Road Safety News

'Dashboard’ presents casualty data by constituency

Thursday 26th March 2015

A new report and ‘road safety dashboard’ published last week (26 March) by PACTS and Direct Line Car Insurance calculates the casualty rates for residents of each Parliamentary constituency relative to the local population.

The Parliamentary Constituency Road Safety Dashboard, published in the run up to the General Election on 7 May, is intended to make it easier for MPs to see the road casualty picture in their constituency.

Using information from DfT STATS19 between 2008 and 2013, the Dashboard presents information on casualties for residents of the constituency, rather than looking only at collision locations.

The Dashboard also provides an analysis of progress over the last six years, comparing each constituency against the national average.

Separate analyses are provided by casualty severity, major road user casualty groups (pedestrians, motorcyclists and car users) and by age groups.

PACTS says the purpose of the Dashboard is to “provide additional specific local information to stimulate action to improve road safety for all road users”.

David Davies, executive director of PACTS, said: “Once the outcome of the General Election is known, the Government will consider its strategy for road safety.

“MPs have vital roles in promoting road safety – passing legislation, scrutinising the Government’s policies and representing the interests of their constituents. As the democratically elected representative for the constituency, an MP will often be the person to champion local road safety issues and to bring together the many public and private organisations involved.

“Good information is an essential basis for effective road safety actions and the UK has a long established and robust system for compiling road casualty data. This information, however, is rarely presented at Parliamentary constituency level. Indeed, until recently this was not technically possible. Instead, it is shown using local authority boundaries or other bases which rarely coincide with constituency boundaries. This makes it harder for an MP to see the overall picture for their constituency, to track local progress or to compare one constituency with another.

“To address this problem, PACTS commissioned Road Safety Analysis to create a set of online reports – one for each Parliamentary constituency - that present key information in a simple format at constituency level. The result is the Constituency Road Safety Dashboard.

“There will be many reasons why casualty rates and progress are higher or lower in certain constituencies. The figures provided here are not measures of performance. They are intended to highlight differences and to prompt questions about how progress can be made.”


Comment on this story
Report a reader comment

What's your view - comment on this story:

I confirm that I have read and accept the moderation policy and house rules relating to comments posted on this website.
Your comment:
Your name and location:
Your email:

Might be too late for anyone to read this but worth a go!

To my mind there is not much theoretical difference in targeting physical measures at sites where casualties are clustered above an "average amount" compared with targeting educational measures at a specific demographic group if they are involved at a rate above the average or over represented in casualty statistics when compared to other demographic groups. We need to target resources more efficiently than ever before so using MOSAIC to identify the most appropriate way of communicating with people more likely to be injured in a particular way than others makes sense to me. Get to them before they crash/collide/have an accident etc
Nick, Lancashire

Agree (0) | Disagree (0)

It might have been the way I phrased it Jeremy, but when I said "targeting such behaviour on the roads as and when it happens" I didn't mean being reactive (to accidents) I meant being proactive to bad behaviour i.e as a preventative measure and there's no way you could identify these individuals by referring to the demographic of those who have been in accidents. When you say 'select communities for targeting', do you mean in the residential sense or some other?
Hugh Jones, Cheshire

Agree (0) | Disagree (0)

Postcode level origination of the road user involved in a collision does not in itself reveal the individual to be targeted. Nor does Mosaic and I would hope not to have suggested either.

Although you could target at or near household level with the postcode data there are a number of reasons why you would not - arguably two of the most important being that it's (a) more cost effective to target at an aggregated community level and (b) that targeting the individual involved in the collision is wholly reactive (and we already have efficient enforcement and diversionary processes in place that do that). Better then to 'chunk up' the postcode data, select communities for targeting and use Mosaic to help you understand something about what is likely to appeal to that target area before you start.

Targeting the site already occurs – and the approaches (through cluster and route treatments) are complementary to what I’ve been describing, not alternatives. Moreover, those approaches accept the degree of randomness to which you refer and try to account for that in their analysis of the data.

Ironically of course, because we accept that RTCs are, in part, ‘rare and random’ events, simply chasing them and targeting the individuals involved would in itself be a scattergun approach.
Jeremy, Devon

Agree (0) | Disagree (0)


"...getting the right message to the right people in the right way" is absolutely right and was always my motto and was always my preferred 'intervention' anyway. However I don't believe postcode analysis and Mosaic can usefully identify those who need targeting. Once you factor in random chance and circumstance, such analysis goes out of the window and you're back to targeting such behaviour on the roads as and when it happens. Identifying and educating only those individual road users who need it seems to be a more eficient use of resources than the scattergun approach.
Hugh Jones, Cheshire

Agree (1) | Disagree (6)


You’re right, road users to do not become more or less predisposed to collisions because of their postcode. No-one has, I think, suggested that they do. But you need to bear in mind two key operational benefits of knowing from where the where the injured and/or the perpetrator comes.

