Before the Independent Commissioner Under the Resource ... · this evidence and I agree to comply with it in presenting evidence at this hearing. The evidence that I give is within

Before the Independent Commissioner

Under the Resource Management Act 1991

In the matter of an application by Go Media Bacbou Limited for resource consent to install and display a 32m2 electronic billboard at 129 Kāpiti Road, Paraparaumu (RM180154).

Statement of Evidence of Andrew David Carr

20 June 2019

Applicant's solicitors:

Alex Booker

Anderson Lloyd

Level 3, 70 Gloucester Street, Christchurch 8013

PO Box 13831, Armagh, Christchurch 8141

DX Box WX10009

p + 64 3 379 0037 | f + 64 3 379 0039

[email protected]

1903537 | 4441653v01 page 1

Introduction

1 My full name is Andrew (Andy) David Carr.

2 I am a Chartered Professional Engineer and an International Professional

Engineer (New Zealand section of the register). I hold a Masters degree in

Transport Engineering and Operations and also a Masters degree in Business

Administration.

3 I am a member of the national committee of the Resource Management Law

Association and a past Chair of the Canterbury branch of the organisation. I am

also a Member of the Institution of Professional Engineers New Zealand, and an

Associate Member of the New Zealand Planning Institute.

4 I have more than 29 years’ experience in traffic engineering, over which time I

have been responsible for investigating and evaluating the traffic and

transportation impacts of a wide range of land use developments, both in New

Zealand and the United Kingdom.

5 I am presently a director of Carriageway Consulting Ltd, a specialist traffic

engineering and transport planning consultancy which I founded in early 2014.

My role primarily involves undertaking and reviewing traffic analyses for both

resource consent applications and proposed plan changes for a variety of

different development types, for both local authorities and private organisations. I

am also a Hearings Commissioner and have acted in that role for Greater

Wellington Regional Council, Ashburton District Council, Waimakariri District

Council and Christchurch City Council.

6 Prior to forming Carriageway Consulting Ltd, I was employed by traffic

engineering consultancies where I had senior roles in developing the business,

undertaking technical work and supervising project teams primarily within the

South Island.

7 I have carried out a number of commissions which have involved assessing the

transportation-related effects of signage. Within Christchurch, this has included

evaluating applications for digital billboards on Victoria Street, Riccarton Road,

Colombo Street and Grove Road (which are now erected and operating), and on

Fitzgerald Avenue, Blenheim Road and adjacent to Hornby Mall (which are

consented but not yet installed). I have provided advice for proposed digital

billboards in Wellington (two of which are erected and operating), Auckland

(which are presently progressing through the consenting process) and in Dunedin

(consented and operating), and presented evidence for a resource consent

application for off-site static signage in Nelson (which was subsequently

approved).

1903537 | 4441653v01 page 2

8 Also of relevance to this application, I have previously been commissioned to

investigate the road safety records around existing digital billboards within New

Zealand. In brief, the records do not support any conclusion that the operation of

such billboards give rise to an increase in traffic crashes. Independent of any

specific commission, I conducted my own research into the road safety effects of

existing digital billboards within New Zealand by refining the methodology used in

the earlier investigation. This also did not support any conclusion that the

operation of such billboards give rise to an increase in traffic crashes. I discuss

both of these within this evidence.

9 I confirm that I have read the Code of Conduct for Expert Witnesses contained in

the Environment Court Practice Note 2014. I have complied with it in preparing

this evidence and I agree to comply with it in presenting evidence at this hearing.

The evidence that I give is within my area of expertise except where I state that

my evidence is given in reliance on another person’s evidence. I have considered

all material facts that are known to me that might alter or detract from the

opinions that I express in this evidence.

Scope of Evidence

10 In this matter, I have been asked by the applicant to:

(a) Review and comment on the submissions received in respect of its

application; and

(b) Review and comment on the Statements of Evidence of Dr Walton and Mrs

Rushmere.

11 In order to fully address the Statements of Evidence of Dr Walton and Mrs

Rushmere, I have set out a detailed review of the research regarding digital

billboards. This forms the first part of my evidence.

12 For clarity, I was not involved in preparing the documentation or assessments

accompanying the application. However, I can confirm that I have reviewed the

transportation aspects of the application, and particularly the report of Mr Wanty,

and am in agreement with his conclusions that the billboard could proceed

subject to Conditions of Consent. I disagree though with some of these

Conditions, and as a result I have addressed those matters in detail

subsequently.

13 In preparing this statement of evidence I have considered the following

documents:

(a) the AEE accompanying the Application;

(b) submissions relevant to my area of expertise;

1903537 | 4441653v01 page 3

(c) the statement of evidence of Dr Darren Walton dated 9 June 2019;

(d) the statement of evidence of Mrs Suzanne Rushmere dated 6 June 2019;

(e) section 42A report; and

(f) a variety of research articles addressing the road safety effects of digital

billboards (these are referenced within my evidence).

Executive Summary

14 Having evaluated and assessed the transportation aspects of the proposal,

particularly the matter of road safety, I consider that the adverse transportation-

related effects arising from the digital billboard will be negligible. This view is

supported by my review of the available research literature, coupled with my own

research.

15 In this regard, I am of a different opinion to the Council consultant transportation

planners/engineers, Dr Walton and Mrs Rushmere, who consider that adverse

safety effects are likely to arise if the application was to be granted.

The Road Safety Risks of Digital Billboards

16 Prior to addressing the Statements of Evidence and reports of Dr Walton and Mrs

Rushmere, I consider it is helpful to provide some background regarding the road

safety effects associated with digital billboards.

17 I have previously carried out a literature review of documents to assess the likely

road safety effects of digital billboards, and this is attached as Annexure A to my

evidence1. In summary, based on my review, I am of the opinion that the

research shows digital billboards attract driver attention to a greater extent than

static billboards, but that the extent of this is not sufficient to result in a significant

change in behaviour such that there is then a consequential increase in the crash

rate.

18 I have been involved in assessing the crash records in the vicinity of two digital

billboards in Christchurch, where both resource consents had Conditions of

Consent requiring post-construction monitoring (as they were granted at a time

when the effects were not as well understood). These billboards were installed in

November 2015 and October 2015 respectively, but my review of the crash

1 For clarity, the papers listed in this Annexure are those identified By Dr J Wachtel entitled ’Compendium of Recent Research Studies on Distraction from Commercial Electronic Variable Message Signs (CEVMS)’.

1903537 | 4441653v01 page 4

records at each showed that there was no evidence that either had resulted in

adverse road safety effects arising.

19 In 2016, I was commissioned to carry out a comparative review of crash records

before and after digital billboards were installed at 11 sites within Auckland. My

detailed review is included as Annexure B to my evidence. In brief, I reviewed

the number of crashes in the five-year period prior to the installation of the digital

billboards, and compared this with the crash records after the installation. The

majority of the billboards were located at signalised intersections, and in many

cases, the billboards formed a background to the traffic lights themselves.

Notwithstanding this, my analysis showed a reduction in crash rates at the vast

majority of the sites following the installation of the digital billboards.

20 In passing, the aggregation of these Auckland sites meant that a total of 152

months of data had been evaluated, representing 55.4 million views of digital

billboards by passing drivers.

21 Importantly in respect of the current application, ten of the 14 sites that I reviewed

were at intersections and all of these were within the ‘cone of vision’ of the driver.

In five of these ten cases, the digital billboard formed a background to the traffic

signals heads themselves. Despite this, I was unable to find any evidence of an

increase in crashes.

22 My results are the same as a 2010 study carried out by Tantala Associates LLC

Consulting Engineers2 that examined the crash rates at ten locations in the USA

before and after digital billboards were installed. This concluded that there were

no statistically significant differences between the crash rates. Moreover, the

crash rates were seen to decline, as with my study.

23 I subsequently revised the study, and the results are shown in Annexure C.

24 In brief, in this second study I identified the five billboards that had been installed

for the longest period of time (in Auckland) and at signalised intersections. The

crash records at each location were then reviewed for the two years prior to the

installation of the digital billboard and for two years after their installation, and the

records were compared to see whether there had been an increase in crash

numbers.

25 At the same time, ‘control’ sites were selected in close proximity to each of the

digital billboards. One control site had a static billboard which had been in place

throughout the four years of crash records. One control site had no billboard

erected in this time period.

2 http://www.ohiooutdoor.org/wp-content/uploads/2011/01/TSA-Richmond-2010.pdf referenced on 30 July 2016

1903537 | 4441653v01 page 5

26 A critical part of this study is that it was done on a ‘double blind’ basis to eliminate

any bias. To do this, I contracted independent traffic engineers who were

provided with a list of the intersections which I had randomised (using a random

number generator to avoid any bias). They were not aware of which intersection

had a digital billboard, which had a static billboard and which had no billboard

(the first ‘blinding’).

27 I then asked them to identify all reported crashes using the NZTA Crash Analysis

System within 75m of the intersection where driver distraction may have been a

contributing cause. That is, driver distraction did not need to be specifically noted

as a factor in the record, but the type of crash recorded was of a type that could

have involved distraction. For example, a nose-to-tail crash in the queue

approaching the intersection or a driver failing to stop at a red signal were

identified as crashes which could have been due to distraction. Crashes where a

driver had a sudden medical event or vehicle failure were not assumed to be

caused by distraction.

28 The independent engineers then carried out a statistical test (a t-test at the 95%

level) to compare the results for any statistically significant differences. However,

they kindly agreed to undertake the assessment without knowing the purpose of

the study until all analyses had been complete (the second ‘blinding’). This

meant that there could be no subconscious bias in the selection or rejection of

crashes, or in the analysis itself.

29 The results of the analysis showed that there was no statistical difference

between the performance of the intersections with digital billboards when

compared to other intersections, and no statistical difference when the safety

records of the intersections before and after installation were compared.

30 On the basis of my literature review and my own research, I consider that I am

able to draw reliable conclusions as to whether digital billboards give rise to an

increase in traffic crashes. In my view, based on the evidence, they do not.

Submissions

31 I have reviewed the submissions that relate to transportation matters and have

discussed these below. For clarity, these are not listed in any particular order

and I have grouped together submissions which raise the same concern.

Submitter concern: The billboard may distract drivers

32 I discuss this in detail when addressing the Statements of Evidence of Dr Walton

and Mrs Rushmere.

1903537 | 4441653v01 page 6

Submitter concern: There will be adverse road safety effects due to the proximity

to the intersection


and Mrs Rushmere.

Submitter concern: Even a short distraction of the driver could create a crash


and Mrs Rushmere.

Submitter concern: The billboard should not block property accesses

35 I confirm that the billboard will be elevated and installed within a grassed verge,

meaning that no accesses will be blocked and sightlines will remain unaffected.

Submitter concern: The billboard should not include movement since it is

instinctive to look at movement

36 I confirm that a Condition of Consent will be sought such that no animation of the

images will be permitted.

Response to Evidence and Reports of Dr Walton

Reports of Dr Walton

37 I have read the three reports of Dr Walton, a consultant road safety engineer to

the Council. However, there are no material differences between the first and

second reports, and therefore I have not referred to the first report further.

38 I note that Dr Walton has taken a first principles approach when assessing the

billboard. Consequently, he has drawn his conclusions based on a limited amount

of research, plus some guidance documents (the veracity of which is not easily

independently checked). I have firstly reviewed the primary research which he

has used.

39 My review of the literature references by Dr Walton is attached as Annexure D to

this evidence. In short, I do not consider that many of these studies are

specifically relevant to this application, and in other cases, that the conclusions

are not supportive of digital billboards giving rise to road safety issues. Indeed, of

the seven papers referenced, one explicitly notes that billboards are not a

distractor for drivers, one notes that a statistical correlation could not be found

between road crashes and advertising signs, and one sets out that whether

billboards lead to crashes cannot be conclusively answered based on the

information available.

1903537 | 4441653v01 page 7

40 However, in his reports Dr Walton (rightly) highlights that not finding any evidence

of a relationship is not evidence that there is no relationship. That is, not being

able to show that there is a relationship between digital billboards and adverse

road safety outcomes does not mean that there is no relationship. Equally

though, it is not possible to ‘prove a negative’. In other words, if there is no

relationship then it is not possible to conclusively prove it.

41 With that in mind, the approach which I have taken in order to satisfy myself that

that there are no adverse road safety effects is twofold.

42 Firstly, I have sourced as many primary research papers as I have been able to

find. In total, I have around 40, and in my view, none of these show a clear

relationship between adverse road safety outcomes and digital billboards.

Annexure A of my evidence includes a summary of a selection of these.

43 There are a range of other papers (not reviewed in detail in this Statement of

Evidence) which support the view that there is no demonstrable relationship

between digital billboards and adverse road safety outcomes, for example,

Samsa3 and Perez et al4.

44 In assessing the effects of digital billboards, it is important to differentiate

between ‘distraction’ and ‘attraction’. A useful definition5 is that distraction occurs

“when a driver’s attention is diverted away from the driving task by an object,

activity, event, or other person, to such a degree that the driver no longer is

capable of performing the driving task in a safe manner”. That is, a driver may

have their attention attracted by something but unless this means that they are no

longer able to drive in a safe manner, they are not distracted.

45 In my view, it is not unreasonable to anticipate that if a driver is distracted, and

that they are no longer able to drive in a safe manner, this will then give rise to an

increased number or frequency of crashes. However, from my review of the

literature, and my own research, this does not appear to be the case.

