  
• Page 13 of the Evaluation Report notes that as of March 2015 shuttle operators had 479 registered vehicles in the pilot program 399 — 80% — of which are intercity motor coaches that are typically 40-feet to 45-feet long, including double-decker buses. Assuming a 41% increase in the number of registered vehicles if the program is made permanent, that could mean the addition of another 164 giant-size intercity motor coaches on the streets of San Francisco that may increase the likelihood of serious or fatal collisions, and increase greenhouse gas emissions even further — without a full EIR of the planned increase in shuttle bus activity!
  
  
• On page 16 of its Evaluation Report, Muni reports that 546 of the 8,500 estimated daily riders — just 6.4% — had completed a survey distributed via shuttle operators and employer sponsors to assess transportation choices.
  
  The report noted that the vast majority of survey respondents work in the Peninsula/South Bay and half of the respondents work in Menlo Park, suggesting that many Facebook employees had completed the survey instrument. Why aren’t billionaire Mark Zuckerberg and Facebook building workforce housing in Menlo Park for Facebook employees so they don’t need to commute?
  
  
• On page 18 of the report, Muni noted that just 372 of 2,302 daily stop-events — 16% — had been observed during the June 2014 pre-pilot data collection period, and just 706 of 2,978 of daily stop events — 24% — had been observed in July 2015 during-pilot data collection period. That means that just 1,078 of 5,280 stop-events (just 20%) were observed.
  
  
• Page 20 of the report notes that data collectors were so buy recording data — including recording blocked travel lanes, blocked bicycle lanes, blocked right-turning cares from seeing pedestrians crossing, double-parking, whether shuttles couldn’t access a stop due to another vehicle blocking access, Muni buses being prevented from accessing a Muni stop, loading and unloading in the street, among other conflicts — that the data collectors were left with limited (or no) time to record instances of people with disabilities being denied access to a Muni bus.
  
  
• The report notes on page 21 that shuttle bus frequency (as measured by stop-events) at observed zones increased by nearly 80% from June 2014 to June 2015. Any time a phenomena increases by 80%, you can expect that there would be negative impacts that may well be going undocumented!
  
  
• Brazenly, on page 23 of the Evaluation Report, Muni claims twelve of the during-pilot zones saw no Muni buses blocked at all. The report failed to note that 7 of the 12 locations were shuttle-only zones, suggesting Muni buses weren’t allowed to stop there, anyway, and likely wouldn’t have experienced Muni buses being blocked.
  
  Shockingly, the report claims that compared to the pre-pilot period, there were zero blocked Muni vehicles per hour at Van Ness Avenue and California Street, which is complete nonsense, in part give that Van Ness Avenue and California Street Muni bus stop nearly tripled from 10 shuttle buses per hour during the pre-pilot period, to 28 per hour in the during-pilot period.
  
  As the dozen photographs this author has taken since October 26 at the Van Ness Avenue and California Street Muni bus stop shown in Appendix A on pages 9and 10 of this secondary analysis illustrate, there’s more-than-ample evidence that Muni buses are routinely blocked by commuter shuttle buses at this stop! You just can’t make this stuff up, given photographic evidence like this.
  
  
• Page 23 of the report also acknowledges that two locations — Valencia at 24th Street, and Van Ness Avenue at Union Street — saw the percentage of Shuttle bus stop-events blocking Muni buses increase 10% and 18% of the time between the pre-pilot and during-pilot phases, respectively, with Muni blocked buses at Van Ness and Union occurring nearly one-fifth of the time, which is, obviously, totally unacceptable! That may be because shuttle stop-events increased by 56% at Valencia and 24th Street, and by a whopping 238% at Van Ness and Union during the evening rush hour.
  
  
• On page 25, Muni again claimed that although the Van Ness Avenue and California Street Muni stop sees 13.5 Muni busses per hour, there we no — zero — blocked Muni bus conflicts. Take your pick: This is either sheer nonsense, or an outright lie!
  
