If ever there was a policy driven by sentiment rather than facts, it's the idea that 20mph limits are a 'good thing'. My consultation paper on Haringey's proposed 20mph limit is reproduced below

CONSULTATION RESPONSE ON THE PROPOSAL FOR A 20 MPH SPEED LIMIT ACROSS THE BOROUGH

ABSTRACT

Anticipating that most consultation responses are likely to be qualitative and subjective, I have sought here to look at the proposal more quantitatively and objectively.

To help in the subsequent analysis I have prepared a short digest of salient statistics.

· Looking first at the current road traffic accident rate in Haringey, there is just over one adult pedestrian/cyclist fatality every year and roughly one child pedestrian/cyclist death every three years.

· Looking second at the quantity of traffic in Haringey, the total traffic is 650 million vehicle km annually.

It is likely that the public perception of road traffic casualties overstates the actual incidence of serious accidents. There is one pedestrian or cyclist killed or seriously injured for every 23.7 million km driven on Haringey’s roads.

Costs and benefits can be estimated for an average speed reduction from 30mph to 20mph. In reality, the speed reduction would not be as great as the 10mph assumed. However, for smaller reductions in speed the costs and benefits would scale with each other, implying that relative costs and benefits would remain similar even if the absolute values did not.

The costs and benefits of a 20mph limit being applied to the 50% of roads that do not currently have a 20 mph limit (apart from red routes managed as part of the TLRN) have been estimated from published data and statistics as follows:

Summation of annual costs and benefits of the proposed scheme

Benefits	£	Costs	£
Road safety impacts	5,132,000	Increased fuel consumption	6,750,000
		Carbon emissions	13,800
		Journey times	9,418,229
Total	5,132,000	Total	16,182,029

At almost £16m the total costs are three times the benefits of £5m. The preliminary analysis here does not support the introduction of a borough-wide 20mph limit as proposed.

An aspect of 20mph limits that is not generally appreciated is that they are NOT environmentally friendly. Road vehicles are less efficient at 20mph than 30mph and emit more carbon dioxide per km and more NOX pollutants. A borough-wide limit would increase pollution and contribute towards global warming.

The analysis supports the need for selectivity, as suggested in RoSPA’s factsheet –‘20 mph areas should initially be prioritised to places where they are most needed, for example, in areas of social deprivation which have high populations, areas which consistently display accident problems or have other issues which a 20mph zone could alleviate, and in residential areas around locations which are common urban destinations.’

I remain concerned that Haringey Council has failed to undertake either an environmental impact assessment or an economic assessment for the proposed scheme. Most significantly, it has not even attempted to use the DfT’s Speed Limit Appraisal Tool. While my analysis may be crude in some respects, my preliminary conclusions point to the need for the borough to undertake its own analysis before deciding to implement an area-wide speed limit. My reading of the existing evidence from the DfT and elsewhere is that a targeted use of 20 mph zones, with appropriate traffic calming, would be a better use of resources and provide greater benefits at lower costs than area-wide 20 mph limits with no associated traffic calming or enforcement. In effect, a borough-wide limit is a blunt instrument to solve a problem that is better addressed through more targeted measures.

If Haringey is serious about reducing road casualties then 20mph zones in selected hotspots will be far more effective. This is the route that Haringey should follow.

THE PROPOSAL

The proposal according to the public consultation document is to adopt a default 20 mph limit across the borough. According to this document, already some 50% of roads have 20 mph limits. The proposal in effect is for the remaining 50% of roads to have a 20 mph limit apart from red routes managed as part of the TLRN.

The proposal is justified by the Council mainly in terms of pedestrian and cyclist safety, particularly the reduction in the numbers of pedestrians fatally or seriously injured in road traffic accidents.

STRUCTURE OF THIS RESPONSE

Anticipating that most consultation responses are likely to be qualitative and subjective, I have sought here to look at the proposal more quantitatively.

I am grateful for the assistance of Mr Greville Percival, Frontline Consultation, London Borough of Haringey, of whom I have sought clarification on a number of points regarding the consultation document.

One such aspect of the proposal I questioned is the existence or otherwise of any additional information on the scheme such as an environmental impact assessment, an economic assessment, or a costing document. I was told that no such documents exist.

BASELINE STATISTICS

To help in the subsequent analysis I have prepared a digest of salient statistics – the current road traffic accident rate in Haringey and the quantity of traffic in Haringey.