Firstly, as Bob has intimated below, it offers a guide as to where to put the resource. In fact the motorcycling example that Bob presents is very pertinent for me. Devon has a motorcyclist injury rate that warrants attention. I know where the incidents occur and where possible I treat those routes or sites. But, we are predominantly a rural county flanked on three side by clusters of higher density populations. Our postcode analysis tells us that many riders come out of one of these, but not all three. One of them is within my highway authority boundary, the others are not. It would be easy to assume that if I were to try to support the rider I would go to the city within my boundary – that’s where my responsibilities lie, don’t they? To do so would be a mistake. Knowing where my bikers come from, not just where they have their collisions, is central to targeting the behavioural work.

Secondly, the postcode can tell you a lot about who you are targeting. Postcode level analysis of, for example, Mosaic datasets is now fairly commonplace to help professionals understand who they are trying to reach. That might affect what you appeal to; how you pitch your message; what style of language you use etc. For me, this is an arena we have played with in the past and will be returning to in future - because it offers exciting opportunities for getting the right message to the right people in the right way. And it starts with knowing where they live.
Jeremy, Devon

Agree (8) | Disagree (0)

I have often wondered about this myself, it's a bit of a conundrum.

What if an area has a great problem with motorcyclists and their accident rates. It's obviously identified by the area officers and measures are put into place in order to reduce the death toll. Motorcyclists in that area are offered discounts on Advanced Riding courses and possibly other measures and interventions are explored and possibly activated. Lots of money and time and effort.

But, and its a big but, what if it's not the local motorcyclists that are at the heart of the problem? It's that the neighbouring area has a lot of them and that area has pretty poor motorcycling roads and so they meet up and head to the first area for their pleasure riding. Because they all in the main riding elsewhere they are not recognised as being a problem in the area in which they live. Therefore no action is deemed necessary to improve the riding ability in that area.

That is the conundrum. Which area do you attack or respond with interventions? The one where there are incidents or the one where the perpetrators of those incidents live, or both?
Bob Craven lancs.....Space is Safe Campainer.

Agree (4) | Disagree (0)

Road users do not mysteriously become less, or more accident prone, because of their home address! If we start to believe that, we might as well catagorise casualties by the colour of their hair for all the relevance it has.

Take a hundred drivers/riders in any residential road, anywhere in the country - some will be more likely than others to be involved in an accident somewhere - it could happen outside their house or it might be two hundred miles away - the element of chance plays a big part for the accident prone.
Hugh Jones, Cheshire

Agree (2) | Disagree (8)

In my experience any number of different stakeholders will want their data sliced and diced in a way that makes sense to them and often this means by variations on the theme of ‘local’ – local by parish; by district; by county; highway authority; health board; super output areas etc. Now by Parliamentary Constituency.

I see no issue with this and the dashboards seem to be what dashboards should be – accessible, at-a-glance summaries capable to raising interest and generating discussion.

At this level one would like to think they may spark discussion between constituency offices and Police & Crime Commissioners; LA Chief Executives; Directors of Public Health; Chief Fire Officers and, of course, civil servants within departments (Home Office and DfT specifically) capable of influencing geographical and thematic expenditure.

I don’t see road safety professionals using this for road safety operational purposes – the boundaries are too arbitrary and the data presentation at too high a level. But that’s not what it’s for.

As for presenting the data sets based on the home location of those involved in a collision, that approach is routine in many LHAs which understand that you can treat the highway where the incident occurred but you may also wish to treat the behaviour where you can access the road user you want to influence. In Devon our District level reports and collision site analyses do just that.

MPs are responsible not just for their constituency as a geography but also for the constituency as body of people who may be found to have undesirable propensities to become involved in RTCs. I’d expect any of us wanting to treat a problem at source to take an interest in that.
Jeremy, Devon

Agree (5) | Disagree (0)

Bob, lets flip that round then. If there were 400 motorcycle casualties from collisions all over the country and you found that 40 of them lived in one specific geographical area, would you not want to target that area's population with regard to motorcycle safety education?
Matt Staton, Cambridgeshire

Agree (5) | Disagree (1)

So what we are saying is that of all the accidents around the country within a given time, six casualties live within a postal area of Blackpool and so what? Are MPs going to target that area with waste paper outlining the danger of being away from home?

For my own part with motorcycling as my speciality. If there were 40 accidents involving motorcyclists on a stretch of road and it was established that all the 40 casualties lived in totally different areas, are we to send someone round to see them and warn them not to do it again, or are we going to look at various interventions on the road that they came to grief on? I know where my money is on.
Bob Craven Lancs....Space is Safe Campaigner

Agree (5) | Disagree (3)

As resources are reduced it is even more important to ensure that they are targeted at the most appropriate location or groups of people. For instance if the majority of casualties near to Blackpool Pleasure Beach are not from Blackpool then there may little point in Blackpool Council educating their local residents about the dangers (notwithstanding the need to inform them of the potential increased chances of encountering errant pedestrians in the carriageway!).