3 Samsa, C. (2015) “Digital billboards ‘down under’: are they distracting to drivers and can industry and regulators work together for a successful road safety outcome?” Proceedings of the 2015 Australasian Road Safety Conference 14 – 16 October, Gold Coast, Australia

4 Perez, Bertola, Kennedy & Molino (2013) “Driver Visual Behaviour in the Presence of Commercial Electronic Variable Message Signs (CEVMS)”, US Department of Transportation Federal Highways Administration.

5 Young, K.L. and M.A. Regan, Driver distraction: A review of the literature, in Distracted driving. , I.J. Faulks, et al., Editors. 2007, Australasian College of Road Safety: Sydney, NSW. p. 379-405.

1903537 | 4441653v01 page 8

46 By way of example, one study which is sometimes cited is that of Sisiopiku, VP,

Islam, M, Haleem, K, Alluri, P. & Gan, A. (2014)6. This compares the crash

records upstream and downstream of digital billboards on high speed roads in the

USA. When the data is aggregated, it purports to show an overall increase in the

number of crashes prior to the billboard (where it can be seen by the driver)

compared to downstream (where it is not visible). However, at 50% of the sites

examined, the crash records showed that the road was safer prior to the billboard

than after, with the outcomes reversed at the other 50% of the sites. In my view,

this is not the consistent pattern that would be expected from distracted drivers.

47 A second issue which arises when considering the research into digital billboards

is the extent to which animation is permitted (or was present when the research

was undertaken). In some cases, the papers are silent on this matter and it its

therefore not possible to easily apply the results of the research to scenarios

where there is no animation. However, in my view there is sufficient research

available to show that animation does capture driver attention to an increased

extent such that an elevated road safety risk arises. The applicant has agreed

that animation will not be permitted on the proposed billboard.

48 Within the New Zealand context, roading engineers often refer to the Austroads

organisation and many of its guides form part of the suite of design parameters

for NZ roads. There is one Austroads document which is helpful in this regard,

which is a research report concerning all types of roadside advertising from static

billboards to those that interact with a driver and those which are projections of

large images onto buildings (Austroads Research Report AP-R420-13, “Impact of

Roadside Advertising on Road Safety” Section 3).

49 The guide adopts a cautious approach in drawing any conclusions:

“There is compelling evidence that distraction is a major contributor to

crashes. However, studies providing direct evidence that roadside

advertising plays a significant role in these distraction based crashes

are currently not available. The studies that have been conducted show

convincingly that roadside advertising is distracting and that it may lead to

poorer vehicle control. However, the evidence is presently only suggestive

of, although clearly consistent with, the notion that this in turn results in

crashes.

6 Sisiopiku, VP, Islam, M, Haleem, K, Alluri, P. & Gan, A. (2014). Investigation of the Potential Relationship between Crash Occurrence and the Presence of Digital Advertising Billboards in Alabama and Florida. Proceedings of the Transportation Research Board (TRB) 94th Annual Meeting.

1903537 | 4441653v01 page 9

It is also worth noting, on the basis of Klauer et al.’s (2006) results, that

while looking at an external object increased the crash risk by nearly four

times, less than 1% of all crashes and near crashes were from this source

of distraction. A substantial proportion of these external objects would not

have been advertising signs. Thus, while it is not possible to tell from the

reported results, it is reasonable to conclude that far less than 1% of all

crashes and near crashes involved distraction from roadside

advertising.

While the Klauer et al. (2006) study may not be representative of all driving

events, it does suggest that the contribution of roadside advertising

to crashes is likely to be relatively minor.”

(Austroads Research Report AP-R420-13, “Impact of Roadside Advertising

on Road Safety” Section 5.2) (My emphases)

50 Secondly, I have undertaken my own research into the safety effects of

billboards. These are discussed in detail previously in this Statement of Evidence,

but in brief, I have been unable to find any relationship between adverse road

safety outcomes and the presence of digital billboards.

Statement of Evidence of Dr Walton

51 Within his evidence, Dr Walton comments on the use of animation (his paragraph

3.3). I agree with his views in respect of the likely distraction of drivers, and I

therefore agree that animation should not be permitted.

52 In respect of images changing from one to another by way of a 0.5 second

dissolve (Walton paragraph 3.4), the research shows that an instantaneous

change creates a ‘flash’ effect due to the change in luminosity which attracts the

drivers attention involuntarily7. Moreover, Dr Walton’s assertion of drivers waiting

for the next image appears to be based on sequential advertisements rather than

the first and subsequent advertisements being unrelated. However, it is a

straightforward matter to specify the transition time within a Condition of Consent,

if an instantaneous change is desired.

53 I agree that the brightness of the sign is straightforward to specify and address

and amend if necessary (Walton paragraph 3.7).

54 I do not agree that advertisements should avoid telephone numbers, addresses,

email addresses and websites due to a concern that “drivers may attempt to write

these down”. Such an outcome is in my view fanciful, and from a practical

7 Queensland Department of Transport and Main Roads. “Roadside Advertising Guide”, 2013

1903537 | 4441653v01 page 10

perspective, a driver will have already travelled past the billboard by the time that

they have somehow found a pen and paper in their vehicle.

55 In Section 4 of his evidence, Dr Walton discusses a number of aspects of the

Officer’s Reports. In his paragraph 4.3, his states that the road safety risk to

drivers “will accumulate”. This is potentially misleading because statistically, the

potential for a crash does not vary according to the number of road users that

have passed through a particular point. If a particular location has, say, one crash

for every million road users, this does not mean that the millionth road user will

inevitably be involved in a crash. The risk to the first road user is exactly the

same as the risk to the millionth road user – the risk per road user remains

constant.

56 Regardless though, Dr Walton’s assertion is based on a premise that the

billboard will result in adverse road safety effects and in my view, for the reasons

set out above, it will not.

57 With regard to paragraph 4.6 of Dr Walton’s Statement of Evidence, and the

propensity of people to look at billboards, I highlight that video / animation is not

now proposed. It is also worth noting that simply glancing at a billboard is not

indicative of the potential for a crash (rather it is the duration of the gaze) and the

study referenced by Dr Walton did not compare the number of glances at

billboards with the number of glances at other roadside objects or other signage.

The Dukic study8 is helpful in regard to the latter, showing that drivers also look at

roadsigns on a regular basis.

58 Section 5 of Dr Walton’s evidence summarises his main concerns, which appear

to be that the proposed location of the billboard is adjacent to a ‘decision point’,

where a road user’s cognitive load is increased such that competition for that

required resource from an advertising sign would serve as a potential distraction.

59 Helpfully, there are two studies that have assessed the issue of driver attention in

complex scenarios. One of these9 concluded that “generally, billboard-related

distraction appeared to be minor and regulated by drivers as the demands of the

driving task changed” although it also noted that there was scope for further

research.

8 Dukic, T., Ahlstrom, C., Patten, C., Kettwich, C., & Kircher, K. (2013). Effects of electronic billboards on driver distraction. Traffic Injury Prevention, 14(5), 469–476.

9 Decker, JS et al (2015). The Impact of Billboards on Driver Visual Behavior: A Systematic Literature Review, Traffic Injury Prevention Vol 16(3), 234-239

1903537 | 4441653v01 page 11

60 A more recent paper10 confirms this finding based on a study of driver behaviour

on an urban test route, concluding that “the structure and content of drivers'

situation awareness was not appreciably affected by the billboards in any of the

road environments examined. Drivers focused their attention on the billboards

when driving demand was low, such as when driving on the freeway with light to

moderate traffic, in lower speed zones, or when stationary. However, when

drivers were required to perform a manoeuvre or driving demands increased,

drivers directed less attention to the billboards and focussed their awareness on

the immediate driving task”.

61 In my view then, the literature shows that when drivers have limited cognitive

capacity due to complex environments, they focus on the issues that are the most

important (ie driving). Put another way, if this is a complex location, then drivers

will pay less attention to the billboard.

62 In respect of the proximity to a decision point, Dr Walton appears to rely on the

South Australian guidelines (quoted within his third report). However, further

reading of this guide shows that the so-called “Device Restriction Area’ is simply

an area “whereby the installation of advertising signs is required to abide by a

‘stricter’ set of rules in order to assist in alleviating potential additional safety

concerns within these areas”11 That is, the guidelines do not prohibit billboards

within the restriction area. Further, the list of Device Restriction Areas in the guide

does not include diverging lanes of the type at the application site, only diverging

lanes at freeways / expressways (which is not the case here). There is however

a requirement for billboards to be located 60m from the stop-line at an

intersection, and I confirm that this is achieved in this case.

63 Taking all of the matters above into account, I do not agree with Dr Walton’s

views regarding the likely adverse road safety effects of the proposed billboard.

Rather, my assessment of the research leads me to the conclusion that any

adverse effects will be less then minor, because there is no evidence that digital

billboards give rise to adverse safety effects (provided that they are controlled

such that they emulate a static billboard).

Response to Statement of Evidence of Mrs Rushmere

64 In her paragraph 5.6, Mrs Rushmere sets out her concerns with the billboard’s

location.

10 Young, KL et al (2017). Investigating the Impact of Static Roadside Advertising on Drivers’ Situation Awareness, Applied Ergonomics, Vol 60, 136-145

11 Department of Planning, Transport and, and Infrastructure. “Advertising Signs Assessment Guidelines for Road Safety,” 2014, 31. https://www.dpti.sa.gov.au/__data/assets/pdf_file/0019/145333/DPTI-Advertising-Signs- Assessment-Guidelines.pdf

1903537 | 4441653v01 page 12

65 In her third bullet point, Mrs Rushmere sets out that “evidence” shows that

crashes have already occurred at the diverge lanes on Kāpiti Road with

distracted drivers being a contributing factor. Since no further information is

provided by Mrs Rushmere, I have reviewed the crash records from the NZTA

Crash Analysis System.

66 In total, there have been seven crashes recorded between the stop-line of the

Kāpiti Road / Te Roto Drive intersection and 25m north of Arko Place. Four of

these involved northbound vehicles, and therefore the drivers would not have

been facing the billboard (and therefore it could not have influenced any crashes).

One crash occurred within a car park. Only two crashes involved southbound

vehicles on Kāpiti Road:

(a) A southbound driver in the centre lane moved into the left lane and struck

another vehicle that was already in the left lane; and

(b) A southbound driver travelled into the left lane thinking the driver ahead

was travelling into the right lane, except the driver ahead went into left

lane. It is not clear whether the vehicles collided or not.

67 Neither of these crashes resulted in any injuries. In my view then, the crash

record does not highlight any particular road safety deficiencies in this location. I

therefore do not agree that there is evidence of a poor crash record.

68 Mrs Rushmere is concerned that cyclists and pedestrians may be distracted by

the billboard and collide with one another (or with a car). I have been unable to

find any research that suggests pedestrians and cyclists are at greater risk of

crashes as a result of digital billboards. Indeed, in cities such as Melbourne,

digital billboards (with full animation) are located in locations where cyclists and

pedestrians share the road with cars.

69 Mrs Rushmere sets out that in her view, drivers have the potential to view the

billboard at a distance of 450m away. It is extremely unlikely in my view that at

this distance drivers would be able to perceive any wording on the sign and rather

it will simply appear as a lit static billboard. Further, the billboard will simply be

one of many signs and will in effect simply blend into the prevailing environment.

70 In her final bullet point in paragraph 5.6, Mrs Rushmere sets out a brief review of

three documents:

(a) The NZTA Traffic Control Devices Manual (Part 13: Advertising Signs) is

not mandatory but is a guideline. I acknowledge that this sets out that

billboards should not be within 100m of an intersection, but in this case the

distance provided is 84m (according to Dr Walton) and therefore the

1903537 | 4441653v01 page 13

shortfall is small. Further, this document also sets out that “there are many

advertisements close to intersections … apparently causing no problems.”

(b) The Advertising Signs Assessment Guidelines for Road Safety (South

Australia) (August 2014) is described by Mrs Rushmere as “prohibiting”

advertising signs at critical points in the network. This is not the case,

rather, the guidelines set out that stricter rules are expected in those

locations. As I noted previously, diverge lanes that are not on

expressways/freeways are not addressed by this guideline.

(c) The Australia Road Research Board ‘Evidence based guidelines for safe

use of digital billboard installation’ (2013) appears to be a conference

paper. The paper sets out that advertising devices (which include all types

of advertising) “should” not be placed on the approach to an intersection.

However, there is no discussion about what separation distance is suitable

(since in theory every location is on the approach to some intersection) and

further, in this location there is already advertising which is closer to the

intersection than the proposed billboard and which has not presented any

road safety issues to date.

The author of this conference paper previously authored the Austroads

research report to which I referred in paragraph 48 above.

71 It is not clear how many other guidance documents or research papers Mrs

Rushmere has reviewed, or whether her views are based only on these three

documents.

72 Mrs Rushmere queries how controls can be applied which ensure that the

billboard will not appear to be a road sign. This is an important issue, since

guidance is unequivocal that images should not resemble any road sign, and I

discuss this subsequently.

73 Taking all of the matters above into account, I do not agree with Mrs Rushmere

regarding the likely adverse road safety effects of the proposed billboard. Rather,

I consider that any adverse effects will be less then minor, because there is no

evidence that digital billboards give rise to adverse safety effects.

Conditions of Consent

74 I have reviewed the Conditions of Consent set out within Appendix 5 of the s42A

report.