The scale of the problem

Table 1 below presents figures for road traffic injuries and fatalities in Haringey

Table 1: Annual road traffic injuries and fatalities in Haringey

	2008	2009	2010	2011	2012
All casualties	743	929	984	915	890
· Killed or seriously injured	80	98	79	78	107
· Killed	3	6	1	4	3

Child casualties	56	90	91	78	72
· Killed or seriously injured	9	13	5	12	16
· Killed	1	1	0	0	0

Source: RAS30038 and RAS30039, Department for Transport [1]

The above figures are for all road traffic injuries and fatalities in Haringey, not just pedestrians. Figures are not published for pedestrians or cyclists in Haringey alone. However another table, RAS30024 breaks down road traffic casualties by age band, road user type and severity for the UK as a whole. Table 2 below is an extract and analysis of RAS30024 which shows the proportions for the UK in 2012.

Table 2: Proportion of pedestrian and pedal cyclist injuries of all road casualties in designated age bands, UK 2012

	All ages	Aged 0 to 15
Pedestrians
· All severities	13%	41%
· Killed	24%	33%
· KSI	24%	68%

Pedal cyclists
· All severities	10%	13%
· Killed	7%	21%
· KSI	13%	14%

Source: Analysis of RAS30024, Department for Transport[2]. The table shows the proportions of all road casualties in the designated category. For example, 13% of all road traffic casualties are pedestrians.

By applying the proportions in Table 2 to the Haringey data in Table 1 it is possible to estimate the number of pedestrian and pedal cycle casualties in Haringey. Figures are given in Table 3 for the mean annual rate of road traffic casualties over the five years 2008-2012.

Table 3: Annual pedestrian and pedal cycle casualties in Haringey, estimated

Pedestrian and pedal cycle casualties in Haringey	Annual mean 2008-2012
All casualties	205.2
· Killed or seriously injured	27.4
· Killed	1.3

Child casualties	41.8
· Killed or seriously injured	5.9
· Killed	0.3

Source: Calculations based on Department for Transport data

I believe that this table puts the problem in context. For example, there is just over one adult pedestrian/cyclist fatality every year and roughly one child pedestrian/cyclist death every three years in Haringey.

Traffic quantification in Haringey

It is important to know just how much traffic there is in Haringey. The traffic flow in Haringey is taken from GLA Economics Working Paper 34. The table below shows that Haringey has 22 million vehicle km per square km. This in effect is a measure of traffic density and speed within the borough.

Figure 1: Traffic flow densities in London boroughs (Source: p37, Mayor’s economic unit[3])

Multiplying this figure by the area of Haringey[4], 29.59 km², we calculate that the total traffic in Haringey is 650 million vehicle km.

Relating this figure to the accident statistics is interesting. There is just one pedestrian or cyclist killed or seriously injured for every 23.7 million km driven on Haringey’s roads.

COSTS AND BENEFITS

With these basic statistics it is possible to start to quantify costs and benefits.

Costs and benefits are estimated for an average speed reduction from 30mph to 20mph. In reality, the speed reduction would not be as great as the 10mph assumed. However, for smaller reductions in speed the costs and benefits would roughly scale with each other, ignoring non-linearities, implying that relative costs and benefits would remain similar even if the absolute values did not.

Road safety impacts

The principal benefit is acknowledged to be pedestrian and cyclist safety. The consultation document claims that ‘The Royal Society for Prevention of Accidents (RoSPA) reports that if a pedestrian is hit by a vehicle travelling at 20mph there is a less than 3% chance that they will be fatally injured, compared to a 20% chance at 30mph.’ This amounts to a reduction in risk of 17%.

In fact, RoSPA notes that this was one of the very first studies (dating back to research published in 1979 by Ashton and Mackay). Indeed, RoSPA goes on to say that a ‘recent review identified the studies which had produced the most reliable modern estimates’. The results from one of these is presented in Figure 2 (reproduced from the RoSPA factsheet[5]). The solid line shows the expected fatality risk. The risk at 30 mph (50km/h) is 8% and at 20 mph (30km/h) is 1.5%, a reduction in risk of 6.5%.

The reason for the difference between the early study and later ones in the UK and elsewhere is attributed by experts to improved medical treatment and improvements in vehicle design driven by EuroNCAP.

Figure 2: Relationship between fatality risk and impact speed (Source: RoSPA[6])

A recent TRRL publication summarises the statistics on fatalities at 30 and 40 mph. Table 4 shows that fatalities at 30mph are approximately 7-9% in four of the five data sets reviewed.

