Tempe Traffic Collisions

There have been 4 bicyclists killed in Tempe traffic collisions on city streets (excludes limited-access highways) over the past 5 years [five year period 2010-2014); at the same time dozens of motorists and 14 pedestrians have been killed.

[UPDATed totals:  There have been a total of 9 bicyclist traffic fatalities in the past 13 years, 2009-2021]

Skip below to #fatals for an incident-by-incident description

Scope of the Problem

For the figures in this report, a 5-year time period, 2010-2014 was selected (the most recent available at the time; 2015 & 16 has since become available). Many additional crashes occur (14,000!) occur in Tempe on limited access highways; these crashes are not included.

There were 21,167 traffic crashes reported in the period, below are the number of people, by person type, involved:

ePersonType	count
DRIVER	41476 *
PEDESTRIAN	436 *
PEDALCYCLIST	1107 *
PASSENGER	14168 *

* not counting limited-access highways: I have the data, I just didn’t bother to pull it. contact me if you’re interested

Below are the number of people, by person type, killed:

Persons	count
MOTORIST	21 *
PEDESTRIAN	13 *
PEDALCYCLIST	4

* not counting limited-access highways: there were an additional 32 motorists, and 1 pedestrian killed on freeways in Tempe city limits.

Listing of each Bicyclist Fatality Since 2009 (#fatal)

The fourteen fatal bicyclist crashes (covering years 2009 – 2022, 2023 partial ), click on the date for more details:

• 2010-05-10 Right angle at signalized intersection. Cyclist disregarded signal.
• 2010-05-17  Right angle at signalized intersection. Motorist disregarded signal.
• 2013-10-03 Right angle at signalized intersection. Wrong-way sidewalk cyclist collided with right-turning motorist.
• 2014-06-12 Right angle at signalized intersection. Wrong-way sidewalk cyclist collided with turning motorist. [this was coded as Guadalupe, apparently in error(?), anyway it’s included here, click thru the date for full details]
• 2016-02-02 Right angle at signalized intersection (freeway off-ramp). Bicyclist is said to have ignored signal.
• 2017-11-16 Bicyclist in crosswalk struck by motorist turning left; both had “green”.
• 2018-12-06 right-angle at signalized intersection; driver arrested DUI; click thru for latest.
• 2019-01-23 A cyclist was struck from behind while attempting to turn left by a distracted driver.
• 2021-02-19 17:13:00; no info other than what is in database.  WB cyclist improper merge was struck by WB motorist.
• 2021-02-21 18:26:00 no info other than what is in database.  WB cyclist improper merge was struck by WB motorist (a motorcyclist to be more specific).
• 2021-04-02 A cyclist was killed at signalized intersection of Mill and Baseline; it involved a hit-and-run driver who was arrested for DUI.
• 2022-02-03 16:38:00 no info other than what is in database; Priest and 10th Pl. Directions given are confusing, and FARS not yet available.
• 2022-08-25 just before midnight: just what is in database; police say a cyclist was crossing University Dr near Perry when a WB hit-run driver of a toyota SUV struck and killed her.
• 2023-02-14 cyclist failed to stop at 4-way stop College and Alameda.

Note that, for arcane statistical purposes, ebikes and motorized bicycle are counted as motorcycle, not as bicyclists.  In addition to the above a motorized bicyclist was killed 2010-07-10 when a motorist left-crossed the bicyclist at a signalized intersection and fled the scene. And in 2020-05-15 a motorized bicyclist was killed while traveling along sidewalk and crosswalk when he was struck by a motorist who had a stop sign.

In any event, consistent with other urban areas, strike-from-behind tends to be rare. Of the eight total fatalities in twelve years (2009-2020) there has been at most one; the Jan 2019 incident was arguably a strike from behind, but not classic since it involved a bicyclist changing lanes)

Amount of Bicycling

Although bicycle use data is sparse, Tempe has more than most US cities, here are a couple of data sources

• Bike counts at selected locations have been conducted for the past several years by the Tempe Bicycle Action Group []. The bike count data is particularly interesting because it also measures the amount of wrong-way, and sidewalk bicycling.
• The amount of commuting by bicycle is often used as a proxy for relative amounts of bicycling.  The ACS (American Community Survey; conducted by the US Census Bureau) shows that 4.2% (margin of error 0.6%) of workers commuted by bike for the period 2008-2012[]. An LAB report[] based on just 2012 ACS says the rate is 4.5%.

2015 Update

Without re-hashing the 5-year figures above; a couple of observations. One is there were no bicyclist fatalities in Tempe in either 2009 or 2015 — leaving the 7-year toll at 4.