75 Condition 8 notes that the dwell time is to be 45 seconds, as set out in Appendix

5 of the Advertising Signs Assessment Guidelines for Road Safety Guidelines

(Government of South Australian Department of Planning, Transport and

1903537 | 4441653v01 page 14

Infrastructure, August 2014). Having reviewed this document, and others, I have

been unable to identify any source from which this figure is derived. There is no

information available which demonstrates why 45 seconds results in reduced

adverse road safety effects compared to shorter dwell times.

76 Current practice in New Zealand is to use an 8 second dwell time (although,

NZTA has recently moved towards standardising this at 10 seconds), with over

100 billboards consented at this duration. This duration of display for the image

has not presented any road safety issues. In my view the duration of display

should be set at 10 seconds.

77 As set out above, in my view the research shows that an instantaneous change in

the image has the potential to distract drivers involuntarily and so I prefer a 0.5

second dissolve (Condition 9). That said, it is a straightforward matter to change

the way that the image is displayed, if the Council prefers a more sudden change.

78 I agree with Condition 10. Restriction of flashing, movement and the like is an

important factor in ensuring that the billboard operates safely.

79 I also agree with Condition 11. However, my suggestion is to amend the wording

to be more precise: “Images displayed shall not incorporate the predominant use

of graphics, colours or shapes that could cause confusion or conflict with any

traffic control device, nor invite or direct a driver to undertake an action”.

80 Condition 12 seeks to limit the number of words displayed at any time and

appears to be based on an assumption that drivers are expected to read every

word on the image. This is the case with road signs and the like, but in regard to

commercial advertising, parts of the text are not intended to be read. Rather, it

either combines with the graphic to form an image and it is the image that the

drivers observe, or the text is deliberately small such that it cannot be read (such

as the ‘small print’ associated with bank loans) but has to be displayed by law.

81 The Samsa study I referred to above showed that the average glance time when

drivers choose to look at a sign or billboard is 0.75 seconds. This is sufficient only

for drivers to see the overall image, rather than to read all text which may be

present in the image. From a driver’s perspective, any text that is not readily

legible, and which obviously forms a part of an image, will not be read. It is not

the case that drivers will slow down or gaze for longer to try to read more words.

82 By way of example, I attach a still image of a video taken recently on Kāpiti Road.

As a driver travelling along the road, the short period of time available means that

all that can be read is that there is a site for lease. The names and contact details

of the realtors are simply disregarded.

1903537 | 4441653v01 page 15

83 In respect of this Condition, it is also unnecessary in my view to prohibit

telephone numbers, addresses, email addresses and websites from being

displayed in case drivers try to write them down. This simply does not happen.

84 I therefore recommend that Condition 12 is deleted.

85 Condition 13 sets out that any lettering, colour combinations and or symbols will

be clearly legible and will contrast with the background. In my view this is

unnecessary to specify since advertising where the message is not legible will be

ineffective – put another way, it is in the advertiser’s interest to ensure that the

message is legible. However, it is a matter of debate as to what lettering is

considered to be clearly legible and whether in fact this is relevant. By way of

example, the Coca-Cola font is arguably not easily legible, but it is not seen as

wording but as part of a logo. Other examples include the ‘Z’ in Z Energy which is

heavily stylised, and ANZ which has each letter split into two horizontally.

86 This Condition refers to the NZTA Manual of Traffic Signs and Markings Part 1

(August 2010) and part 3 (January 2011). I am unclear what this is intended to

achieve since the Manual of Traffic Signs and Markings sets out a raft of

signs/markings for roads which are irrelevant to a commercial billboard. If it is

intended to limit the type of font used to the same as used on standard road

signs, then this is impractical. Moreover, even the road safety message signs that

NZTA erects alongside state highways do not use this font.

87 I therefore recommend that Condition 13 is deleted.

88 Condition 14 ensures that there is to be no objectionable or offensive material

displayed on the billboard. This provision is already a requirement of the

Advertising Standards Code, and it is therefore unnecessary to include it as a

1903537 | 4441653v01 page 16

Condition. Similarly, it is my understanding that the Code of Ethics is now

included within the overarching Code, and therefore specific reference to it is not

needed.

89 This Condition also requires that the billboard “shall comply with the regulations

of the Traffic Control Devices Manual-Part 3: Advertising, published by NZTA

(2011)”. For clarity, this is a guideline and it is not possible to comply with a

guideline where much of the content represents recommendations or matters that

are optional. That said, this document does include some mandatory

requirements, and in my experience it is increasingly the case that such

measures are included as Advice Notes.

90 I therefore recommend that Condition 14 is deleted but is replaced with the

following Advice Notes:

The consent holder is advised that reflective materials should not be used

for the digital display units, should the digital display unit or any parts be

replaced in the future, and would potentially be contrary to relevant

legislation.

The consent holder is advised of the legal framework that is set out in

clauses 3.1 and 3.2 of the NZTA’s Traffic Control Devices Manual Part 3

Advertising Signs 2011.

91 I support Condition 15. In the event of a malfunction, it is possible for billboards to

fail in a flashing state or with only part of a message displayed, both of which

could cause distraction to drivers, and hence this Condition is required to ensure

safe operation.

92 I have been involved in a number of applications where the Council has

requested a monitoring condition is put in place. These are increasingly

uncommon as Councils around the country increase their understanding of the

operation of billboards. However, in my view Condition 16 is unworkable in its

current form because:

(a) The first monitoring at six months after installation is unlikely to provide any

benefit because it is too short a period for any change in the crash record

to be evident;

(b) Data from ACC is not available to third parties due to personal privacy

issues;

(c) Council complaints and requests do not represent a road safety issue, but

rather they represent a perception of a safety issue; and

1903537 | 4441653v01 page 17

(d) Observational studies are of extremely limited value unless they occur over

an extended period of time. Further, the methodology (monitoring the

speed, headway and lane position of 10,000 vehicles per day) is simply

impractical.

93 It appears that the purpose of Condition 16 is to understand the effects of the

billboard. In my view this type of onerous methodology is not needed because

there is no evidence that a road safety issue will arise. Further, even if the study

showed that there was a small change in (say) vehicle headway, it would not be

possible to definitively attribute this to the effects of the billboard as the change

may be due to an external source (for example, slightly higher traffic volumes on

Kāpiti Road as a result of natural daily variation in the flow). It would also not be

possible to translate changes in speed, headway and/or lane position to an

adverse road safety effect unless the effects were so significant that they led to

an actual crash (in which case it would be evident in the NZTA Crash Analysis

System). Bluntly, many factors such as speed and headways change on a daily /

hourly basis without adverse safety effects arising.

94 If there is to be a monitoring condition, in my view it has to be practical and

provide outputs that are meaningful. Consequently, my suggestion would be the

following:

Prior to the operation of the billboard, the consent holder shall appoint a

suitably qualified traffic engineer to establish the number, type and location

of injury crashes recorded in the New Zealand Transport Agency Crash

Analysis System that have been reported in the previous five years on

Kāpiti Road from the stop-line at the intersection with Te Roto Drive, to a

location 250m to the north (ie just south of the roundabout). The suitably

qualified traffic engineer shall prepare a report detailing those findings. This

‘baseline report’ shall be provided to the District Council within five working

days of it being finalised

Each year for a period of three years from the date that the billboard

commences operation, and within one month of the anniversary of that

date, the consent holder will provide to the District Council a review of the

injury crashes recorded in the New Zealand Transport Agency Crash

Analysis System that have been reported within the section of Kāpiti Road

described above over the preceding year, prepared by a suitably qualified

traffic engineer. Where the number of injury crashes exceeds the range

established in the baseline report, the following will be identified:

a) Whether there have been any changes in traffic management on

Kāpiti Road and at the Kāpiti Road / Te Roto Drive intersection;

1903537 | 4441653v01 page 18

b) Whether one or more of the road users could potentially have seen

the image on the billboard at the location of the crash;

c) Whether the crash record included an “attention diverted” cause

code;

d) Whether the written description on the police record indicates that

distraction by something outside the vehicle was a factor;

e) Whether there has been an increase in this type of crash (of all

injury severities) compared to the previous five years; and

f) Whether, taking all of these factors into account, the billboard is

likely (in the opinion of a suitably qualified traffic engineer) to have

been a contributing factor and if so, to what degree.

Where, in the opinion of a suitably qualified traffic engineer, the billboard is

likely to have been a contributing factor then the traffic engineer will make

appropriate recommendations for any mitigation measures. These may

include, but not be limited to:

a) Reducing the brightness (luminance) of the billboard;

b) Increasing the duration of the images displayed at selected or all

times of the day;

c) Changing the transition time for the images; and/or

d) Revising the controls on image content

Conclusion

95 Having evaluated and assessed the transportation aspects of the proposal,

particularly the matter of road safety, I consider that the adverse transportation-

related effects arising from the digital billboard will be negligible. My view is based

on a comprehensive review of the primary research literature, coupled with my

own research. In short, the proposed billboard will be controlled through

Conditions of Consent which mean that many of the factors which could lead to

driver distraction and crashes (such as animation, sequential images, bright

lighting in darkness, etc) will be eliminated. The billboard will simply appear to be

a static billboard, albeit one where the images changes.

96 I am of a different opinion to the Council transportation planners/engineers, Dr

Walton and Mrs Rushmere, who consider that adverse safety effects are likely to

arise if the application was to be granted. In my view though, such conclusions

are not supported by the evidence.

«MatterNo» | 4441653v01 page 19

97 I have reviewed the proposed Conditions of Consent. In some cases I support the

proposed provisions as they are relevant for controlling the operation of the

billboard in order to ensure that adverse road safety effects do not arise. In other

cases though, the Conditions of Consent are unnecessary, unworkable or

replicate matters which the consent-holder is already obliged to achieve.

98 Overall, and subject to my comments above regarding Conditions of Consent, I

consider that there are no reasons why the application for the installation and

operation of a digital billboard at 129 Kāpiti Road could not be approved.

Andy Carr

20 June 2019


Annexure A

Review of Selected Research Literature

Backer-Grøndahl, A., & Sagberg, F. (2009). “Relative crash involvement risk associated with different sources of driver distraction.” Presented at the First international Conference on Driver Distraction and Inattention. Gothenburg, Sweden: Chalmers University.

Summary: A sample of 4,307 drivers who had been involved in a crash in the previous 12 months filled in a web-based questionnaire about distractions during the crash. For each of the potential 13 distraction factors presented, the drivers indicated whether or not they were distracted by that specific factor at the time of the crash. ‘Distracted by billboard’ was one factor of the 13.

“Even though the results from this study indicate that looking at billboards and searching for addresses/street names are the distractions associated with highest accident risk, it is also important to look at the prevalence of the risk factor. These two factors were reported to have been distracting only 0.3 and 0.6 percent of drivers (i.e., in the whole sample) respectively. This means that, as measured by the rate to which billboards distract drivers, this is not a large risk factor from a population perspective.

When considering the prevalence of the risk factors in addition to the relative accident involvement, talking with passenger(s) and attending to children in the back seat are the distraction factors that perhaps are most likely to make the largest contributions to the number of crashes”

Belyusar, D., Reimer, B., Mehler, B., & Coughlin, JF. (2016). “A field study on the effects of digital billboards on glance behavior during highway driving.” Accident Analysis and Prevention, 88, 88-96.

Summary: The study reports the glance and driving behaviour of 123 drivers who were exposed to two digital billboards on a segment of an eight-lane (four lanes in each direction) highway subject to a100km/h speed limit. Other than the billboards, the highway was largely free from extraneous signage.

“The results presented in this report require further confirmation in different environments to establish the generalizability of findings. However, combined with previous literature, they clearly suggest that digital billboards alter driver attention. The degree to which this diversion of attention impacts safety is not clear.”

Bendak, S., & Al-Saleh, K. (2010). “The role of advertising signs in distracting drivers.” International Journal of Industrial Ergonomics, 40, 233-236.

Summary: Twelve volunteers (all male) were asked to use a car simulator to drive around a 9.3km route and five metrics were observed - their tendency to drift from their own traffic lane, recklessly crossing dangerous intersections, not signalling, speeding and tailgating. In the presence of roadside advertising there were statistically significant differences with drifting from their own traffic lane and recklessly crossing dangerous intersections but not the other three metrics.

“In this current study, roadside advertising signs refer specifically to electric signs (which are illuminated by internal lights), animated signs (which refer to any sign that moves or gives the effect of a moving display), banners (which are portable signs usually made of fabric), shop fronts, billboards (that consist of a number of standard-sized poster panels) and changing message signs (which are animated signs consisting of messages changing in sequence). These signs can be located within the road boundaries, on private property near the road or mounted on vehicles.”

Comment: It is not clear in the study whether the roadside advertising used in the simulator was solely limited to billboards or included full motion and/or other types.

Chan, E., Pradhan, AK, Knodler, MA, Jr., Pollatsek, A. & Fisher, DL. (2008). “Empirical Evaluation on a Driving Simulator of the Effect of Distractions Inside and Outside the Vehicle on Drivers’ Eye Behaviors,” Washington, DC: 87th Annual Meeting of the Transportation Research Board of the National Academies.

Paper could not be sourced

Chattington, M., Reed, N., Basacik, D., Flint, A., & Parkes, A. (2009). “Investigating Driver Distraction: The Effects of Video and Static Advertising.” Report No. RPN256. United Kingdom: Transport Research Laboratory

Summary: This study compared video billboards (defined as billboards where animation or full motion is shown, and akin to a television) to static billboards (where there is no movement), using a sample of 48 drivers and a driving simulator.