Table 4: Fatality risks at 30 and 40mph

Source: Department for Transport[7]

Just looking at fatality probabilities at the two reference speeds is a poor way of estimating real-life outcomes which are affected by levels of enforcement and compliance, and road conditions. It is better to look at the measured effects of speed limits and zones. In their factsheet[8], RoSPA present various statistics of this sort. Perhaps most relevant, a 2007 review of half of the 20mph zones which had been implemented in London (78 zones) found that they reduced injury accidents by about 42% and fatal or serious accidents by 53%.

There is a difference between a 20mph zone and a 20mph limit. Zones are typically associated with traffic calming and/or enforcement, and so are more effective than area –wide limits. To illustrate this point it is interesting to compare Islington which has a 20mph limit with Haringey which has hitherto had a number of 20mph zones but not an area-wide limit. Table 5 shows the mean annual rate of road traffic casualties for the two boroughs.

Table 5: Comparison between Islington and Haringey of mean annual road traffic casualties

Average annual rate of road casualties 2008-2012	Islington	Haringey
· All casualties	836 ± 110	892 ± 90
· Killed or seriously injured	91 ± 20	88.4 ± 13
· Killed	2.8 ± 1.3	3.4 ± 1.8

Source: RAS30038 https://www.gov.uk/government/statistical-data-sets/ras30-reported-casualties-in-road-accidents#table-ras30038. The error bounds are the standard deviations within the data.

It can be seen that the accident rates for the two boroughs are within the error bounds and are therefore not statistically different. If an area-wide limit had reduced accident rates in the adjacent borough to Haringey then the rates would have been statistically different. The proportion of all road traffic casualties that are fatal is also similar – 0.38% in Haringey and 0.33% in Islington.

The Road Safety Observatory[9] found in one study that child pedestrian injuries were reduced by 70 per cent by 20mph zones.

Though Haringey is proposing a limit rather than a zone, for the purposes of this analysis I shall use the figure of 70% which probably overestimates the safety impact. Also, I have ignored the finding that the fatality risk reduction that applies to cyclists is lower than the reduction that applies to pedestrians.

Unfortunately the next section is slightly complicated. As 50% of roads already have a 20 mph limit, the 70% reduction factor cannot be merely applied to the current Haringey accident statistics. Instead, account must be taken of the existing 20 mph zones which would be expected to have already led to a reduction of this order.

Assuming a pre-20mph accident rate across the borough of X, if 50% of roads already have 20mph limits that reduce the accident rate by 70%, the current accident rate (Y) is (X/2+X/2 x .3) = 0.65X.

Assuming a borough-wide 20mph limit would then reduce the accident rate to 0.3X. The proportionate accident rate relative to current levels would therefore be 0.3/0.65 = 46%. The reduction from current levels would be 54%. The forecast reductions are shown in Table 6 below.

Table 6: Forecast annual reduction in road traffic casualties in Haringey

Pedestrian and pedal cycle casualties in Haringey	Mean 2008-2012 (Y)	Forecast reduction using 54% factor
All casualties	205.2	110.8
· Killed or seriously injured	27.4	14.8
· Killed	1.3	0.7

Child casualties	41.8	22.6
· Killed or seriously injured	5.9	3.2
· Killed	0.3	0.2

Source: Calculated from Table 5, see text for methodology

Quantified benefit of safety impacts

It is normal, though slightly disconcerting, to place monetary values on deaths and injuries when looking at transport safety investments. This reflects the reality that there is a limit to how much people are prepared to pay for safety. The average monetary value of fatal and different injuries is shown in the Table 7 which was produced by by the Department for Transport.

Table 7: Average value of prevention per casualty by severity

Source: Department for Transport[10]

Applying the above values to the forecast reductions, and making adjustments to the data to avoid double counting, we can calculate the monetary benefit of reduced road casualties in Haringey – see Table 8 below.

Table 8: Annual monetary equivalent benefit of pedestrian and pedal cycle road traffic casualties

Pedestrian and pedal cycle casualties in Haringey	Forecast reduction using 54% factor	Value (£)	Totals (£m)
All casualties
· Slight	96.7	14,320	1.385
· Seriously injured	14.1	185,831	2.624
· Killed	0.7	1,653,687	1.123
Total			5.132

This shows that the quantified benefits of improved safety according to the scheme are £5.1m annually.

Fuel costs

The relationship between speed and fuel consumption are shown in the graph of Figure 3. Cars average 0.11 litres per mile at 28 mph, increasing to 0.13 litres per mile at 22 mph, a difference of 0.02 litres per mile.