Understanding where casualties live is important in targeting resources efficiently and this type of analysis is used by some Local Authorities to achieve that. Some postcode areas are over-represented in the casualty stats and it is only right that we are aware of that and try to do something about it. I understand that algorithms are used to counter the fact a number of casualty postcodes are incomplete but someone else is in a much better position to explain how they work! As with all data it needs presenting accurately and un-biased to ensure a true picture is obtained. Some press releases are better than others!
Nick Hughes. Lancashire

Agree (11) | Disagree (1)

Like Hugh, Bob and Derek I fail to see the point of this analysis-by-constituency, more like a make-work scheme for consultants, it seems to me.

Nor is the Stats19 system robust, much of the data I have on file includes obvious errors like location codes identifying crashes that happen far outside the boundaries of the particular police force if not well out to sea. Errors of that kind should be prevented by the data entry software and are easy to see by sorting the files on Easting and Northing, thus shifting outliers to the top or the bottom.

In any case the idea that individual MP's could find the time to understand the data for their own constituencies well enough to base policies on it, is naive even if those constituency to constituency comparisons had any statistical significance.
Idris Francis Fight Back With Facts Petersfield

Agree (8) | Disagree (12)

I agree with Bob and Derek. This again seems to be analysis for analysis sake without there being anything useful to come of it, apart from confirming what we already know - that people are involved in road accidents in different places. Whichever administrative area or constituency it occured in can only be of academic interest.
Hugh Jones, Cheshire

Agree (7) | Disagree (6)

If what you say is true Bruce, and I have no doubt about its veracity, then can I question just what is gained by such a publication? To know that a Blackpool resident has been involved in a accident in Torquay and what significance this has, or of what interest to the local MP. Can anyone please elucidate?
Bob Craven Lancs .... Space is Safe Campaigner

Agree (9) | Disagree (8)

Never has there been so much information available to mis-inform and confuse. So a driver from London has an accident with a driver in Blackpool, and the data records an accident to a resident of London and an accident to a resident in Blackpool. And this is of what use? We have a pre-election 'silly season'.
Derek Reynolds, Salop.

Agree (7) | Disagree (9)

The report is based on residency, not on collision location - these figures refer to injured residents of Blackpool South involved in a collision on any UK road. Any tourists injured in Blackpool would be counted under their own home constituency. Obviously traffic conditions in constituencies do affect residents - but visitor casualties have no bearing on the results.
Bruce Walton

Agree (7) | Disagree (0)

Can I ask, what would happen if the area was a holiday resort? Would that not effect the local accident stats as opposed to somewhere that had a more sedentary population? I see my home Blackpool South has apparently a bad position in two tables but it also includes the Pleasure Beach and is particularly busy during summer months which could account for many of the accidents reported. Then there is the illuminations in the darker nights with massive traffic queues and pedestrians crossing.
Bob Craven Lancs....Space is Safe Campaigner

Agree (3) | Disagree (0)

The quote by Mr Davies does say "There will be many reasons why casualty rates and progress are higher or lower in certain constituencies" and you added that the rates of under-reporting vary according to the road users involved anyway, so with any presentation of road casualty analysis where comparisons are expected to be made, that particular caveat should at least be mentioned surely?
Hugh Jones, Cheshire

Agree (5) | Disagree (1)

If you read the report it will tell you exactly how the comparisons were made and why they don't use stark figures about collisions in an area.

Under-reporting is a know phenomenon which happens throughout the country with rates varying according to the road users involved. The quality of data received is very good, far in advance of just about every other nation in the world and I think we should use all available evidence to help inform our opinions and activities.
Richard Owen, Banbury

Agree (13) | Disagree (2)

Shouldn't the MPs at the same time, also be made aware of the short-comings of official accident statistics, in that due to reputed under-reporting of road accidents generally (according to the Dft anyway), stark figures for supposed casualties in defined areas may not necessarily paint a true picture, especially if they do want to make comparisons?
Hugh Jones, Cheshire

Agree (5) | Disagree (9)

You can find another useful comparison website on road casualties at:

This provides comparisons by local authority and police force.
Rod King, 20's Plenty for Us

Agree (4) | Disagree (1)

I’d like to echo David’s comments – this report and dashboard are designed to engage politicians in local road safety in the run-up to the election and I sincerely hope any debate arising from the publication will result in our politicians taking an interest.

No two constituencies are the same and with this evidence available for the first time MPs will have something solid to inform their work.

I’m looking forward to the first PACTS Members’ meeting after the election with a bumper crop of MPs in attendance!

It’s ironic that PACTS have launched this comparative dashboard the day before the DfT close down their English Authority comparison site.
Richard Owen, Banbury

Agree (7) | Disagree (0)