There continues to be a noticeable drop in the overall number of bicyclist incidents which began on the state-level in 2012 (more here). In the city of Tempe, the year over year drop in 2015 to 132 represents a drop of almost 33%. It’s not known why there was such a sharp drop (less being reported?), but it is well below the 2009-2014 trend, here are those figures ( get query here ):

204 / 198 / 223 / 245 / 238 / 210 / 132 (and more recently 2016/2017 were 122/122)

Does Targeted Enforcement Work? (#target)

OK, that’s a teaser headline because I have no idea. It does seem like it would be wise to link enforcement to traffic collision history. Unfortunately this does not seem to ever be the case.  E.g. it was reported on 7/26/2017 there there’s some sort of concerted effort by Tempe PD going on all week at the intersection of College and Alameda, a 4-way stop. I have noted elsewhere, with videos, that almost 100% of drivers (and presumably bicyclists) do not make a full stop when there is no conflicting traffic. This is a violation.

This is the intersection of two low-speed streets; with relatively low (vehicle) volume (and very high bicycle volume, as College is a major bicycling artery); and since 4-way stops tend to be inherently very safe… it would be expected to be very safe, and lo and behold… What does the data show? Reported in EIGHT YEARS:

• There were ZERO bike-MV collisions
• There were five MV collisions; four were no injury, one possible injury
• There was one ped-MV where the ped suffered a non-incap injury. The driver was at fault for running the stop sign, and hit-and-ran (this is a felony).

There were other reports from spring 2018 of similar-sounding enforcement actions at 13th and Farmer, and 5th and Farmer.

Here’s a 2019 news article about Tempe PD receiving grants for extra enforcement, in it TPD there’s a hot-spot map of high-crash areas, all of which of course involve where high-speed roads meet, and the claim that they are going to do “targeted enforcement in those areas”; let’s make sure they do, and do NOT target low-speed, low-volume intersections like those mentioned above. Here is the map (the numbered places are the hot spots, Top 10 Tempe Collision Locations, 2016-2019, I added a lightening bolt at the places where it was mentioned there appeared to be bicyclist stings in the past)

I’ll mention this here because I don’t know where else to put it: The other highly-likely selective enforcement example I am aware of is the all-way stop intersection at 3rd Street and Treat Avenue in Tucson (news article)(bicycletucson.com notes that it’s been upgraded to a roundabout in Jul 2022). Also a stop sign, also low-speed streets, also high-volume of bicyclists (4,000/day according to the news story), also more-or-less perfectly statistically safe… nine years (2009-17) of crash records show one reported bike-MV collision, resulting in no injuries. That works out to over 70 million traversals of that intersection by bicyclists with one no-injury crash reported. [Update 2022; City of Tucson will be installing a traffic circle at this location, see below]

Crash Map

Below is a  static snap-shot of the location of each fatal collision in the city (excludes freeways), for each fatality where a location was reported (i need to double check; I’m guessing all fatals reported a location). I’m looking for better icons. (the graphic is based on 2009-16; there are more now).

Sorry, Google discontinued fusion tables in 2019 🙁
The image below does correctly reflect the period 2009-2017

Click  here for a fusion table/map of all Tempe crashes. (currently covers the period 2009-2017) There you can filter by victim-type (ped, bicyclist, motorist, etc), and any other of many fields. To get only City of Tempe PD, to exclude freeway crashes, simply filter on Oncic and un-tick the DPS box…

.

References

Tempe Bicycle Action Group 2014 Count Data.

Modes Less Traveled—Bicycling and Walking to Work in the United States: 2008–2012 [.pdf][supplemental tables]. Also, 3.7% (margin of error 0.7%) walked to work, see supplemental Table 5.

WHERE WE RIDE Analysis of bicycling in American cities. the LAB. [.pdf]

I also learned the PQI (Pavement Quality Index), as well as ADT volume data, and a whole bunch of other city-related datasets are all freely available (nice job, City of Tempe!):

• Links moved to azbikelaw.org/warner-resurfacing/#PQI (these links apparently go dead often) Note that the methodology used for PQI data has a flaw in that it measures quality in the curb lane. If there is a bike lane, the quality of the bike lane can be, and tends to be, worse than the adjacent lane.

(just as a cross-reference; as of 2019 City of Phoenix has an extensive interactive pavement management plan page linked here but I don’t see numerical PQ data)

Queries

The first set of three are Tempe PD only (excludes DPS / highways), Count people by persontype, and total number of crashes; the second set  counts by city — i.e. includes both Tempe PD, and DPS investigated (and a small number of other agencies, e.g. sheriff’s office).

SELECT count(*)
FROM ( 2012_incident AS i )
LEFT OUTER JOIN LOVNcic ON i.OfficerNcic = LOVNcic.ID
WHERE LOVNcic.name ='Tempe' AND i.IncidentYear IN ( 2010, 2011, 2012, 2013, 2014, 2015, 2016) ;

SELECT p.ePersonType, count(*) FROM ( 2012_incident AS i JOIN 2012_person AS p ON i.IncidentID = p.IncidentID) LEFT OUTER JOIN LOVNcic ON i.OfficerNcic = LOVNcic.ID WHERE LOVNcic.name ='Tempe'  AND i.IncidentYear IN (2010, 2011, 2012, 2013, 2014, 2015, 2016) GROUP BY 1;