  Muni contends on page 25 that “when shuttle [buses] are provided exclusive zones for loading and unloading, conflicts with Muni [buses] are erased almost completely”! Who’s writing this nonsense for Muni? Ron Conway and his billionaire pals? Mark Zuckerberg? Our erstwhile Mayor? Or someone at Muni looking for a promotion by making up totally ridiculous claims?
  
  
• Although the report noted on page 26 that shuttle buses that fail to pull all the way to the curb cause traffic conflicts by blocking other vehicle traffic lanes and bike lanes, the Pilot Program attempted to mitigate this problem by confining shuttle buses as much as possible to low-frequency Muni zones where they [would be] less likely to encounter a Muni bus. Obviously, the Van Ness Avenue and California Street Muni bus stop is not a low-frequency Muni stop, particularly since it’s situated on an arterial street!
  
  
• Page 26 of the Evaluation Report presents a table purporting to assess the percentage of blocked travel and bike lanes hourly as a percentage of hourly stop-events, but 8 of the 20 zones reported an increase, often substantial in terms of percentages of increased blocked travel and bike lanes. The percentage of blocked travel and bike lanes at the Muni stop at Valencia and 24th jumped from 29% during the pre-pilot to 105% in the during-pilot phase, because two shuttle buses managed to block both the bike lane and travel lane at the same time. This author has repeatedly witnessed two shuttle buses simultaneously blocking the traffic lanes at Van Ness at California Street, as well. There are reports from all over the City that multiple shuttles often access the same stop concurrently.
  
  
• As noted above, page 27 of the Evaluation Report reports that Muni’s own data suggests shuttle buses block travel and bike lanes 35% of the time. That’s way too often for a City attempting to get to Vision Zero!
  

Apparently Muni may never have heard the phrase “Garbage In, Garbage Out,” which seems to be the most glaring problem with its premature Evaluation Report.

There’s plenty of flaws-in-the-glass snake oil throughout this Evaluation Report, but no explanation or presentation of any data that may have been collected between June 2015 and November 2015. Are we to believe no data was collected during the intervening four months since June 2015 because Muni has been on hiatus?

Lawsuit Reached Trial in Superior Court

The lawsuit filed by the Coalition for Fair, Legal, and Environmental Transit on May 1, 2014 was finally heard by Superior Court Judge Garrett Wong on Friday, November 13, 2015.

Courtroom observers, including this author, were rather startled when Judge Wong asked at the outset whether if he ruled on the so-called “pre-emption” argument in the case, whether he then needed to rule on the CEQA environmental review aspect of the lawsuit. His question was seen by some observers to be a good sign that he may have concerns about the “pre-emption” issue, which involves whether local ordinances can “trump” State law, which of course,
they can’t.

As plaintiff’s in the lawsuit noted, California Vehicle Code §22500(i) stipulates:

“No person shall stop, park, or leave standing any vehicle in” … or “alongside curb space authorized for the loading and unloading of passengers of a bus engaged as a common carrier in local transportation when indicated by a sign or red paint on the curb erected or painted by local authorities pursuant to an ordinance.”

Common carriers are individuals or businesses that advertise to the public that they are available for hire to transport people in exchange for a fee. Obviously, the tech commuter shuttle buses are not common carriers, and have no business stopping in bus zones deemed by local ordinance.

So it’s quite clear that California’s vehicle code pre-empts the City ordinance that authorized the shuttle pilot program to have begun in August 2014, and it appears Judge Wong understands this.

For that matter, public records show that as far back as October 4, 2011, SFMTA program manager Carli Paine had e-mailed SFPD Captain Al Casciato and two other SFPD staff providing the citation and full text of Vehicle Code §22500 to Casciato, and using yellow highlighting on §22500(i).

