7-22-17

Venue:  Red light camera contract, July 25, Garden Grove City Council, item 6a

Honorable Mayor and Councilmembers:

I.   I am concerned that the Garden Grove red light camera program has a ticket quota.  Per the big month-by-month ticketing table on the Garden Grove Docs page at highwayrobbery [dot] net,  ticketing in Garden Grove has been extraordinarily level for the last eleven years, except for 2015 when ticketing was down by half; ticketing dropped suddenly in January 2015 - probably because the yellows were lengthened at that time - and then in February 2016 ticketing came back up to within a few percent of previous levels.

II.  Safety should be # 1 - it is the only honest reason to operate red light cameras - and an objective measure of safety is a tabulation of the number of injuries over the years - including adjustment for the other factors that were active during the time period and a comparison to "control" intersections - yet the closest the current staff report comes to any numbers is the statement, "...shown to reduce severe collisions and save lives."  No numbers at all!  So, we must look at the record.  The last discussion of the safety effect of the City's cameras was in 2015.   The matter was on the agenda at two council meetings, and there was a number of studies and emails, as follows.

    A. The staff report for the Aug. 25 meeting became available to the public on Aug. 20.  (I will not be attaching copies of these six reports and emails today as the total file size would be excessive.  They are available at links on the Garden Grove Docs page at highwayrobbery [dot] net.)   

    B.  On Aug. 23 I submitted an email in which I questioned the amount of the rent and also staff's inclusion of accidents coming from all four compass directions when only one or two were enforced by cameras.  An attachment to that email was an email I sent the City in April 2015, in which I had previously questioned the amount of the rent.

    C.  On Aug. 25 Jay Beeber of SaferStreetsLA submitted a study showing that safety improved at only one intersection.  At their meeting later that day the City Council asked staff to do a further study.

    D.  Staff's new study, which was to be heard at the Oct. 27 Council meeting, became available to the public on Oct. 23.

    E.  On Oct. 25 I submitted an email suggesting that much of the improvement claimed in Garden Grove could be attributed to the same things - air bags, anti-lock brakes, fewer DUI drivers - that brought a 20% statewide reduction in injury accidents over ten years. 

    F.  On or about Oct. 27 Jay Beeber submitted his analysis of staff's new study, and pointed out the big drop in ticketing that began in January of that year. 

Personally, I identify with what Mark Twain wrote about statistics:  "Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: 'There are three kinds of lies: lies, damned lies and statistics.' "   So, when the question is, "Do cameras reduce injuries," the study I like the most - because it does not rely on statistical analysis - is the one San Francisco first published in August 2015 to satisfy the new CVC 21455.5(i) requirement to publish an annual report.  The October 2016 edition of that study is attached.  It takes about five minutes to read.
In a May 2016 letter to me, San Francisco's Chief Traffic Engineer Ricardo Olea said:

    "You are correct that engineering changes are the most effective way to reduce red light running crashes. We’ve had a long-standing record of improving intersection safety through signal upgrade improvements and signal timing changes."  "We are the process of starting a new Red Light Camera contract which will reduce the total number of approaches being enforced in San Francisco, keeping some locations we believe are still needed based on crash and citation history."

Later in 2016 San Francisco reduced ticketing by 76%; during the five-month period September 2016 thru January 2017 they issued just 1273 tickets  compared to the average 5310 tickets they issued in the same five-month periods a year and two years before.  (Garden Grove issued 2912 tickets in September 2016 thru January 2017.)

How did San Francisco arrive at their decision to downsize?  Their annual report, mentioned above, found that the installation of a red light camera seldom was followed by a drop in accidents.  Instead, the drops occurred after engineering improvements like making the yellows longer, adding an all-red interval (both of which are cheap to do), or a general upgrade to the signal. (In one instance - see page 12 of the report - staff conceded what one of the graphs shows, that the camera may have had no effect whatsoever.)

In August 2015 Redflex reported to you that 79% of Garden Grove's tickets were going to visitors.  Garden Grove is in the middle of a megalopolis with an ever-changing population of visitors who will never have their driving behavior improved by cameras they don't know are there.  If a city genuinely wants to minimize running and accidents by visitors (and "locals" too), it should do the following things to make the problematic intersections stand out, look more important.

       a.  Put up more visible signal lights (larger diameter, with bigger backboards, with more of them placed on the "near" side of the wider intersections).
       b.  Paint "signal ahead" on the pavement.
       c.  Install lighted overhead street signs for the cross street (also placed on the "near" side), and larger bulbs in the streetlights at the intersection.

Then there is the issue of rolling right turns, which were 48% of Garden Grove's camera tickets in 2015.  Lengthening the yellow won't reduce the number of rolling right turn violations, nor will heavy ticketing (blame the visitors), so I suggest that the council ask staff to identify the specific intersections where the danger from rolling right turns remains high and consider installing "blank out" signs programmed to light up and prohibit all right turns during the riskiest portions of the signal cycle.  If pedestrians or bicyclists being run down, the City should not stand by and allow rolling right turns to continue unimpeded.