SELECT SUM(TotalMotoristsFatalities), SUM(TotalNonMotoristsFatalities),
SUM(TotalMotoristsInjuries), SUM(TotalNonMotoristsInjuries)
FROM ( 2012_incident AS i )
LEFT OUTER JOIN LOVNcic ON i.OfficerNcic = LOVNcic.ID
WHERE LOVNcic.name ='Tempe' AND i.IncidentYear IN (2010, 2011, 2012, 2013, 2014, 2015, 2016) ;

SELECT count(*)
FROM ( 2012_incident AS i )
LEFT OUTER JOIN LOVCity ON i.CityId = LOVCity.id
WHERE LOVCity.name ='Tempe' AND i.IncidentYear IN (2010, 2011, 2012, 2013, 2014, 2015, 2016) ;

SELECT p.ePersonType, count(*)
FROM ( 2012_incident AS i JOIN 2012_person AS p ON i.IncidentID = p.IncidentID)
LEFT OUTER JOIN LOVCity ON i.CityId = LOVCity.id
WHERE LOVCity.name ='Tempe' AND i.IncidentYear IN (2010, 2011, 2012, 2013, 2014, 2015, 2016)  GROUP BY 1;

SELECT SUM(TotalMotoristsFatalities), SUM(TotalNonMotoristsFatalities),
SUM(TotalMotoristsInjuries), SUM(TotalNonMotoristsInjuries)
FROM ( 2012_incident AS i )
LEFT OUTER JOIN LOVCity ON i.CityId = LOVCity.Id
WHERE LOVCity.name ='Tempe' AND i.IncidentYear IN (2010, 2011, 2012, 2013, 2014, 2015, 2016) ;

select * from incident where (OnRoad LIKE "Alameda%" and CrossingFeature LIKE "College%") OR (OnRoad LIKE "College%" and CrossingFeature LIKE "Alameda%");

todo, make this into a fusion table (or do OffNcic=tempe?):
SELECT
i.IncidentID, i.IncidentDateTime, i.IncidentYear,
sF_Bicycle, sF_Pedestrian, sF_Motorist, sF_OtherUnk, sF_Motorcycle,
i.OfficerId, i.OfficerNcic, i.ExtendedNcic,
OffNcic.name Oncic,ExtNcic.name Encic,
i.eCollisionManner,
i.eLightCondition,
i.TotalUnits,
i.TotalMotorists, i.TotalNonMotorists,
i.TotalInjuries, i.TotalFatalities, i.TotalMotoristsInjuries, i.TotalNonMotoristsInjuries, 
i.TotalMotoristsFatalities, TotalNonMotoristsFatalities,
i.eInjurySeverity,
i.InjurySeverity,
i.HitAndRunFlag,
LOVCity.name City, 
i.OnRoad, i.CrossingFeature,
i.Latitude, i.Longitude,
i.eTrafficWayType,
i.eIntersectionType,
i.eJunctionRelation,
i.eWeather
FROM ( 2012_incident AS i )
LEFT OUTER JOIN LOVNcic AS OffNcic ON i.OfficerNcic = OffNcic.id
LEFT OUTER JOIN LOVNcic AS ExtNcic ON i.ExtendedNcic = ExtNcic.id
LEFT OUTER JOIN LOVCity ON i.CityId = LOVCity.id
WHERE LOVCity.name ='Tempe' AND i.IncidentYear IN ( 2010, 2011, 2012, 2013, 2014, 2015, 2016) AND Latitude>0 ;

This is for the fusion table's info window layout:
<div class='googft-info-window'>
<b>IncidentID:</b> <a target="_blank" href="http://mysql.azbikelaw.org/asdmphp/queryAsdmIncident.php?incident={IncidentID}">{IncidentID}</a><br>
<b>IncidentDateTime:</b> {IncidentDateTime}<br>
<b>sF_Bicycle:</b> {sF_Bicycle}<br>
<b>sF_Pedestrian:</b> {sF_Pedestrian}<br>
<b>sF_Motorcycle:</b> {sF_Motorcycle}<br>
<b>OfficerNcic:</b> {OfficerNcic}
<b> ExtendedNcic:</b> {ExtendedNcic}<br>
<b>CollisionManner:</b> {eCollisionManner}<br>
<b>LightCond:</b> {eLightCondition}
<b>TotalNumUnits:</b> {TotalUnits}<br>
<b>InjurySeverity:</b> {eInjurySeverity}<br>
<b>HitAndRunFlag:</b> {HitAndRunFlag}
<b>Weather:</b> {eWeather}<br>
<b>OnRoad:</b> {OnRoad}
<b> Cross St:</b> {CrossingFeature}<br>
<b>JuncRelation:</b> {eJunctionRelation}<br>
</div>

These are excel formulas for icons in the fusion table:
=IF(AA2=1;"small_green";IF(AA2=2;"small_blue";IF(AA2=3;"small_yellow";IF(AA2=4;"small_red";IF(AA2=5;"placemark_circle";"null") ))))
=IF(D2=1;"cycling";IF(E2=1;"man";IF(H2=1;"motorcycling";"cab")))
and also add a column to the right CopyOfLatitude (i don't remember why?)