However, their relevance to a 20mph speed limit on urban residential roads needs qualifying in two respects. Firstly most of these roads are rarely congested and therefore the stop/start impact from congestion will be much smaller then implied by the curves. Secondly urban road networks have much more frequent intersections than the main roads to which the curves relate - and these require slowing/speeding negotiation at almost any speed above say 10 mph and the slower the ambient speed the less severe these slowing speeding manoeuvres will be. For these reasons the emissions and fuel consumption effects of a 20mph speed limit are likely to be significantly less than a simple interpolation from these curves would indicate[11].

Figure 3: Relationship between speed and fuel consumption (Source: RAC Foundation[12])

I do not have access to data to quantify the degree to which the difference of 0.02 litres should be downgraded but for the purposes of this analysis I shall use a further factor of 50%, ie the difference in fuel consumption is adjusted downwards to 0.01 litres/mile.

Assuming that 50% of the borough’s mileage is already in 20mph zones, the total mileage calculated above needs to be multiplied by 50%. Then the total increase in fuel consumption would be 650 vehicle km x 0.5 x 8/5 (to convert to miles) x 0.01 litres = 5.2 million litres of fuel. At the current price of £1.30 per litre the total cost would be £6.75m.

Carbon impacts

The emissions of carbon dioxide per km are shown in Figure 4.

At 28 mph the emission of CO₂ is 140 g/km but at the lower speed of 22 mph it rises to 160 g/km, a difference of 20 g/km.

Using the methodology previously described, I have applied a 50% reduction to take account of congestion. The annual increase in CO₂ would therefore be

650 million x 0.5 x 10g /1000000 (to convert to tonnes) = 3,250 tonnes CO₂.

Using a typical price for carbon trading credits of 5 euro/tonne, the carbon cost would be 3,250 x 5 x 0.85 = £13,800 annually.

Figure 4: Pollutant emissions as a function of speed (Source: RAC Foundation [13])

NOX concentrations

Haringey has a serious problem with air quality[14] - with 7 monitoring sites where the NO₂ level is measured and exceeds the annual objective of 40µg/m³.

The emissions of NOX per km are shown by the purple line in in Figure 4. It is notable that NOX emissions increase dramatically below 9 mph. Such speeds are typically necessitated when vehicles have to repeatedly start and stop. This in turn is caused by congestion. As average speeds would fall under the proposals, congestion would inevitably increase. We would expect more stopping and starting, and therefore higher NOX emissions.

Accurate modelling is impossible without further data, however NOX emissions would certainly increase, taking Haringey further from its NOX targets.

Economic losses due to travel time increases

If vehicles travel more slowly they take longer to complete their journeys. People will have to set out earlier to arrive at the same time as previously. This has an economic cost and the Department for Transport has given guidance on the valuation of increased journey times.

According to the DfT publication Values of Time and Vehicle Operating Costs[15], the market price value of time for an average vehicle is £13.91 per hour, 2010 prices and values.

The calculation of increased journey times and their economic value is shown in Table 9 below. Each row shows a successive stage in the calculation.

Table 9: Calculation of increased journey times and their annual economic value

Quantity	Value	Units
Vehicle km	650,000,000	km
Adjust vehicle km to account for 50% of borough already being 20mph	325,000,000	km
Time at 30 mph	6,770,833	hours
Time at 20 mph	10,156,250	hours
Difference	3,385,417	hours
Adjustment factor to take account of other effects on journey time which are not traceable to speed limits (eg waiting timews at junctions)	0.20
Adjusted difference in total journey times	677,083	hours
Monetary value per hour	13.91	£
Total monetary value of increased journey times	9,418,229	£

To summarise, drivers and their passengers in Haringey would spend an extra 677,000 hours travelling and at the lowest value per hour the economic loss would be £9.4m.

POLICY CONTEXT

In researching this consultation response I spoke to the Urban Speed Limits section of the Department for Transport. I was referred to the Department for Transport Circular 01/2013 ‘SETTING LOCAL SPEED LIMITS’ [16].

The DfT told me that they were “lukewarm about 20 mph zones” [17]. This is because in practice 20mph zones without calming or enforcement tend not to reduce speeds very much. Quoting from Circular 01/2013:

· (para 95) ‘Research into signed-only 20 mph speed limits shows that they generally lead to only small reductions in traffic speeds. Signed-only 20 mph speed limits are therefore most appropriate for areas where vehicle speeds are already low.’

· (para 96) ‘The early evidence suggests that it is likely that some speed and casualty reductions have taken place and this is consistent with previous research that has indicated that 20 mph limits without traffic calming reduce mean speeds by about 1 mph on average.’

