1. Impacts Session 4 SOP Chapter 7
Seminar Session 4 first summarizes impacts of education and enforcement, regulatory, and psycho-perception measures on vehicle operating speeds
Session 4 then reviews in more detail the evidence of traffic calming impacts on:
• Vehicle operating speeds
• Traffic volumes
• Collision rates
• Crime rates
• Land values, walking and bicycling activity levels and noise levels
Literally hundreds of before-and-after studies from communities around the U.S. are used to estimate speed and volume impact models
A smaller sample of studies is used to test for statistically significant impacts on collisions

2. Impacts of Regulatory Measures
Stop Signs
Turn Restrictions
Restrictive One-Way Streets
Conversion Back to Two-Way Operation
“Rest on Green” and “Rest on Red”
Mixed Impact on Midblock
Speeds -- Some Reduction in
Cut-Through Volumes
Significant Reduction in
Peak-Hour Volumes But Many Violations
No Evidence Available
SOP Chapter 5
Table includes a rough summary of the evidence
For details, see SOP report

3. Impacts of Education and Enforcement Measures
Neighborhood Traffic Safety Campaigns
Radar Speed Display Units
Neighborhood Speed Watch
Targeted Police Enforcement
Photo-Radar Speed Enforcement
No Evidence Available
Short-Lived Reduction in
Speeds
Generally Little Impact on
Speeds (with one
exception)
Reduction in Speeds Only
During Enforcement Period
Significant Reduction in
Speeds in Some Cases
but Uneven
SOP Chapter 5
Table includes a rough summary of the evidence
For details, see SOP report

4. Impacts of Psycho-Perception Measures
Centerline and Edgeline Striping
Transverse Markings
Small Building Setbacks
Street Trees
No Change or Increase in
Speeds
Speed Reduction According
to One Study
No Effect on Speeds
According to One Study
SOP Chapter 5
Table includes a rough summary of the evidence
For details, see SOP report

5. Overall Assessment of Non-Physical Measures
Most Have Uneven or Limited Impacts on Speeding
High-Tech Options Have Yet to Prove Themselves
The More Promising Approaches Are Costly
No reference in SOP report
Slide summarizes the previous three slides

6. Speed and Volume Impacts -- Somewhat Predictable
SOP Chapter 5
Impacts of traffic calming measures are case-specific, depending on:
Geometrics and spacing of measures
Availability of alternate routes
Treatment of other streets in areawide applications
Many other factors
Still, impacts can be generalized to a degree:
Semi-diverter on left would be expected to, and does, greatly reduce traffic volumes by prohibiting most movements along this local street in Sarasota
Semi-diverter also would be expected to, and does, reduce vehicle operating speeds by limiting the straightaway to a few hundred feet
Speed tables on right would be expected to, and do, have more modest impact on speeds and volumes along this main thoroughfare in Charlotte
Speed tables produce 85th percentile speeds at the tables themselves of 27 mph, and spaced hundreds of feet apart, have much higher midpoint speeds
Diversion in this case is limited by lack of nearby parallel routes and by the maintenance of moderately high speeds along this route
38 -> 23 mph
224 -> 92 vpd
40 ->37 mph
13,000 -> 10,300 vpd

7. Speeds Between Slow Points
(85th Percentile)
No. Speed Change Percent Sites After in Speed Change
12' Humps mph -7.6 mph -22%
14' Humps
22' Tables
Circles
Narrowings
Half Closures
Diverters
SOP Chapter 5
In an effort to generalize impacts, before-and-after results from hundreds of studies have been analyzed
Summary statistics for seven types of traffic calming measures are shown in this table
Statistics for other types of measures are presented in the SOP report
For example, after installation of 12’ humps:
85th percentile midpoint speeds drop on average to 27 mph
85th percentile midpoint speeds drop by an average of almost 8 mph or 22%
Contrast this with 22’ tables -- midpoint speeds after installation are higher with tables than with humps, and absolute and percentage declines are less
This is as expected -- speed tables are applied to higher speed streets than are speed humps, and have much higher crossing speeds themselves
Such ballpark estimates of impact are subject to caveats:
Large standard deviations from average values
No indication of where and when speed measurements were taken
Small sample sizes for some measures