Now - before a new four-year contract is signed - the council should be provided with a report about whether the cameras actually improve safety.  That report should also include details of any engineering changes staff may have made to improve safety.  A suggested format would be that used in the San Francisco study.

If staff or Redflex submits accident statistics the morning of the 25th, I would have two objections:

1.  The last minute presentation of said statistics would deny the public a suitable opportunity to review their veracity.

2.  With all due respect to City and police staff, any statistical analysis should be done by a professional who has academic credentials in the field of statistics and has not done previous work for the City.

Please also consider these safety comments from other SoCal cities.

    Gardena (cameras installed in 2005, removed in 2011):  "Our research in Gardena has revealed there is no significant traffic safety impact as a result of the use of the red light cameras. At almost every intersection where we have cameras, collisions have remained the same, decreased very slightly, or increased depending on the intersection you examine. When combining the statistics of all the intersections, the overall consensus is that there is not a noticeable safety enhancement to the public."  Chief of Police Edward Medrano, in memo presented at 2-9-10 council meeting.

    Bell Gardens (cameras installed in 2009, removed in 2012):  "To date, 95% of the funds collected from verifiable violations have been paid to RedFlex Traffic Systems for operating the cameras.  The remaining 5% of funds collected have been utilized to partially offset costs of personnel to manage the system.  The red light camera program has contributed to a moderate decrease in the overall number of accidents; however, no change in the overall number of injury accidents. Furthermore, the police department has recognized unanticipated personnel costs to manage the program.  Based on this analysis, the red light camera program is not significant enough of a community safety benefit to justify the continuation of the program beyond the existing three (3) year agreement term that expires on March 29, 2012."  Staff report presented at 9-26-11 council meeting.

    South Gate (cameras installed in 2003, removed in 2013):  "The most disappointing thing from staff's perspective is the lack of change in behavior at the intersections." "If you look at the statistics that were provided by RedFlex, you didn't see a dramatic impact in the behavior over the years.  In fact, a limited correlation between the implementation of RedFlex and the change in behavior.  That's disappointing in the deployment, not just in this city, but everywhere."  City Manager Michael Flad at council meeting of 9-10-13.

    El Monte (cameras installed in 2003, removed in 2008):  "A comparison of traffic collisions at Redflex monitored intersections vs. non-Redflex monitored intersections revealed that there is no statistical difference in the number of traffic collisions because of Redflex monitoring."   Chief of Police Ken Weldon, in memo presented at 10-21-08 council meeting.  More from El Monte:  "We're spending a lot of staff time on this just to gain $2000 a month." "It doesn't reduce accidents -- that's what our studies and results have come back."  City Manager James W. Mussenden.

    Whittier (cameras installed in 2004, removed in 2010):  "Initially, the red-light program did change behaviors because it did lessen the number of red-light violations but over the long term it didn't appear to lessen the number of injury accidents."  Assistant City Manager Nancy Mendez.

    City of Laguna Woods, California (cameras installed in 2005, closed June 2014):  "Staff studied incidents over a 10-year period of time and found that the number of collisions related to signal violations at the two photo enforced intersections fluctuated slightly, but did not change in any significant manner after initiation of the red light photo enforcement program."  City Manager Christopher Macon in staff report prepared for 5-28-14 council item.

    City of Escondido, California (cameras installed in 2004, removed in 2013):  "Staff's analysis is, the data on accident rates is inconclusive." "We didn't find any change between photo enforced intersections and citywide. You're just as likely to be injured at a photo enforced intersection as you are citywide. So we didn't find anything to demonstrate that severity had been reduced."  "Photo enforcement has the highest cost of all the countermeasures."  Escondido Assistant Director of Public Works Julie Procopio. 

III.   Because Senator Hill's bill to halve the fine for rolling right turns is moving thru the Legislature, it is important to watch the finances, including the proposed monthly rent to be paid to Redflex.  At $2200 per active camera (after adjustment for three free cameras) the rent is at least 47% above market.  Del Mar pays $1578 for each of its three cameras.  Redflex' contract with Elk Grove - which has five cameras - sets a rent of $1500 for cameras ten years old or more.  Laguna Woods, which had five cameras, was paying Redflex $1750 per camera.  Why was Garden Grove not able to obtain a better price, considering its larger scale, the advanced age of the equipment and the City's willingness to sign a four year no-escape contract?  If the City agrees to pay the rent proposed, it will pay $470,400 extra rent over the four years (when compared to a feasible target rent of $1500) and will need to issue an extra 4704 tickets in order to cover that extra rent.  (Assuming the City receives $100 revenue from each ticket issued.)

Conclusion

All of the above suggests that it would be unwise, and premature, to sign a four year contract at this time. 

Sincerely,

(highwayrobbery.net)

Attached: 

San Francisco annual report