The introduction to Circular 01/2013: states that ‘Speed limits should be evidence-led and self-explaining and seek to reinforce people's assessment of what is a safe speed to travel. They should encourage self-compliance’.

Chapter 5 of Circular 01/2013 describes a new Speed Limit Appraisal Tool to help local authorities assess the full costs and benefits of any proposed schemes and help make evidence-based decisions to introduce local speeds that reflect the needs of all road users. The tool is available at https://www.gov.uk/government/publications/speed-limit-appraisal-tool and local authorities are invited, though not required, to use it. Its use is free of charge and is not restricted to local authorities. The tool has been designed to enable local highway authority officers and other professionals to: (among others) appraise changes in speed limits to 20mph, 30mph, 40mph, 50mph, 60mph and, on dual carriageways, 70mph.

It is regrettable that Haringey has evidently not chosen to use the speed appraisal tool, nor even attempt an evidence-based approach to this topic.

CONCLUSIONS

Quantitative cost benefit conclusions

The summated annual costs and benefits are shown in Table 10.

Table 10: Summation of annual costs and benefits of the proposed scheme

Benefits	£	Costs	£
Road safety impacts	5,132,000	Increased fuel consumption	6,750,000
		Carbon emissions	13,800
		Journey times	9,418,229
Total	5,132,000	Total	16,182,029

At £16m the total costs would appear to be more than three times the benefits of £5m. The preliminary analysis here does not support the introduction of a borough-wide 20mph limit as proposed.

The main disbenefit is in journey times. People would have to leave earlier to make their journeys. This will be true whether those journeys are by private car or bus. Many of Haringey’s buses, (including the 34, 143 and W7 near me) travel along 30mph roads which are not red routes. Deliveries to shops and businesses will take longer and in time will be more expensive. Ultimately the cost of slower transport will be borne by consumers.

Vehicles tend to be less efficient at 20 mph than 30mph. More fuel will be used and there will be concomitantly higher emissions of CO₂ and other pollutants.

Though relatively detailed, the above analysis is admittedly crude in many respects. For example I have been unable to quantify the increased NOX emissions from slower moving vehicles due to the lack of empirical data. There are health effects of NOX and particulates which are not captured in my analysis.

In general I have given the ‘pro 20 mph’ side of the argument the benefit of the doubt in this paper. I have used estimation parameters that overstate benefits and understate costs.

It is likely that if Haringey had used the DfT Speed Limit Appraisal Tool then the costs and benefits could have been modelled more accurately.

Qualitative conclusion

RoSPA’s conclusion[18] is that

‘Local Authorities should take advantage of opportunities to introduce [20mph zones and limits] where they are needed. 20 mph areas should initially be prioritised to places where they are most needed, for example, in areas of social deprivation which have high populations, areas which consistently display accident problems or have other issues which a 20mph zone could alleviate, and in residential areas around locations which are common urban destinations. The need for 20mph zones can be examined when developing safer routes to school.’

I support RoSPA’s conclusion. Targeted local speed limits where accident statistics indicate the need for a lower limit (such as near schools) would be appropriate to Haringey. In this regard however I note that there are already 20 mph limits in place in many of these areas.

An aspect of 20mph limits that is not generally appreciated is that they are NOT environment-friendly. Road vehicles are less efficient at 20mph than 30mph and emit more carbon dioxide per km and more NOX pollutants. A borough-wide limit would increase pollution and contribute towards global warming.

The DfT is also more positive about 20 mph zones[19].:

(para 89) ‘20 mph zones are very effective at reducing collisions and injuries. Research in 1996 showed that overall average annual collision frequency could fall by around 60%, and the number of collisions involving injury to children could be reduced by up to two-thirds. Zones may also bring further benefits, such as a modal shift towards more walking and cycling and overall reductions in traffic flow, where research has shown a reduction by over a quarter.’

The need for an evidence-based approach

I remain concerned that Haringey Council has failed to undertake either an environmental impact assessment or an economic assessment for the proposed scheme. Most significantly, it has not even attempted to use the DfT’s Speed Limit Appraisal Tool. While my analysis may be crude in some respects, my preliminary conclusions point to the need for the borough to undertake its own analysis before deciding to implement an area-wide speed limit. My reading of the existing evidence from the DfT and elsewhere is that a targeted use of zones, with appropriate traffic calming, would be a better use of resources and provide greater benefits at lower costs than area-wide limits with no associated traffic calming or enforcement.

If Haringey is serious about reducing road casualties then this is the route it should follow.

From Jeremy Klein BSc MBA PhD, 28 October 2013