In other words, SFMTA knew three years before rolling out the shuttle project in August 2014 that not only were the commuter shuttles not common carriers, SFMTA knew State law preempts Muni from allowing the commuter shuttle buses to use Muni red zones. More brazenly, when San Francisco’s Board of Supervisors denied a CEQA review of the Shuttle Pilot Program, they also had to have been aware a local ordinance cannot preempt or “trump” a State law.

Shockingly, at the conclusion of the Superior Court trial on November 13, a Mr. Carr, a lawyer for Genentech††, stupidly asserted after the period for taking all evidence in the case had officially ended, that the plaintiffs have “no beneficial interest” and “no public interest exception” — no standing — for bringing their lawsuit to Superior Court, in another brazen attempt to seek summary judgment, asking Judge Wong dismiss the lawsuit altogether.

Plaintiff’s attorney Richard Drury rebutted Carr’s nonsense to Judge Wong and did so very competently, asserting that the plaintiff’s First Amended Verified petition had provided more than sufficient justification of “standing” to have filed the case, so hopefully Judge Wong will accept Drury’s line of rebuttal.

Carr’s tactic of waiting to announce the “standing” issue until after the period for submitting evidence in the trial had ended amounted to a legal trick called “sandbagging,” which involves trial lawyers noticing, but not mentioning during a trial, potential errors in hopes of using it during a subsequent appeal. “Sandbagging” usually doesn’t keep an issue alive, and Trial Judges typically don’t like it when counsel on either side attempts to sandbag the process by not pointing out potential errors in a timely manner. Carr should have brought this issue up before the period for submitting evidence in the trial had ended.

Leave it to Genentech and the City of San Francisco to pull such a brazen stunt! Hopefully, Judge Wong will not only uphold the plaintiff’s standing, he’ll rule that the pre-emption provision has merit separately from the CEQA aspect of the lawsuit.

Fast-Tracking Permanent Shuttle Program Using Flawed Data

As I noted in my extended article, “Tech Buses” Drive Displacement located at www.stopLHHdownsize.com, in response to public records requests Sue Vaughan had placed, e-mail records received reveal City planners repeatedly referred to making the program permanent on a “tight schedule,” and that in an effort of adhere to the tight schedule, some data of the program were made on an “educated guess” basis by planners, rather than relying on real data. SFMTA claimed it disagreed that the program was being “fast tracked.”

Indeed, during SFMTA’s CAC (Citizens’ Advisory Council) meeting on November 5, Ms. Vaughan introduced a motion bravely recommending that the commuter shuttle buses be removed from Muni stops. Unfortunately, her motion failed to pass — with just one vote shy of passage. One vote!

Sounds to me like there’s a Citizen on this Advisory Council who may not have my — and other Muni passenger’s — best interests at heart.

Neither SFMTA’s Evaluation Report nor the Planning Department’s October 22 new EIR exemption letter for making the shuttle program permanent mention there is a distinct probability the shuttle program will face no upward limits, and there may be unlimited expansion of the shuttle program in the future.

Given the many flaws and gaps in reliable data in the premature Evaluation Report, SFMTA’s Board of Directors should reject making the shuttle program permanent.

Monette-Shaw is an open-government accountability advocate, a patient advocate, and a member of California’s First Amendment Coalition. He received the Society of Professional Journalists-Northern California Chapter’s James Madison Freedom of Information Award in the Advocacy category in March 2012.   Feedback: monette-shaw@westsideobserver.

  † Memo - Potential Increase in Commuter Shuttle Activity, from Hank Willson (SFMTA) to Melinda Hue (SF Planning Department), dated October 8, 2015.
†† Genentech and several other tech companies, including Google and Apple, had initially been named as defendants in the lawsuit, but fought bitterly to be removed, and were. Later, the tech companies reapplied to become "intervenors" in the lawsuit; intervenors are third parties who were not originally a party to a lawsuit who become a party if approved by the Court. Why Genentech first fought to get out, and then later rejoined, isn't known. Perhaps Genentech concluded San Francisco Deputy City Attorney Audrey Pearson was handling the defense badly.

Top