8. Speed Profile of a 14-Foot Hump
SOP Chapter 5
Effects of geometrics and spacing are captured in this figure and the next
From the Portland Bureau of Traffic Management
Before they were traffic calmed, streets treated with 14-foot speed humps had 85th percentile speeds averaging 32 mph
After traffic calming, 85th percentile speeds fell to about 21 mph at the humps themselves, 26 mph a hundred feet upstream, and 25 mph one hundred feet downstream
Note:
Drivers decelerate rapidly upon approaching slow points, and accelerate rapidly upon departing them
Nonetheless, speeds do not rise to their pre-treatment levels even at considerable distances from slow points
Source: City of Portland, Oregon

9. Speed Profile of a 22-Foot Table
SOP Chapter 5
22’ speed tables are deployed on higher speed streets than are 14’ humps
Before they were traffic calmed, streets treated with 22’ tables had 85th percentile speeds averaging 40 mph
After traffic calming, 85th percentile speeds fell to 27 mph at the tables themselves, 33 mph one hundred feet upstream, and 30 mph one hundred feet downstream
As with 14’ humps, deceleration and acceleration are rapid, yet speeds do not appear to return to pre-treatment levels even at considerable distances from tables
Source: City of Portland, Oregon

10. Midpoint Speed vs. Hump Spacing
Before
After
SOP Chapter 5
Given results like Portland’s, it should be possible to model midpoint speeds in terms of speeds crossing slow points and distances between slow points
Many before-and-after studies report the spacing of slow points -- specifically humps and tables -- as well as before and after speeds
These studies provide a database sufficient to model midpoint speeds
As depicted in the figure, speeds between slow points increase about ½ to 1 mph for each 100 feet of separation distance. Even with wide spacing, midpoint speeds are still lower than before traffic calming.
The best fit equation was
V85 = 0.005s
where V85 = 85th percentile speed in mph
s = spacing in feet
Because of limitations on data outlined on page 103, the fit was not very strong (R2 = )

11. Volume Impacts of Traffic Calming Measures
No. Average Change Percent Sites in Volume Change
12' Humps %
14' Humps
22' Tables
Circles
Narrowings
Half Closures
Diverters
SOP Chapter 5
In an effort to generalize impacts, before-and-after results from hundreds of studies have been analyzed
Summary statistics for seven types of traffic calming measures are shown in this table
Statistics for additional measures are presented in the SOP report
Streets with humps experience on average an 18% decrease in volume depending on available alternate routes
Closures produce the greatest traffic diversion and traffic circles produce the least
This is as expected -- speed humps are known to divert traffic to a greater extent than do circles -- indeed, conventional wisdom holds that circles simply slow traffic without diversion
Such ballpark estimates of impact are subject to caveats:
Large standard deviations from average values, much larger for volume impacts than speed impacts
Inconsistent measurement locations, sometimes in the same block as the treatment, other times separated by a block or two
Small sample sizes for some measures

12. Impact of Alternate Routes
Speed Change Volume Change
(mph) (veh/day)
162nd Ave > 27 1,472 -> 1,071
SE 63rd St > 25 2,456 -> 2,593
SOP Chapter 5
Volume impacts are much more complex and case-specific than are speed impacts
They depend on the entire network of which a street is a part, not just the characteristics of the street itself
Particularly important are the availability of alternate routes and the application of other measures in areawide schemes
Compare volume impacts of two virtually identical speed hump installations in Bellevue:
SE 63rd St without convenient alternate routes
162nd Ave SE with convenient alternatives