“This study set out to investigate the relative level of driver distraction caused by a range of billboard advertising configurations with a particular focus on the effect of video adverts compared to static adverts”

“While it is clear there are some effects of position and duration of exposure, the main findings is that video adverts provide a greater distraction than that currently caused by drivers approaching equivalent static adverts”

Comment: It is not considered that this study is particularly relevant since it addresses only digital billboards displaying full motion.

Divekar, G., Pradhan, AK, Pollatsek, A., & Fisher, DL. (2013). “External Distractions”: Evaluations of their effect on younger novice and experienced drivers’ behavior and vehicle control.” Transportation Research Record, Journal of the Transportation Research Board No. 2321. Washington, DC: Transportation Research Board of the National Academies.

Summary: This paper discusses the potential for long glances away from the forward roadway create safety problems. The methodology used was to compare 24 novice drivers and 24 experienced drivers as they drove a car simulator and were engaged in a task of counting the number of times that a specific letter appeared within a 5 x 5 grid of letters presented to them in the form of a (simulated) roadside billboard. Metrics such as speed, lane deviation, braking and acceleration were measured. There were 11 such grids on a 4.3km long route (one grid every 390m).

“The major finding was that the long glances of both experienced and novice drivers came at the cost of identifying potential hidden hazards and seeing exposed moving threats.”

Comment: It is not considered that this study is particularly relevant to digital billboards since it relates to distractions of all types outside the vehicle.

Dukic, T., Ahlstrom, C., Patten, C., Kettwich, C., & Kircher, K. (2012). “Effects of Electronic Billboards on Driver Distraction.” Journal of Traffic Injury Prevention, 14, 469-476.

Summary: The study used a sample of 41 drivers in Sweden to drive a route passing four electronic billboards during day and night conditions. A driver was considered visually distracted when looking at a billboard continuously for more than 2 s or if the driver looked away from the road for a high percentage of time. The amount of time that they looked at the billboards was compared to the amount of time spent looking at standard road signs.

“To conclude, electronic billboards appear to have an effect on gaze behavior because they attract more and longer glances than standard road signs. Whether they attract too much attention and constitute a bona fide traffic safety hazard cannot be answered conclusively based on the present data”

Edquist, J., Horberry, T., Hosking, S. & Johnston, I. (2011). “Advertising billboards impair change detection in road scenes.” Paper presented at the 2011 Australasian Road Safety Research, Education & Policing Conference.

Summary: The study used a sample of 45 drivers to view photographs of road scenes on a screen, with an image of a scene displayed for 0.4 seconds, before a second image was shown for the same period of time. The first image was then displayed again, followed by the second image and so on. The was essentially the same, except that one of the pair had been modified such that a car, a road sign, or some other item was missing or its size had been changed. Participants were required to identify the nature of the change in the fastest time possible. However, some of the pairs of images included static billboards and some did not.

“When both built and designed clutter were high, adding billboards did not have a significant effect on time to detect change. When built clutter was high but designed clutter was low or vice versa, drivers took longer to detect changes in scenes with billboards than in scenes without billboards. When both built and designed clutter were low, drivers were faster to detect changes in scenes with billboards”

(Note: “designed clutter” is defined as built clutter is the objects that road authorities use to communicate with the driver, such as road markings, traffic signs and signals. “Built clutter” is buildings and other infrastructure, shop signage etc that make the scene visually complex)

“The present study is limited in that it did not include a driving task, merely a surrogate measure for visual subtasks required during driving.”

Gitelman, V., Zaidel, D., & Doveh, E. (2012). “Influence of Billboards on Driving Behavior and Road Safety,” Presented at: Fifth International Conference on Traffic and Transportation Psychology. Groningen, The Netherlands: University of Groningen.

Summary: The paper presents an analysis of the impact of advertising billboards adjacent to the Ayalon Highway (Israel) on the occurrence of crashes on that highway. Two periods are compared: "before" – when the billboards were present along the roadside (years 2006-2007) and "after" - when the billboards were covered (2008). A literature survey is also included.

“The literature survey shows that both early and recent studies found a negative impact of advertising billboards on safety. However, a critical analysis of the studies reveals that many studies were not methodologically adequate. Recent studies were more rigorous, and while the findings were also in the same direction, the results were often not statistically significant.”

“Laboratory experiments, including simulator studies, have shown deteriorating driving performance in the presence of advertising billboards and messages, especially dynamic advertising media. However, the findings of field studies do not provide consistent evidence for the negative effects of billboards on driver behaviour”

“Due to reservations regarding the data, the uniqueness of the Ayalon Highway and the Treatment characteristics, it is recommended not to attach undue weight to the (relative large) derived statistical value for the percentage reduction in accidents following the removal / cover of advertising billboards. However, the downward trend in accidents in the “after“ period was robust and consistent, in all examinations, particularly for injury crashes. Therefore we can conclude that under Israeli road conditions, there is empirical evidence of a link between the removal of advertising signs and the improvement of road safety on an urban / suburban highway.”

Gitelman, V., Zaidel, D., Doveh, E., & Zilberstein, R. (2014) “The Impact of Billboards on Road Accidents on Ayalon Highway Three Periods Comparison – Billboards Present, Removed, and Returned.” Report to the Israeli National Road Authority.

Summary: As per the above, except that an additional data set is included – of the crash records over 3.5 years when the billboards were returned

Ayalon Highway “is used as a local, metropolitan, and national level traffic route. • 21 interchanges (rather dense). 750, 000 vehicles travel it daily. Trains run in parallel to the highway”

“Ayalon Highway carries high volume of traffic and has high density of interchanges generating frequent need for lane changes and speed adjustment”

“Advertising industry, Lobbyists and Politicians argued for or against roadside advertising. No side in the debate could present compelling accident data”

Hawkins, HG, Jr., Kuo, PF, & Lord, D. (2014). “Statistical Analysis of the Traffic Safety Impacts of On-Premise Digital Signs.” Paper No: 14-2772. Presented at the 93rd Annual Meeting of the Transportation Research Board.

Summary: A total of 135 on-premise digital signs were identified, and the number of crashes which occurred before and after the sign was installed were compared. Three to four years of ‘before’ data plus three to four years of ‘after’ data were used. Control sites were used to account for changes in crashes unrelated to the presence of the sign.

“The results show that there was no statistically significant change in crash frequency associated with the installation of on‐premise digital signs. Thus, there seems to be no evidence the installation of on-premise signs at these locations led to an automatic increase in the number of crashes”.

Herrstedt, L., Greibe, P. & Andersson, P. (2013). “Roadside Advertising Affects Driver Attention and Road Safety.” Proceedings

of the 3rd International Conference on Driver Distraction and Inattention, Gothenburg, Sweden.

Summary: The purpose was to study whether static roadside advertising in rural areas captures and keeps drivers’ attention to the extent that it affects driver behaviour and thereby traffic safety, using a sample of 32 drivers on high speed roads.

“The roadside advertising signs were selected amongst the most striking conventional rural roadside advertising signs”

Comment: It is not considered that this study is particularly relevant as it solely addresses billboards on high speed rural roads, where other research shows drivers tend to be attracted to the billboard due to a lack of other stimuli.

Horberry, T., Regan, MA, & Edquist, J. (2009). Driver Distraction from Roadside Advertising: The clash of road safety evidence, highway authority guidelines, and commercial advertising pressure.

Summary: This is a literature review and does not appear to have been peer reviewed or published in a journal.

“There has been little research into how and to what extent roadside advertising can cause driver distraction”

“The whole area is difficult to study due to differences in billboard types, drivers, roads, traffic etc”

“There is still a lack of comprehensive research evidence upon which to form guidelines or standards about how much distraction from outside of the vehicle is ‘safe’”

Milloy, SL and Caird, JK. (2011). “External Driver Distractions: The Effects of Video Billboards and Wind Farms on Driver Performance.” Published in: Handbook of Driving Simulation for Engineering, Medicine and Psychology. Edited by: D.L. Fisher, M. Rizzo, J.K. Caird, & J.D. Lee. Boca Raton: CRC Press.

Paper could not be sourced

Perez, WA., Bertola, MA, Kennedy, JF, & Molino, JA. (2012). “Driver Visual Behavior in the Presence of Commercial Electronic Variable Message Signs (CEVMS).” Unnumbered Report, Federal Highway Administration, Washington, DC.

Summary: This study used an instrumented vehicle with an eye tracking system to measure where drivers were looking when driving past digital and standard billboards. These billboards did not contain dynamic video or other dynamic elements, but changed content approximately every 8 to 10 seconds.

“In the present study, the presence of (digital billboards) did not appear to be related to a decrease in looking toward the road ahead.”

“The results did not provide evidence indicating that (digital billboards), as deployed and tested in the two selected cities, were associated with unacceptably long glances away from the road”

“When comparing the probability of a gaze at a (digital billboard) versus a standard billboard, the drivers in this study were generally more likely to gaze at (digital billboards) than at standard billboards.”

“The present data suggest that the drivers in this study directed the majority of their visual attention to areas of the roadway that were relevant to the task at hand (e.g., the driving task). Furthermore, it is possible, and likely, that in the time that the drivers looked away from the forward roadway, they may have elected to glance at other objects in the surrounding environment (in the absence of billboards) that were not relevant to the driving task. When billboards were present, the drivers in this study sometimes looked at them, but not such that overall attention to the forward roadway decreased”

Roberts, P., Boddington, K., & Rodwell, L. (2013). Impact of Roadside Advertising on Road Safety. Austroads Road Research Report: Publication No. AP-R420-13. City: Australia, ARRB Group.

Summary: This report is a review of other research.

“There is compelling evidence that distraction is a major contributor to crashes. However, studies providing direct evidence that roadside advertising plays a significant role in these distraction based crashes are currently not available. The studies that have been conducted show convincingly that roadside advertising is distracting and that it may lead to poorer vehicle control. However, the evidence is presently only suggestive of, although clearly consistent with, the notion that this in turn results in crashes.

It is also worth noting, on the basis of Klauer et al.’s (2006) results, that while looking at an external object increased the crash risk by nearly four times, less than 1% of all crashes and near crashes were from this source of distraction. A substantial proportion of these external objects would not have been advertising signs. Thus, while it is not possible to tell from the reported results, it is reasonable to conclude that far less than 1% of all crashes and near crashes involved distraction from roadside advertising.

While the Klauer et al. (2006) study may not be representative of all driving events, it does suggest that the contribution of roadside advertising to crashes is likely to be relatively minor”

Samsa, C., & Phillips, T. (2015). Digital Billboards ‘Down Under’. Are they Distracting to Drivers and can Industry and Regulators Work Together for a Successful Road Safety Outcome? Paper Presented at the 4th International Conference on Driver Distraction and Inattention, Sydney, Australia.

Summary: A total of 29 participants drove an instrumented vehicle along a 14.6 km route in Brisbane, Queensland passing a number of advertising signs, including digital and static billboards and on-premise signage. Number of fixations and dwell times towards advertising signs were measured, along with lateral deviation and vehicle headway.

“The findings show that digital billboards do not draw drivers’ attention away from the road for dangerously long periods of time compared to the other signage types, and drivers maintained a safe average vehicle headway in the presence of these signs”

Schieber, F., Limrick, K. McCall, R, & Beck, A. (2014). Evaluation of the Visual Demands of Digital Billboards Using a Hybrid Driving Simulator. Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting, 2214-2218.

Summary: The study used a specially-built simulator and a sample of 18 participants to assess driving performance in the presence of a simulated digital billboard at speeds of 40km/h and 80km/h. The participants were required to read aloud the message on the billboard which comprised of 4, 8 or 12 words selected at random, with the billboard simulating a 3.3m wide billboard.

“Video-based driving simulators are not well suited for studying a driver’s ability to extract information from signs at the same distances at which drivers can perform such tasks in the real world. These simulators lack sufficient display resolution to render sign stimuli that are readable at a distance. In the study reported here, we designed, built and evaluated a specialized hybrid simulator.”

“Little or no decrement in lane keeping or reading performance was observed at 40km/h on straight roads (but) performance was significantly degraded when participants were required to read digital billboards with 8 or more words at 80km/h.”

Sisiopiku, VP, Islam, M, Haleem, K, Alluri, P. & Gan, A. (2014). Investigation of the Potential Relationship between Crash Occurrence and the Presence of Digital Advertising Billboards in Alabama and Florida. Proceedings of the Transportation Research Board (TRB)

94th Annual Meeting.

Summary: This is a poster presentation showing the crash rates upstream and downstream of digital billboards in two USA states. When all sites are aggregated, it shows an overall increase in the number of crashes prior to the billboard (where it can be seen by the driver) compared to downstream (where it is not visible). However at 50% of the sites examined, the crash records showed that the road became safer after the billboard was installed than it was prior to installation. This is not the consistent pattern that would be expected is billboards presented a hazard.

Moreover, the crash rates in one state were five times higher than those at the other state (across all sites) suggesting that other road safety factors are involved.

Young, MS, Mahfoud, JM, Stanton, N. Salmon, PM, Jenkins, DP & Walker, GH. (2009). “Conflicts of Interest: The implications of roadside advertising for driver attention.” Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 12(5), 381-388.