13. Impact of Geometrics and Spacing
Hump Speed Changes Volume Changes Design (mph) ( vpd)
Initial 12' x 3-3/4" 39 -> 22 (midpoint) > ' apart (at humps) Redesign 12' x 3" 22 -> 27 (midpoint) > ' apart (at humps)
SOP Chapter 5
Volume impacts also depend on the geometrics and spacing of slow points
Somerset Drive in Bellevue was initially treated with 12-foot humps of 3- inch height spaced an average of 150 feet apart
After many complaints from residents, the humps were reinstalled at a height of 3 inches and average spacing of 340 feet
When first treated, daily traffic volumes dropped by a third, with significant diversion to parallel local streets
When the number and height of humps were reduced, daily volumes nearly returned to their pre-treatment levels
Map shows final treatment

14. Volume Impact Model -- Volume Control Measures
Grand Mean % Volume Change %
Deviations from the Grand Mean Full Closures (19 cases) - 5%
Half Closures (53 cases) - 3%
Diagonal Diverters (27 cases) + 5%
Other (10 cases) + 9%
Coefficient of the Covariate Blocks from Measure + 5%
SOP Chapter 5
Given results like Bellevue’s, it should be possible to model changes in traffic volumes in terms of types of measures employed, degree of speed reduction, and distance from traffic calming measures
Many before-and-after studies report all of the above
These studies provide a database sufficient to model midpoint speeds
Multiple classification analysis was used to model volume impacts of closures, diverters, and other volume control measures
Volume controls categorically reduce traffic volumes by 39 percent
As expected, full closures cause the greatest reduction in traffic volumes -- an additional 5 percent beyond the grand mean reduction or a 44 percent reduction
Other volume controls cause less reduction; though differences in impact are not statistically different (p =.71)
Each additional block from a traffic calming measure lessens the impact on traffic volumes by 5 percent (not statistically significant p = .16)

15. Volume Impact Model -- Speed Control Measures
Grand Mean % Volume Change -15%
Deviations from the Grand Mean
Humps (144 cases) - 5%
Tables (56 cases) + 1%
Circles (40 cases) + 1%
Other (22 cases) + 6%
Coefficient of the Covariate
% Speed Change 0.2
SOP Chapter 5
See notes on previous slide
Multiple classification analysis was also used to model volume impacts of speed humps, traffic circles, and other speed control measures
Speed control measures categorically reduce traffic volumes by about 15 percent
The type of measure has a statistically significant effect on % traffic volume reduction (p = .001)
Speed humps cause the greatest traffic diversion, resulting in an additional 5 percent reduction beyond the grand mean (-15%-5% = -20% on average)
% reduction in traffic volume is only marginally related to the % reduction in speed -- all else being equal, a 10 percent drop in speed will cause a 2 percent drop in volume (not statistically significant p=.33)

16. Cost Savings Due to Accident Reduction (Seattle)
Accidents Cost per Cost Savings Prevented Accident ( ) ( )
Non-Injury Accidents $6,500 $1,774,500
Injury Accidents $30,000 $8,310,000
All Accidents $10,084,500
SOP Chapter 5
The most politically consequential impacts of traffic calming are in the area of safety
Faced with budget cuts in 1996, the Seattle Transportation Division resumed its accident analyses and reiterated safety as a departmental priority
Savings in property and casualty losses were estimated to be in the millions of dollars each year
Traffic calming was spared budget cuts

17. Average Reduction in Collisions by Measure (International Experience)
SOP Chapter 5
By slowing traffic, eliminating conflicting movements, and/or sharpening drivers’ attention, traffic calming may result in fewer collisions
Due to lower speeds, collisions may be less serious when they occur
Among the 43 international case studies, collision frequencies declined by anywhere from 8 to 100 percent
These results should be used with caution
The source does not provide important details such as method of analysis or length of before and after periods
The studies did not take into account traffic diversion, trend over time, regression to the mean or other factors that could affect the validity and reliability of the results
With these caveats in mind
Traffic circles and chicanes had the most favorable impacts on safety, reducing collision frequencies by an average of 82 percent
Other measures, even humps, had very favorable impacts
Physical measures outperformed regulatory measures
Results are somewhat counterintuitive -- see discussion in the SOP report
Source: E. Geddes et al., Safety Benefits of Traffic Calming, Insurance Corporation of British Columbia, Vancouver, 1996, p. 38.

18. Safety Impacts of Traffic Calming Measures (U.S. Experience)
No. Avg. Annual Collisions Sig.
Measure Sites Before After Change Level
12’Humps % .40
14’ Humps % .18
22’ Tables % <.01
Circles w/o Seattle % .04
All Circles % <.01
Overall % <.01
exclude Seattle % <.01
SOP Chapter 5
While positive, the impacts of U.S. traffic calming are less favorable than impacts internationally, perhaps because U.S. practice focuses on low-volume streets with few collisions to begin with
A difference-of-means test for paired samples was used to check for significant changes in collision frequencies after traffic calming
The average annual number of collisions fell with each type of traffic calming measure -- the decline is statistically significant for speed tables and neighborhood traffic circles but not for speed humps
Overall collisions decreased on average by 51%, including Seattle circles, and by 25 percent, excluding Seattle circles -- the declines are significant
These results do not take traffic diversion into account
The Seattle results relate only to intersection collisions

19. Safety Impacts of Traffic Calming After Adjusting for Traffic Diversion
No. % Sig.
Sites Change Level
Accidents % <.01
Accident Rate %
Most studies report a reduction in traffic on traffic calmed streets which may explain the reduction in collisions
Volume data were available for 55 streets in the larger database
For these streets, collisions decreased 27%
However collision rates, which adjust for the reduction in traffic, only declined 4 % (which is not statistically different from 0)
This illustrates the importance of accounting for traffic diversion
Also, some collisions may migrate to other streets as a result of traffic diversion -- for a comprehensive view of the safety impacts of traffic calming, it is important to examine a wide area including streets with and without traffic calming

20. European vs. U.S. Traffic Calming
SOP Chapter 5
Why is European traffic calming more successful in reducing collisions than U.S. traffic calming?
European treatments are often:
more intensive, involving multiple measures and larger speed reductions
applied to higher order roads with more collisions to begin with
Most U.S. treatments are add-ons to local streets which already have low frequency of collisions due to low traffic volumes, off-street parking, better sight lines, etc. (thus less headroom for improvement)
The left photo is a combined speed table, choker, and bend on Lowther Street in York, England
It is part of an areawide program of one ways, full closures, speed humps, chokers, and parking bays
The effect of traffic calming on Lowther Street was to reduce speeds from 27 to 15 mph
The right photo is one in a series of 12’ speed humps on Cleveland Street in Tampa, FL
Humps are spaced 680 feet apart and, at the time of the before-and-after study, were the only traffic calming measure in place
Midpoint speeds fell from 40 to 34 mph
British Installation
Change in Speed -45%
Change in Collisions -57%
U.S. Installation
Change in Speed -15%
Change in Collisions +13%

21. Other Impacts of Traffic Calming
Crime
Street Life
Property Values
Noise
SOP Chapter 5
See slides that follow

22. Five Oaks Neighborhood Stabilization Plan -- Significant Impact on Crime
SOP Chapter 5
Massive intervention in the Five Oaks neighborhood of Dayton was successful in reducing crime
See Seminar Session 1, Slide 10
Not only streets but alleys were closed to through-traffic, as in this slide

23. Street Closures in Ft. Lauderdale -- Little Impact on Crime
SOP Chapter 5
Most traffic calming treatments have not been as successful in reducing crime as has Dayton’s Five Oaks Neighborhood Stabilization Plan
Street closures in the crime-ridden Riverside Park neighborhood of Ft. Lauderdale caused no apparent reduction in serious crimes
Where much of the crime is internally generated, closing off a neighborhood would not be expected to have much impact on crime within
This may be the case in Riverside Park
Also, the Riverside Park intervention consisted only of street closures, while the Five Oaks intervention involved many complementary activities
Closures