Summary: The study used a sample of 48 drivers using a vehicle simulator to travel urban, rural and motorway journeys of 4.8 to 9.1km in length (6 runs per participant). After each run, participants were asked to recall the last road sign and, in the case of billboards, the last billboard which they saw. Metrics including time spent out of lane, number of lane excursions, and average and minimum distance to the vehicle ahead were measured.

“On the basis of research so far, whilst it is clear that roadside advertising has potential effects on driver distraction, it remains difficult to be conclusive about the specific risks. Results from early field studies as well as more recent controlled experiments seem to conflict with each other, whilst concern about the risks is based on estimates and self-report data”.

“This study has found that roadside advertising can be detrimental to performance (in terms of lateral control) and pose a distraction for drivers (in terms of increased mental workload). This conclusion stands apart from previous field research, which has been inconclusive or has not found such an effect”

“Whilst the methodological particulars of such studies have already been discussed earlier in this paper, we must also consider the limitations of the laboratory method in interpreting the present results. For instance, the simulated image can never offer the resolution of the real world, and so there may have been some legibility issues with the billboards that could have affected viewing behaviour. Also, the instructions to recall road signs and/or billboards may have influenced performance on subsequent trials, and whilst there is always a trade-off between the benefits of naïve against informed participants in such circumstances, it is possible that this interfered with the attention data. Finally, the analysis of the eye-tracking data by sector was admittedly a coarse approach, an unfortunate consequence of the equipment used, and future studies would benefit from a more precise method of determining allocation of visual attention.”


Annexure B

Carriageway Consulting Review of ‘Before and After’ Crash Records at Digital

Billboards

Assessment of Road Safety Records

Digital Billboard Installations

ETC Media Limited Review of Digital Billboard Safety Records

Table of Contents Main Report Page 1 Introduction 1 2 Stanley Street (State Highway 16) / Alten Road, Auckland 2 2.1 Background 2 2.2 Traffic Flows 3 2.3 Road Safety Records 3 3 Queen Street / Wakefield Street, Auckland 4 3.1 Background 4 3.2 Traffic Flows 5 3.3 Road Safety Records 6 4 Broadway / Remuera Road, Auckland 6 4.1 Background 6 4.2 Traffic Flows 7 4.3 Road Safety Records 7 5 Karangahape Road / Ponsonby Road, Auckland 8 5.1 Background 8 5.2 Traffic Flows 9 5.3 Road Safety Records 9 6 Khyber Pass Road / Symonds Street, Auckland 11 6.1 Background 11 6.2 Traffic Flows 12 6.3 Road Safety Records 12 7 Khyber Pass Road / Southern Motorway, Auckland 13 7.1 Background 13 7.2 Traffic Flows 13 7.3 Road Safety Records 14 8 Khyber Pass Road / Broadway, Auckland 15 8.1 Background 15 8.2 Traffic Flows 16 8.3 Road Safety Records 16 9 Fanshawe Street / Nelson Street, Auckland 17 9.1 Background 17 9.2 Traffic Flows 18 9.3 Road Safety Records 18 10 Victoria Street / Hobson Street, Auckland 19 10.1 Background 19 10.2 Traffic Flows 20 10.3 Road Safety Records 20


11 Tom Pearce Drive / George Bolt Memorial Drive, Auckland 21 11.1 Background 21 11.2 Traffic Flows 22 11.3 Road Safety Records 22 12 Green Lane West / ASB Showgrounds, Auckland 23 12.1 Background 23 12.2 Traffic Flows 23 12.3 Road Safety Records 24 13 Colombo Street / Moorhouse Avenue, Christchurch 25 13.1 Background 25 13.2 Traffic Flows 26 13.3 Road Safety Records 26 14 Victoria Street / Bealey Avenue, Christchurch 27 14.1 Background 27 14.2 Traffic Flows 27 14.3 Road Safety Records 28 15 Main South Road (Sockburn Roundabout), Christchurch 29 15.1 Background 29 15.2 Traffic Flows 29 15.3 Road Safety Records 29 16 Discussion 31 17 Conclusions 33

CCL file reference 14164-billboardsafety final

Status Final

Issued 23 June 2016


1 / 33P.

1. Introduction

1.1. Digital billboards are increasing in popularity within New Zealand, but in some cases, road controlling authorities (“RCAs”) have raised potential road safety concerns associated with driver distraction as a reason for declining resource consent. The particular concerns of RCAs appear to be in relation to the placement of billboards at locations where drivers are required to make a decision and determine an action in respect of their driving behaviour, such as in close proximity to an intersection.

1.2. One such example is set out in a report commissioned by Christchurch City Council and produced by consultants MWH. This notes that “the location of signs relative to traffic signals and other decision points should be considered based on crash history and potential for crashes if the sign goes in…if a sign is installed close to a decision point it should ideally be situated outside the cone of visibility (COV) so that less drivers (sic) are likely to glance at the sign.”

1.3. This report specifically addresses the matter of road safety at digital billboards through evaluating the prevailing conditions at existing installations within New Zealand. In each case, the location of the billboard is assessed in respect of whether it is close to a decision point and/or in the drivers’ cone of vision, and the prevailing traffic environment is also briefly described. The NZTA Crash Analysis System (“CAS”) has been used to identify the traffic flows on the frontage roads where drivers are able to see the signs, and then to assess whether any changes in the number of type of accidents have arisen since the billboard has been installed.


2 / 33P.

2. Stanley Street (State Highway 16) / Alten Road, Auckland

2.1. Background

Figure 1: Aerial View of Billboard Location (Image ©2016 Google)

2.1.1. The billboard is affixed to the southern side of a three-storey building, at the first floor level, as shown below. It was installed in July 2014.

Figure 2: Billboard Location (Image ©2016 Google)

2.1.2. The billboard is directly within the cone of vision for northbound drivers on the state highway, and also forms a backdrop to the adjacent traffic signals. Since it is located immediately adjacent to an intersection, it is sited at a ‘decision point’.

Stanley Street

Billboard Location

Alten Road

Billboard


3 / 33P.

Figure 3: Billboard Forms Backdrop to Traffic Signals (Image ©2016 Google)

2.2. Traffic Flows

2.2.1. The traffic flows on the frontage roads are set out below.

Road Traffic Volumes (Daily, Two-way)

Alten Road 12,000

Stanley Street (north) 43,300

Stanley Street (south) 43,650

Table 1: Traffic Flows on the Frontage Roads

2.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling northbound on Stanley Street, towards the billboard. This traffic flow is in the order of 18,550 vehicles per day.

2.3. Road Safety Records

2.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, July 2009 to June 2014). In this period there were six accidents recorded on Stanley Street northbound involving vehicles travelling towards the billboard location (an average of 1.2 accidents per year).

2.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, July 2014 to present). In this period there was one accident recorded on Stanley Street northbound, involving vehicles travelling towards the billboard (an average of 0.6 accidents per year).

2.3.3. No accidents have been recorded where distraction due to objects outside the vehicle was noted as a contributing factor.


4 / 33P.

3. Queen Street / Wakefield Street, Auckland

3.1. Background


3.1.1. The billboard is affixed to the western side of a multi-storey building, at the first floor level, as shown below. It was installed in July 2013.


3.1.2. The billboard is within the cone of vision for northbound drivers on Queen Street. Since it is located immediately adjacent to an intersection, it is sited at a ‘decision point’.

Billboard

Billboard Location

WakefieldStreet

Queen Street


5 / 33P.

Figure 6: Billboard Within Driver Cone of Vision (Image ©2016 Google)

3.2. Traffic Flows



Queen Street (south) 22,000

Queen Street (north) 22,000

Wakefield Street 12,000


3.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling northbound on Queen Street towards the billboard. This traffic flow is in the order of 11,000 vehicles per day.


3.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, July 2008 to June 2013). In this period there were nine accidents recorded on Queen Street northbound involving vehicles travelling towards the billboard location (an average of 1.8 accidents per year).

3.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, July 2013 to present). In this period there were four accidents recorded on Queen Street northbound involving vehicles travelling towards the billboard (an average of 1.5 accidents per year).



6 / 33P.

4. Broadway / Remuera Road, Auckland

4.1. Background


4.1.1. The billboard is mounted on the top of a two-storey building, as shown below, facing north. It was installed in December 2014.


4.1.2. The billboard is directly within the cone of vision for southbound drivers on Broadway, and also forms a backdrop to the adjacent traffic signals. Since it is located immediately adjacent to an intersection, it is sited at a ‘decision point’.

Billboard

Broadway

Remuera Road

Billboard Location


7 / 33P.


4.2. Traffic Flows



Broadway (north) 25,000

Broadway (south) 25,000

Remuera Road 19,000


4.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling southbound on Broadway, towards the billboard. This traffic flow is in the order of 12,500 vehicles per day.


4.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, December 2009 to November 2014). In this period there were nine accidents recorded on Broadway southbound involving vehicles travelling towards the billboard location (an average of 1.8 accidents per year).

4.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, December 2014 to the present). In this period there was one accident recorded on Broadway southbound involving vehicles travelling towards the billboard (an average of 0.9 accidents per year).



8 / 33P.

5. Karangahape Road / Ponsonby Road, Auckland

5.1. Background


5.1.1. The billboard is mounted at the top of a two-storey building. The billboard is split into three parts (adjacent to one another), as shown below which wrap around the corner. It was installed in July 2014.


5.1.2. The billboard is directly within the cone of vision for northbound drivers on Newton Road and southbound drivers on Karangahape Road, and also forms a backdrop to the adjacent traffic signals for both traffic streams. Since it is located immediately adjacent to an intersection, it is sited at a ‘decision point’.

Billboard

Karangahape Road

Newton Road

Ponsonby Road

Billboard Location

Great North Road


9 / 33P.


5.2. Traffic Flows



Ponsonby Road 25,000

Karangahape Road 25,000

Newton Road 19,000

Great North Road 22,000


5.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling southbound on Karangahape Road and westbound on Newton Road, towards the billboard. These traffic flows are in the order of 12,500 and 9,500 vehicles per day respectively.


5.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, July 2009 to June 2014). In this period there were nine accidents recorded on Karangahape Road southbound and Newton Road westbound involving vehicles travelling towards the billboard location (an average of 1.8 accidents per year).

5.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, July 2014 to present). In this period there was were two accidents recorded on Karangahape Road southbound and Newton Road westbound involving vehicles travelling towards the billboard (an average of 1.2 accidents per year).

5.3.3. Two accidents have been recorded where distraction due to objects outside the vehicle was noted as a contributing factor. One accident occurred when a driver was distracted by a passing pedestrian and ran into the rear of a vehicle in front, and one occurred when a driver


10 / 33P.

was distracted by the flashing lights of a police car (which was attending an accident which had already occurred at the Karangahape Road / Ponsonby Road intersection).


11 / 33P.

6. Khyber Pass Road / Symonds Street, Auckland

6.1. Background


6.1.1. The billboard is affixed to the top of a three-storey building as shown below, and is in three parts that wrap around the corner. It was installed in July 2014.


6.1.2. The billboard is directly within the cone of vision for northbound drivers on Symonds Street and westbound vehicles on Khyber Pass Road. Since it is located immediately adjacent to an intersection, it is sited at a ‘decision point’.

Billboard

Billboard Location

Khyber Pass Road

Symonds Street

Newton Road


12 / 33P.


6.2. Traffic Flows



Symonds Street (north) 45,000

Symonds Street (south) 30,000

Khyber Pass Road 25,000

Newton Road 13,000


6.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling northbound on Symonds Street and westbound on Khyber Pass Road, towards the billboard. These traffic flows are in the order of 15,000 and 12,500 vehicles per day respectively.


6.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, July 2009 to June 2014). In this period there were seven accidents recorded on Symonds Street northbound and Khyber Pass Road westbound involving vehicles travelling towards the billboard location (an average of 1.4 accidents per year).

6.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, July 2014 to the present). In this period there were two accidents recorded on recorded on Symonds Street northbound and Khyber Pass Road westbound involving vehicles travelling towards the billboard (an average of 1.2 accidents per year).



13 / 33P.

7. Khyber Pass Road / Southern Motorway, Auckland

7.1. Background


7.1.1. The billboard is free standing and faces west. It was installed in August 2015.


7.2. Traffic Flows





Billboard

State Highway 1

Khyber Pass Road

Billboard Location


14 / 33P.

7.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling eastbound on Khyber Pass Road, towards the billboard. This traffic flow is in the order of 15,000 vehicles per day.


7.3.1. CAS has been used to identify the road safety records in the vicinity of the billboard for the five-year period immediately prior to the installation of the billboard (that is, August 2010 to July 2015). In this period there was one accident recorded on Khyber Pass Road eastbound involving vehicles travelling towards the billboard location (an average of 0.2 accidents per year).

7.3.2. CAS has also been used to identify the road safety records in the vicinity of the billboard for the period immediately following its installation (that is, July 2015 to present). In this period, no accidents have been recorded on Khyber Pass Road eastbound involving vehicles travelling towards the billboard.



15 / 33P.

8. Khyber Pass Road / Broadway, Auckland

8.1. Background


8.1.1. The billboard is affixed to the western side of a multi-storey building, at the first floor level, as shown below. It was installed in August 2015.


8.1.2. The billboard is directly within the cone of vision for eastbound drivers on Khyber Pass Road, and also forms a backdrop to the adjacent traffic signals. Since it is located immediately adjacent to an intersection, it is sited at a ‘decision point’.

Billboard

Khyber Pass Road

Billboard Location

Broadway


16 / 33P.


8.2. Traffic Flows



Broadway (north) 6,850

Broadway (south) 25,000



8.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling eastbound on Khyber Pass Road, towards the billboard. This traffic flow is in the order of 12,500 vehicles per day.


8.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, August 2010 to July 2015). In this period there were five accidents recorded on Khyber Pass Road eastbound involving vehicles travelling towards the billboard location (an average of 1.0 accidents per year).

8.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, August 2015 to present). In this period there were no accidents recorded on Khyber Pass Road eastbound involving vehicles travelling towards the billboard.



17 / 33P.

9. Fanshawe Street / Nelson Street, Auckland

9.1. Background


9.1.1. The billboard is affixed to the western side of a multi-storey building as shown below. It was installed in February 2016.

Figure 22: Approximate Billboard Location (Image ©2016 Google)

9.1.2. Of particular note is that the billboard is directly within the cone of vision for eastbound drivers on Fanshawe Street.

Billboard

Fanshawe Street

Market Place

Billboard Location Nelson

Street


18 / 33P.


9.2. Traffic Flows



Fanshawe Street (east) 38,700

Fanshawe Street (west) 35,000

Nelson Street 30,000

Market Square 1,000


9.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling eastbound on Fanshawe Street, towards the billboard. This traffic flow is in the order of 19,350 vehicles per day.


9.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, February 2011 to January 2016). In this period there were five accidents recorded on Fanshawe Street eastbound involving vehicles travelling towards the billboard location (an average of 1.2 accidents per year).

9.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, February 2016 to present). In this period there were no accidents recorded on Fanshawe Street eastbound involving vehicles travelling towards the billboard.



19 / 33P.

10. Victoria Street / Hobson Street, Auckland

10.1. Background


10.1.1. The billboard is affixed to the western side of a multi-storey building and is considerably elevated above the nearby roads. It was installed in September 2015.


10.1.2. The billboard is directly within the cone of vision for eastbound drivers on Victoria Street. Since it is located in close proximity to an intersection, it is sited at a ‘decision point’.

Billboard

Victoria Street

Hobson Street

Billboard Location


20 / 33P.


10.2. Traffic Flows



Victoria Street (east) 25,000

Victoria Street (west) 25,000

Hobson Street (north) 25,000

Hobson Street (south) 25,000


10.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling eastbound on Victoria Street, towards the billboard. This traffic flow is in the order of 12,500 vehicles per day.


10.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, September 2010 to August 2015). In this period there were three accidents recorded on Victoria Street eastbound involving vehicles travelling towards the billboard location (an average of 0.6 accidents per year).

10.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, September 2015 to present). In this period there were no accidents recorded on Victoria Street eastbound, involving vehicles travelling towards the billboard.



21 / 33P.

11. Tom Pearce Drive / George Bolt Memorial Drive, Auckland

11.1. Background


11.1.1. The billboard is free-standing, and was installed in December 2013.


11.1.2. The billboard is directly within the cone of vision for southbound drivers on George Bolt Memorial Drive and eastbound vehicles on Tom Pearce Drive. It also forms a backdrop to the adjacent traffic signals and as it is located immediately adjacent to an intersection, it is sited at a ‘decision point’.

Billboard

George Bolt Memorial Drive

Tom Pearce Drive

Billboard Location


22 / 33P.


11.2. Traffic Flows



George Bolt Memorial Drive (north)

No traffic data available George Bolt Memorial Drive (south)

Tom Pearce Drive (east)

Tom Pearce Drive (west)


11.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling southbound on George Bolt Memorial Drive and eastbound on Tom Pearce Drive, towards the billboard.


11.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, December 2008 to November 2013). In this period there were 19 accidents recorded on George Bolt Memorial Drive southbound and Tom Pearce Drive eastbound involving vehicles travelling towards the billboard location (an average of 3.8 accidents per year).

11.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, December 2013 to present). In this period there were five accidents recorded on George Bolt Memorial Drive southbound and Tom Pearce Drive eastbound involving vehicles travelling towards the billboard (an average of 2.3 accidents per year).



23 / 33P.

12. Green Lane West / ASB Showgrounds, Auckland

12.1. Background


12.1.1. The billboard is free standing. The date of installation is not known, but is understood to be prior to 2014.


12.2. Traffic Flows



Green Lane West 30,000


Billboard

Green Lane West

Billboard Location


24 / 33P.

12.2.2. As the billboard is double-sided, all of these drivers travelling towards the billboard will be able to see it.


12.3.1. CAS has been used to identify the road safety records in the vicinity of the billboard for the five-year period immediately prior to the (assumed) installation of the billboard (that is, January 2009 to December 2013). In this period there were eight accidents recorded on Green Lane West involving vehicles travelling towards the billboard location (an average of 1.6 accidents per year).

12.3.2. CAS has also been used to identify the road safety records in the vicinity of the billboard for the period immediately following its installation (that is, January 2014 to present). In this period there were five accidents recorded on Green Lane West involving vehicles travelling towards the billboard (an average of 2.3 accidents per year).



25 / 33P.

13. Colombo Street / Moorhouse Avenue, Christchurch

13.1. Background


13.1.1. The billboard free-standing, and is located in the same position at a previous (static) billboard, but elevated to a greater height, as indicatively shown below. It was installed in January 2016.

Figure 33: Approximate Billboard Location (Image ©2016 Google)

13.1.2. Of particular note is that the billboard is directly within the cone of vision for westbound drivers on Moorhouse Avenue and also turning onto Colombo Street.

Billboard Position

Moorhouse Avenue

Billboard Location Colombo

Street


26 / 33P.


13.2. Traffic Flows



Moorhouse Avenue (east) 35,500

Moorhouse Avenue (westbound slip) 3,750

Colombo Street (north) 15,000

Colombo Street (south) 15,100


13.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling westbound on Moorhouse Avenue or the off-ramp, towards the billboard. These traffic flows are in the order of 17,500 and 3,750 vehicles per day respectively.


13.3.1. CAS has been used to identify the road safety records in the vicinity of the billboard for the five-year period immediately prior to the installation of the billboard (that is, January 2011 to December 2015). In this period there were no accidents recorded involving vehicles travelling towards the billboard location. However this may be due in part to the extensive repairs that have been underway on the Moorhouse Avenue overbridge over much of this period, and lower temporary speed limit that has been in place.

13.3.2. CAS also been used to identify the road safety records in the vicinity of the billboard for the period immediately following its installation (that is, January 2016 to present). In this period there were no accidents recorded involving vehicles travelling towards the billboard.



27 / 33P.

14. Victoria Street / Bealey Avenue, Christchurch

14.1. Background


14.1.1. The billboard is free-standing is located just to the north of a large multi-storey building. It was installed in November 2015.

Figure 36: Billboard Location

14.1.2. The billboard is directly within the cone of vision for southbound drivers on Papanui Road, and as it is located immediately adjacent to an intersection, it is sited at a ‘decision point’.

14.2. Traffic Flows


Victoria Street

Billboard Location

Bealey Avenue

Papanui Road

Billboard


28 / 33P.


Bealey Avenue (east) 37,800

Bealey Avenue (west) 35,100

Victoria Street 15,000

Papanui Road 24,200


14.2.2. Not all of these drivers will be able to see the billboard but rather, it is considered that it will be visible only to those drivers that are travelling southbound on Papanui Road, towards the billboard. This traffic flow is in the order of 12,100 vehicles per day.


14.3.1. CAS has been used to identify the road safety records at the intersection for the five-year period immediately prior to the installation of the billboard (that is, November 2010 to October 2015). In this period there were two accidents recorded on Papanui Road southbound involving vehicles travelling towards the billboard location (an average of 0.4 accidents per year).

14.3.2. CAS has also been used to identify the road safety records at the intersection for the period immediately following the installation of the billboard (that is, November 2015 to present). In this period there were no accidents recorded on Papanui Road southbound involving vehicles travelling towards the billboard.



29 / 33P.

15. Main South Road (Sockburn Roundabout), Christchurch

15.1. Background


15.1.1. The billboard is free-standing and is located on the northern side of Main South Road. It is double-sides and therefore visible to traffic approaching from both directions. It was installed in October 2015.

15.1.2. The billboard is directly within the cone of vision for northbound drivers on Main South Road.

15.2. Traffic Flows



Main South Road 27,700


15.2.2. All of these drivers travelling towards the billboard will be able to see it.


15.3.1. CAS has been used to identify the road safety records in the vicinity of the billboard for the five-year period immediately prior to the installation of the billboard (that is, October 2010 to September 2015). In this period there were five accidents recorded on Main South Road involving vehicles travelling towards the billboard location (an average of 1.0 accidents per year).

15.3.2. CAS has also been used to identify the road safety records in the vicinity of the billboard for the period immediately following its installation (that is, October 2015 to present). In this period there have been no accidents recorded on Main South Road involving vehicles travelling towards the billboard.

Billboard Location

Main South Road


30 / 33P.



31 / 33P.

16. Discussion

16.1. Data Summary

16.1.1. The locational information, traffic volumes and accident information set out in each section above has been summarised, below.

Location At Decision Point?

Within Cone of Vision?

Backgroundto Signals?

Views Per Day (Traffic

Volume)

Accident Rates

Before Billboard

After Billboard

Stanley Street (State Highway 16) / Alten Road, Auckland Yes Yes Yes 18,550 1.2

(5-year avg) 0.6

(1.7-year avg)

Queen Street / Wakefield Street, Auckland Yes Yes No 11,000 1.4

(5-year avg) 1.5

(2.7-year avg)

Broadway / Remuera Road, Auckland Yes Yes Yes 12,500 1.8

(5-year avg) 0.9

(1.3-year avg)

Karangahape Road / Ponsonby Road, Auckland Yes Yes Yes 22,000 1.8

(5-year avg) 1.2

(1.7-year avg)

Khyber Pass Road / Symonds Street, Auckland Yes Yes No 27,500 1.4

(5-year avg) 1.2

(1.7-year avg)

Khyber Pass Road / Southern Motorway, Auckland No Yes No 15,000 0.2

(5-year avg) 0.0

(0.6-year avg)

Khyber Pass Road / Broadway, Auckland Yes Yes Yes 12,500 1.0

(5-year avg) 0.0

(0.6-year avg)

Fanshawe Street / Nelson Street, Auckland Yes Yes No 19,350 1.2

(5-year avg) 0.0

(0.1-year avg)

Victoria Street / Hobson Street, Auckland Yes Yes No 12,500 0.6

(5-year avg) 0.0

(0.5-year avg)

Tom Pearce Drive / George Bolt Memorial Drive, Auckland Yes Yes Yes unknown 3.8

(5-year avg) 2.3

(2.3-year avg)

Green Lane West / ASB Showgrounds, Auckland No Yes No 30,000 1.6

(5-year avg) 2.3

(2.2-year avg)

Colombo Street / Moorhouse Avenue, Christchurch No Yes No 21,250 0.0

(5-year avg) 0.0

(0.2-year avg)

Victoria Street / Bealey Avenue, Christchurch Yes Yes No 12,100 0.4

(5-year avg) 0.0

(0.3-year avg)

Main South Road (Sockburn Roundabout), Christchurch No Yes No 27,700 1.0

(5-year avg) 0.0

(0.3-year avg)

Table 15: Summary of Characteristics of Billboards

16.2. Discussion

16.2.1. It can be seen that of the 14 sites considered within this report, ten are located at driver ‘decision points’, that is, at intersections. All of these billboards are sited within the ‘cone of vision’ of the driver, and in five of these ten cases, the digital billboard forms a background to the traffic signals heads themselves. Being within the cone of vision and the billboard being a background to traffic signals are typically matters that are highlighted as road safety concerns/risks when an application for a new digital billboard is made. Accordingly, it would be expected that accident rates should increase with the billboard in place.

16.2.2. However the data for nine of these ten sites shows that there has been a decrease in the accident rates, subsequent to the billboard being installed.


32 / 33P.

16.2.3. Accidents are by their nature random and infrequent events, and the timeframes for which the post-construction accident analysis has been carried out are relatively short. Under the NZTA Economic Evaluation Manual, a five-year timeframe is required for the determination of an accident rate for any given site, but no digital billboards have been installed for this length of time which precludes such an assessment. Nevertheless, the aggregation of these ten sites means that a total of 152 months of data has been evaluated, which is considered to be a suitably robust data set, especially given that the sites each have particularly ‘risky’ factors.

16.2.4. A further assessment has been carried out to rebase the data in terms of the number of accidents per million ‘views’ by drivers (that is, the number of vehicles passing the site where a driver would have the potential to view the billboard).



Background to Signals?

Accidents Per Million Views

Before Billboard After Billboard

Stanley Street (State Highway 16) / Alten Road, Auckland

Yes Yes Yes

0.24 0.09

Broadway / Remuera Road, Auckland 0.39 0.18

Karangahape Road / Ponsonby Road, Auckland 0.22 0.15

Khyber Pass Road / Broadway, Auckland 0.22 0.00

Subtotal 0.26 0.12

Queen Street / Wakefield Street, Auckland

Yes Yes No

0.35 0.37

Khyber Pass Road / Symonds Street, Auckland 0.14 0.12

Fanshawe Street / Nelson Street, Auckland 0.17 0.00

Victoria Street / Hobson Street, Auckland 0.13 0.00

Victoria Street / Bealey Avenue, Christchurch 0.09 0.00

Subtotal 0.17 0.19

Total 0.20 0.15

Table 16: Accident Rates at Each Site at a ‘Decision Point’

16.2.5. The subtotals and total have been calculated by a weighted sum approach, meaning that there is a bias towards those sites where data has been collected over a longer period of time.