24. Speed Humps in Berkeley -- Mixed Impact on Crime
SOP Chapter 5
About 100 Berkeley blocks have been treated with speed humps in an anti-drug campaign
Humps have been found to impede escape, generally deter reckless driving, and demonstrate the city’s commitment to the area
They are often combined with better lighting, street clean-up, neighborhood crime watch, and heightened police presence
Of seven blocks for which multi-year crime data are available, drug-related crimes were down for five of them, and up for two
The random pattern caused the Police Special Enforcement Unit to conclude that “speed humps generally have no discernable impact on the amount of criminal activity on a street”
That assessment may be too pessimistic, particularly given resident perceptions that humps have discouraged drug dealing

25. European Impacts on Bicycling and Walking
SOP Chapter 5
European studies suggest that quieter and safer street environments after traffic calming encourage walking, bicycling, and street life
Pictured street is in Vinderup, Denmark, where traffic calming of an intercity route through this small town had the indicated effect on bicyclist and pedestrian traffic counts
Motor vehicle counts are for 24 hours, the bicyclist and pedestrian counts for 8 hours

26. Mode Shifts in Berkeley
SOP Chapter 5
Europeans calm traffic with such flare for design, and in such pedestrian-friendly environments to begin with, that their results may have little applicability to the U.S.
The only known U.S. study of impacts on bicycling and walking is from Berkeley, where Milvia Street was reconstructed as a “slow street” with neckdowns, chicanes, speed humps, and center islands
Mode shifts were as indicated

27. Impacts of Traffic Calming on Property Values
SOP Chapter 5
In theory, traffic calming could raise property values by reducing the negative effects of the automobile, or could lower property values by advertising a traffic problem
There are two rigorous U.S. studies of traffic calming impacts on property values
One showed significant appreciation following traffic calming, the other did not
This figure, from the former study, shows residential properties appreciating by more than 5 percent over 13 years
Most of the appreciating properties are in the Dickinson Neighborhood of Grand Rapids, MI, which was treated with diagonal diverters (righthand neighborhood)
A nearly identical neighborhood, which remained untreated, experienced less property appreciation (lefthand neighborhood)
The other rigorous study, from Gwinnett County, showed no consistent impact on residential property values in neighborhoods treated with speed tables
Different results of the two studies are doubtless due to different measures being employed under different conditions -- see discussion in SOP report
Source: D.G. Bagby, "The Effects of Traffic Flow on Residential Property Values," Journal of the American Planning Association, Vol. 46, 1980, pp

28. Anecdotal Evidence from West Palm Beach
SOP Chapter 5
Beyond the two rigorous studies, there is only anecdotal evidence that traffic calming increases real estate values, as in Downtown West Palm Beach
See Seminar Session 1, Slide 11

29. Impacts of Traffic Calming on Noise Levels
Measure Usual Level Peak Level
None dB 72 dB
(unobstructed traffic)
4-Way Stop
Traffic Circle
Raised Crossing
SOP Chapter 5
Lower speeds that result from proper design and application of traffic calming measures should lower noise levels
The one exception is just downstream of the measures themselves, particularly when cargo-carrying trucks make up a significant fraction of the traffic stream
Evidence of favorable noise impacts comes from Boulder, Charlotte, and San Jose
From the Boulder noise study, traffic circles and raised crossings were perceptibly (more than 3 dB(A)) less noisy than untreated streets
See caveat on SOP report, p. 116

30. Undocumented Impacts (at least in the U.S.)
Energy Consumption
Air Pollution
Retail Trade
Residential Satisfaction
Neighborliness and Sense of Community
No reference in SOP report
Favorable impacts on all of the above have been documented in other countries, but not in the U.S.
These represent areas for further research