16.2.6. Overall, the rate of accidents per million view decreases post installation of the billboards, and a decrease is seen for those sites where the billboard forms a background to the traffic signal heads. There is a slight increase seen at the Queen Street / Wakefield Street site, and this skews the overall result for those sites where the billboard does not form a background to the traffic signals since the data from this site represents slightly more than half of the data set.

16.2.7. The review also identified that no accidents were recorded at any of the 14 sites assessed where distraction due to an external source (which includes, but is not limited to, digital billboards) was a factor.


33 / 33P.

17. Conclusions

17.1. This report has identified, evaluated and assessed the road safety records and traffic characteristics of 14 sites where digital billboards presently operate. Of the 14 sites, ten were at ‘decision points’ for drivers (that is, at intersections) and five of these were at locations where approaching drivers are able to see the billboard directly behind the traffic signal head. Both of these factors are commonly mentioned as presenting a particular road safety risk when resource consent applications are made for new digital billboards. As such, it would be expected that the accident rates increase once the billboard is in place.

17.2. However, the data shows that the accident rate observed after the billboard was operating is lower than the rate observed prior to the billboard being installed and commissioned.

17.3. Data is required for a five-year period at each site for a robust determination of an accident rate, but no digital billboards have been in place for this length of time. As a result, the analyses have been based on the aggregation of 152 months of data. While this is less than the ideal of five years of data per site, the data set is considered to be suitably robust, especially given that each of the sites have particular ‘high risk’ factors and so any adverse trends in safety should be more evident.

17.4. In view of the lack of any increase in accident rates after the digital billboards are installed, there is no evidence from the CAS data that the operation of digital billboards gives rise to an increase in the number of accidents.

Carriageway Consulting Limited May 2016

traffic engineering | transport planning

A. PO Box 29623, Christchurch, 8540 P. 03 377 7010 E. [email protected]


Annexure C

Carriageway Consulting Review of ‘Before and After’ Crash Records at Digital

Billboards Using Control Sites

A.

P.

E.

CCL Ref: 14164-080817-mccaleb.docx 8 August 2017 Mike McCaleb ETC Media Limited By e-mail only: [email protected] Dear Mike

Review of Safety Records at Existing Digital Billboards

Further to our various e-mails and discussions, I am writing to formalise the outcomes of my additional research into the safety records around digital billboards.

Initial Study

As you know, in 2016 we carried out a review of the reported crashes around 14 existing sites in New Zealand. The methodology involved reviewing the crash records in the vicinity of each site for a period of time before the digital billboards were installed and for a period of time after the billboard was installed, and comparing the two sets of data. The results are summarised below:




Views Per Day (Traffic

Volume)

Accident Rates

Before Billboard

After Billboard

Stanley Street (State Highway 16) / Alten Road, Auckland Yes Yes Yes 18,550 1.2

(5-year avg) 0.6

(1.7-year avg)

Queen Street / Wakefield Street, Auckland Yes Yes No 11,000 1.4

(5-year avg) 1.5

(2.7-year avg)

Broadway / Remuera Road, Auckland Yes Yes Yes 12,500 1.8 (5-year avg)

0.9 (1.3-year avg)

Karangahape Road / Ponsonby Road, Auckland Yes Yes Yes 22,000 1.8

(5-year avg) 1.2

(1.7-year avg)

Khyber Pass Road / Symonds Street, Auckland Yes Yes No 27,500 1.4

(5-year avg) 1.2

(1.7-year avg)

Khyber Pass Road / Southern Motorway, Auckland No Yes No 15,000 0.2

(5-year avg) 0.0

(0.6-year avg)

Khyber Pass Road / Broadway, Auckland Yes Yes Yes 12,500 1.0

(5-year avg) 0.0

(0.6-year avg)

Fanshawe Street / Nelson Street, Auckland Yes Yes No 19,350 1.2

(5-year avg) 0.0

(0.1-year avg)

Victoria Street / Hobson Street, Auckland Yes Yes No 12,500 0.6

(5-year avg) 0.0

(0.5-year avg)

Tom Pearce Drive / George Bolt Memorial Drive, Auckland Yes Yes Yes unknown 3.8

(5-year avg) 2.3

(2.3-year avg)

Green Lane West / ASB Showgrounds, Auckland No Yes No 30,000 1.6

(5-year avg) 2.3

(2.2-year avg)

Colombo Street / Moorhouse Avenue, Christchurch No Yes No 21,250 0.0

(5-year avg) 0.0

(0.2-year avg)

Victoria Street / Bealey Avenue, Christchurch Yes Yes No 12,100 0.4

(5-year avg) 0.0

(0.3-year avg)

Main South Road (Sockburn Roundabout), Christchurch No Yes No 27,700 0.6

(5-year avg) 0.0

(0.3-year avg)

Table 1: Summary of Characteristics of Billboards

2 / 9P.

A total of 14 sites were considered within this report, of which ten were located at driver ‘decision points’, that is, at intersections. All of these billboards were sited within the ‘cone of vision’ of the driver, and in five of these ten cases, the digital billboard formed a background to the traffic signals heads themselves. Being within the cone of vision and the billboard being a background to traffic signals are typically matters that are highlighted as road safety concerns/risks when an application for a new digital billboard is made. Accordingly, it would be expected that accident rates should increase with the billboard in place.

However the data for nine of these ten sites showed that there had been a decrease in the accident rates, subsequent to the billboard being installed:




Accidents Per Million Views

Before Billboard After Billboard

Stanley Street (State Highway 16) / Alten Road, Auckland

Yes Yes Yes

0.24 0.09

Broadway / Remuera Road, Auckland 0.39 0.18

Karangahape Road / Ponsonby Road, Auckland 0.22 0.15

Khyber Pass Road / Broadway, Auckland 0.22 0.00

Subtotal 0.26 0.12

Queen Street / Wakefield Street, Auckland

Yes Yes No

0.35 0.37

Khyber Pass Road / Symonds Street, Auckland 0.14 0.12

Fanshawe Street / Nelson Street, Auckland 0.17 0.00

Victoria Street / Hobson Street, Auckland 0.13 0.00

Victoria Street / Bealey Avenue, Christchurch 0.09 0.00

Subtotal 0.17 0.19

Total 0.20 0.15

Table 2: Accident Rates at Each Site at a ‘Decision Point’

Accidents are by their nature random and infrequent events, and the timeframes for which the post-construction accident analysis was carried out were relatively short. Under the NZTA Economic Evaluation Manual, a five-year timeframe is required for the determination of an accident rate for any given site, but no digital billboards have been installed for this length of time which in turn precludes such an assessment. Nevertheless, the aggregation of these ten sites meant that a total of 152 months of data had been evaluated, which was considered to be a suitably robust data set, especially given that the sites each had particularly ‘risky’ factors. We also noted that the aggregation of data in this manner is an approach that has been used by other research into digital billboards1.

Accordingly, we concluded that there was no evidence that the operation of the digital billboards examined gave rise to an increase in the number of accidents.

1 See for example Gitelman, V., Zaidel, D., Doveh, E., & Zilberstein, R. (2014) “The Impact of Billboards on Road Accidents on Ayalon Highway Three Periods Comparison – Billboards Present, Removed, and Returned.” Report to the Israeli National Road Authority. The periods used in this paper were 2 years (‘present’), 1.5 years (‘removed’) and 3.5 years (‘returned’).

3 / 9P.

Methodology Refinement

The study described above was critiqued by Dr Shane Turner of MWH (now part of Stantec)2 and his concerns regarding the robustness of the methodology were taken into account within a updated and refined methodology for the study.

Under the updated/refined approach, five digital billboards in Auckland were identified, together with the date when they were installed. The crash records at each location were then reviewed for the two years prior to their installation and the two years after their installation, and the records were compared to see whether there had been an increase in crash numbers. The locations were selected to be all signalised intersections.

At the same time, control sites were selected in close proximity to each of the digital billboards. One control site had a static billboard which had been in place throughout the four years of crash records. One control site had no billboard erected in this time period.

The locations chosen were:

Site Group Locations

Digital Static No Billboard

1 Stanley Street (State Highway 16) / Alten Road

Stanley Street (State Highway 16) / Beach Road /

Parnell Rise

The Strand (State Highway 16) / Tamaki Drive

2 Queen Street / Wakefield Street Mayoral Drive / Queen Street Mayoral Drive / Wakefield

Street

3 Karangahape Road / Ponsonby Road

Ponsonby Road / Richmond Road / Picton Street

Ponsonby Road / Williamson Ave / Hepburn Street

4 Khyber Pass Road / Symonds Street

Symonds Street / Alex Evans Street / Overbridge

Symonds Street / Mt Eden Road

5 George Bolt Memorial Drive / Tom Pearce Drive - George Bolt Memorial Drive /

Montgomerie Road

Table 3: Locations Assessed for Road Safety in Updated Study

No static billboard could be found within Group 5 and so no assessment was carried out of this.

Comparisons of the road safety records at each location were carried out were as follows:

At intersections where a digital billboard had been installed: Two years of accident records prior to the installation and two years after installation;

At intersections where a static billboard was present throughout: Two years of accident records prior to the installation of the nearby digital billboard and two years after installation; and

At intersections where no billboard was present throughout: Two years of accident records prior to the installation of the nearby digital billboard and two years after installation.

Traffic flows at each location were kindly provided by Auckland Transport.

Flow Transportation Consultants agreed to undertake the analysis. They were provided with a list of the intersections which had been randomised (using a random number generator to avoid any bias). They were then asked to identify all reported crashes using the NZTA Crash Analysis System within 75m of the intersection where driver distraction may have been a contributing cause. That 2 Shane Turner letter dated 1 December 2016

4 / 9P.

is, driver distraction did not need to be specifically noted as a factor in the record, but the type of crash recorded was of a type that could have involved distraction. For example, a nose-to-tail crash in the queue approaching the intersection or a driver failing to stop at a red signal were identified as crashes which could have been due to distraction. Crashes where a driver had a sudden medical event or vehicle failure were not assumed to be caused by distraction.

Flow then carried out a statistical test (a t-test at the 95% level) to compare the results for any statistically significant differences.

Of critical importance to this study, Flow graciously agreed to undertake the assessment on a ‘double-blind’ basis. That is, as well as not knowing which intersections had digital billboards, static billboards or no billboards, Flow was not informed of the purpose of this study until all analyses had been complete. This meant that there could be no subconscious bias in the selection or rejection of crashes, or in the analysis itself.

Turner Critique

The concerns raised by Dr Turner in his critique of the earlier study are set out below, together with the ways in which they have been addressed in the revised methodology.

Turner Concern: The impact of advertising signs will be different in each case depending on intersection complexity and other crash causing features

Abley Transportation Consultants have recently completed some research for the New Zealand Transport Agency3 which addressed the factors at urban signalised intersections which were observed to contribute to crashes. Accordingly, to address this concern, Abley was given a list of intersections some of which had digital billboards, some had static billboards and some had no billboards, and was asked to assess each intersection according to their methodology for rating the various risk factors. The list given to Abley was randomised, and although the project manager of the work (Paul Durdin) was told the purpose of the study, he was instructed not to divulge this to the member of staff who was to carry out the work. This approach avoids any bias in the results.

Turner Concern: Crashes are rare and so it may not be possible to discern any trend from the data

The approach of assessing crash records has been adopted by at least three independent research papers that Dr Turner has relied upon when informing his views on the road safety effects of billboards4. Since this is research on which weight has been placed by Dr Turner and others (for example, Wachtel), we consider that the same methodology can be used in this instance also.

3 NZTA Research Report 588 ‘Why are some urban traffic signals much less safe than others?’ September 2016 4 Hawkins, HG, Jr., Kuo, PF, & Lord, D. (2014). “Statistical Analysis of the Traffic Safety Impacts of On-Premise Digital Signs.” Paper No: 14-2772. Presented at the 93rd Annual Meeting of the Transportation Research Board. Gitelman, V., Zaidel, D., Doveh, E., & Zilberstein, R. (2014) “The Impact of Billboards on Road Accidents on Ayalon Highway Three Periods Comparison – Billboards Present, Removed, and Returned.” Report to the Israeli National Road Authority. Sisiopiku, VP, Islam, M, Haleem, K, Alluri, P. & Gan, A. (2014). Investigation of the Potential Relationship between Crash Occurrence and the Presence of Digital Advertising Billboards in Alabama and Florida. Proceedings of the Transportation Research Board (TRB) 94th Annual Meeting.

5 / 9P.

Turner Concern: The police records may not explicitly set out distraction as a factor, so there may be under-reporting of this

The approach used by Flow was not to determine whether distraction was a factor from the crash records but whether the type of crash could have distraction involved. This eliminates the potential for any reporting bias in the police records.

Turner Concern: The sites do not have a minimum of three years ‘after’ data and none have the ideal of five years that is usually used for such studies

The nature of digital billboards is that at the time of the study, none had been installed for three years or more, which meant that it was not possible to use data from a longer period of time. We note though that there are studies relied on by Dr Turner5 which also use short time periods. Nevertheless we agree that if longer periods were available, then this would improve the robustness of the results.

Turner Concern: The analysis approach does not take into account the trend in crashes (most likely downward) that might have occur at the site in the ‘after’ period. The correct approach is to use control sites to understand how the modified site may have performed crash-wise

The study includes two control sites for most of the digital billboard locations, one with a static billboard and one with no billboard.

Turner Concern: There may have been other changes to the site during the analysis period that has influenced the number of crashes in the after period

We confirm that to our knowledge, there have been no changes to the intersections that are assessed in the updated/refined study.

Turner Concern: The data set is small - most studies of this type would typically have in excess of 50 sites

We agree that the data set is limited, but in large part this is due to there being only a small number of digital billboards that have been installed for any length of time in New Zealand. That said, we have reviewed a substantial number of peer-reviewed research papers into digital billboard safety and note that many of these also use a limited data set.

Turner Concern: Test statistics need to be used to determine whether the change in crashes observed is statistical significant

The work carried out by Flow used a t-test to check for statistical significance at the 95% confidence level.

Overall, we consider that the updated/refined approach has appropriately responded to Dr Turner’s concerns with our earlier work, taking into account that an optimum approach of having at least three years of ‘before’ and ‘after’ data for at least 50 sites simply is not possible at the present time.

5 See for example Gitelman, V., Zaidel, D., Doveh, E., & Zilberstein, R. (2014) “The Impact of Billboards on Road Accidents on Ayalon Highway Three Periods Comparison – Billboards Present, Removed, and Returned.” Report to the Israeli National Road Authority. The periods used in this paper were 2 years (‘present’), 1.5 years (‘removed’) and 3.5 years (‘returned’).

6 / 9P.

Results

In total, the analyses of the ‘before’ period included a total of 140 crashes and 279,000 vehicle movements per day. The ‘’after’ period included a total of 89 crashes and 343,000 vehicle movements per day.

When looking at each intersection individually (that is, simply comparing each intersection approach against the other approaches at the same intersection), Flow found that there were statistically significant differences at the following locations:

Type Type Before After

Digital

Stanley Street (SH16) / Alten Road (no data) No difference

Queen Street / Wakefield Street Wakefield Street Wakefield Street

Karangahape Road / Ponsonby Road No difference K Road

Khyber Pass Road / Symonds Street No difference No difference

George Bolt Memorial Drive / Tom Pearce Drive No difference No difference

Static

Stanley St (SH16) / Beach Road / Parnell Rise No difference No difference

Mayoral Drive / Queen Street No difference Mayoral Drive (east)

Ponsonby Road / Richmond Road / Picton Street No difference No difference

Symonds Street / Alex Evans Street / Overbridge No difference No difference

No billboard

The Strand (SH16) / Tamaki Drive (no data) No difference

Mayoral Drive / Wakefield Street Mayoral Drive (west) No difference

Ponsonby Road / Williamson Ave / Hepburn St No difference No difference

Symonds Street / Mt Eden Road No difference No difference

George Bolt Memorial Drive / Montgomerie Road No difference No difference

Table 4: Whether Any Approaches had Statistically Different Numbers of Crashes Recorded

It can be seen that for the most part, there were no intersections where the crash rate on one particular approach was greater than on the other approaches to the intersection. In respect of the locations where digital billboards were installed, there has been a change at only one approach – Karangahape Road. We can confirm that this is a location from which the digital billboard can be seen. However the billboard can also be seen from the Newton Road approach at this intersection, and this did not have any statistically different safety record after the digital billboard was installed compared to before installation.

Flow then compared the crash record at each intersection against the crash records at the other intersections using a t-test (at the 95% confidence level) and the following results were obtained. Note that this test did not compare the ‘before’ and ‘after’ data in each location. Rather, it compared whether for a particular intersection in the ‘before’ period, the overall crash rate was different to the other intersections in the ‘before’ period. It then compared whether for a particular intersection in the ‘after’ period, the overall crash rate was different to the other intersections in the ‘after’ period.

7 / 9P.

Type Type Before After

Digital

Stanley Street (SH16) / Alten Road (no data) No difference

Queen Street / Wakefield Street No difference No difference

Karangahape Road / Ponsonby Road No difference No difference



Static


Mayoral Drive / Queen Street No difference No difference



No billboard

The Strand (SH16) / Tamaki Drive (no data) No difference

Mayoral Drive / Wakefield Street No difference No difference




Table 4: Whether Any Intersections had Statistically Different Numbers of Crashes Recorded

The results show that no intersections (as a whole) had a crash record that was statistically different to any others in either the before or after period.

Finally, Flow compared the crash record of each intersection in the ‘before’ period with the ‘after’ period. Initially this was done for each approach, and then for the intersection as a whole.

Type Type Approach Intersection

Digital

Stanley Street (SH16) / Alten Road (no data) (no data)

Queen Street / Wakefield Street No difference No difference

Karangahape Road / Ponsonby Road K Road No difference



Static


Mayoral Drive / Queen Street Mayoral Drive (east) No difference



No billboard

The Strand (SH16) / Tamaki Drive (no data) (no data)

Mayoral Drive / Wakefield Street Mayoral Drive (west) No difference




Table 5: Whether Any Intersections had Statistically Different Numbers of Crashes Recorded ‘Before’ and ‘After’

It can be seen that in each of the types of sites (digital, static and no billboard), there was one approach where a statistically significant increase in crashes occurred. Clearly though, there is no effect due to billboards at the ‘no billboard’ location, and no effects due to digital billboards at the

8 / 9P.

‘static’’ locations. Overall, none of the intersections as a whole had a statistically significant change in crashes between the ‘before’ and ‘after’ periods.

Risk Factors at Intersections

As set out above, Abley Transportation Consultants has assessed each location and identified the Level of Safety Service according to NZTA research. The results of this are set out below.

Type Type Approach Intersection Level of Safety

Service

Digital

Stanley Street (SH16) / Alten Road (no data) (no data) Medium

Queen Street / Wakefield Street No difference No difference Medium

Karangahape Road / Ponsonby Road K Road No difference High

Khyber Pass Road / Symonds Street No difference No difference Medium

George Bolt Memorial Dr / Tom Pearce Dr No difference No difference n/a

Static

Stanley St (SH16) / Beach Rd / Parnell Rise No difference No difference Low-medium

Mayoral Drive / Queen Street Mayoral Dr (east) No difference Low-medium

Ponsonby Rd / Richmond Rd / Picton St No difference No difference Medium

Symonds St / Alex Evans St / Overbridge No difference No difference Low-medium

No billboard

The Strand (SH16) / Tamaki Drive (no data) (no data) Low-medium

Mayoral Drive / Wakefield Street Mayoral Dr (west) No difference Low

Ponsonby Rd / Williamson Ave / Hepburn St No difference No difference Medium

Symonds Street / Mt Eden Road No difference No difference Medium

George Bolt Memorial Dr / Montgomerie Rd No difference No difference Medium-high

Table 6: Level of Safety Service Compared to ‘Before’ and ‘After’ Crash Performance

The results show that the Level of Safety Service does not correspond to the presence of digital billboards. While there is one location (Karangahape Road / Ponsonby Road) which is ranked as having a high risk, the remainder are medium risk. The locations with static billboards and no billboards typically have medium-high, medium, or medium-low risks.

Discussion

The data set for the assessment is necessarily limited by the number of billboards that have been operating in New Zealand for any length of time, and the anticipated five sites was ultimately reduced to four sites due to an absence of traffic volume information at one location prior to the billboard being installed.

In three locations where digital billboards were installed, there was no statistically significant difference in the number of crashes after installation compared to before installation. In one location, one approach from which a digital billboard is visible showed an statistically significant increase in crashes. However a second approach at the same intersection from which the billboard is also visible did not show any such increase. None of the intersections as a whole showed any statistically significant increase.

The comparison of the safety performance between the various intersections demonstrates that the sample of control sites does not differ significantly from the digital or static billboard sites. We

9 / 9P.

are confident then that the control sites appropriately reflect a ‘baseline’ performance against which the digital billboards can be compared.

Based on this study, we conclude that there is no evidence at the four sites examined that digital billboards have resulted in increased road safety risks at these intersections.

We trust that this is of assistance, but please do not hesitate to contact me if there are any issues that you wish to discuss or if you would like clarification of any matters.

Kind regards Carriageway Consulting Limited

Andy Carr Traffic Engineer | Director

Mobile 027 561 1967 Email [email protected]


Annexure D

Review of Primary Literature Referenced by Dr Walton

Beijer, D., Smiley, A., & Eizenman, M. (2004). Observed driver glance behavior at

roadside advertising signs. Transportation Research Record: Journal of the

Transportation Research Board, (1899), 96–103.

1 This report draws a distinction between active signs, which include moveable

displays or components, and passive signs which do not. Dr Walton sets out that

in this study, almost all drivers glanced at the advertising along the route.

However, this is not surprising since the section of road which was used for the

study contained 37 advertising signs.

2 Importantly, the number of glances towards static signs was significantly lower

than towards active signs. As noted above, this application now proposes a static

sign (since animation of moving components is to be prohibited), albeit one where

the image changes on a regular basis.

3 The average glance duration observed in the study was less than 0.6 seconds.

This duration of glance is important because there is research which shows that

there is an elevated risk of a crash where drivers take their eyes of the road for

more than 2 seconds12.

4 No assessment was made in this study regarding driver glances at other signage,

such as road signs.

Crundall, D., Van Loon, E., & Underwood, G. (2006). Attraction and distraction of

attention with roadside advertisements. Accident Analysis & Prevention, 38(4), 671–

677.

5 This paper evaluated whether advertisements at street level (the height of a

typical bus shelter) or elevated 3m above street level attracted driver attention to

a great or lesser extent. The authors concluded “street level advertisements

attract and hold attention at inappropriate times compared to raised-level

advertisements”. This application is for a raised-level advertisement, and

therefore (based solely on this study) is the safer option of the two.

12 5. S.G. Klauer, T. a. Dingus, V.L. Neale, J.D. Sudweeks, D.J. Ramsey, The Impact of Driver Inattention on Near-Crash/Crash Risk: An Analysis using the 100-car Naturalistic Driving Study Data. Administration Report No. DOT HS 810 594 (National Highway Traffic Safety Administration, Washington, D.C., 2006) 6. T.H. Rockwell, Spare visual capacity. New empirical results for an old idea, Vis. Veh. -- II. pp. 317–324 (1988)


6 It is important to note that this study did not compare digital billboards with static

billboards, and in fact the study indicates that the advertisements were static. No

assessment was carried out of whether billboards per se give rise to road safety

issues.

Domke, K., Wandachowicz, K., Zalesińska, M., Mroczkowska, S., & Skrzypczak, P.

(2011). Digital billboards and road safety. Light in Engineering, Architecture and the

Environment, 87, 119–131.

7 This paper describes that where images are displayed in sequence, the observer

is “provoked” to wait to see the remaining part of the message, leading to a much

worse concentration of the driver. This propensity of a person to wait to see the

end of a task that has already been started is known as the Zeigarnik effect.

8 In this application though, sequential images will not be displayed on the

billboard, but rather, adverts which follow one another will be unrelated.

Dukic, T., Ahlstrom, C., Patten, C., Kettwich, C., & Kircher, K. (2013). Effects of

electronic billboards on driver distraction. Traffic Injury Prevention, 14(5), 469–476.

9 Rather than setting out the details of this paper, I simply note the researchers

conclusion that “whether the electronic billboards attract too much attention and

constitute a traffic safety hazard cannot be answered conclusively based on the

present data”.

Megías, A., Maldonado, A., Catena, A., Di Stasi, L. L., Serrano, J., & Cándido, A.

(2011). Modulation of attention and urgent decisions by affect-laden roadside

advertisement in risky driving scenarios. Safety Science, 49(10), 1388–1393.

10 This paper compared whether the emotional content of a billboard image affected

driver attention. It did not assess whether digital billboards per se gave rise to

increased road safety risk.

11 In this case, the images displayed will meet the relevant regulations for

advertisers and therefore will not attract driver attention to any greater extent than

is already permitted under those regulations.

12 Dr Walton sets out that this paper specifically refers to NZ advertisements. This is

correct, but the authors are explicitly referring to government roadside advertising

campaigns to improve road safety, and not to general commercial advertising.


Stutts, J. C., Reinfurt, D. W., Staplin, L., & Rodgman, E. A. (2001). The role of driver

distraction in traffic crashes. Report prepared for AAA Foundation for Traffic Safety,

Washington, DC.

13 I concur with Dr Walton’s comments regarding this study, that the authors were

unable to identify billboards as being a distraction to drivers. In fact the authors

noted “some items - such as CB radios, billboards and temperature controls - are

not significant distractions”.

14 Dr Walton cautions that not finding any evidence is not evidence that there is no

relationship. I agree, and have addressed this in my evidence.

Yannis, G., Papadimitriou, E., Papantoniou, P., & Voulgari, C. (2013). A statistical

analysis of the impact of advertising signs on road safety. International Journal of Injury

Control and Safety Promotion, 20(2), 111–120.

https://doi.org/10.1080/17457300.2012.686042

15 I concur with Dr Walton’s comments regarding this study, that the authors were

unable to find any statistical correlation between road crashes and advertising

signs.

Documents

Before the Independent Commissioner Under the Resource ... · this evidence and I agree to comply with it in presenting evidence at this hearing. The evidence that I give is within