Cross with care: Evidence-less policy making

Back in July, Transport for London announced it was consulting on the removal of 145 sets of traffic lights across the capital, half of which included lights with pedestrian facilities. All part of the Mayor’s obsession with his mythical goal of ‘smoothing traffic flow’, which increasingly seems to take priority over the needs of every other road user. The outcry from residents, the London Assembly and borough councils is well documented elsewhere. I was alerted to the remarkable nature of some of the proposed removals by the fact that several were within spitting distance of my then central London workplace, involving pedestrian crossings on a busy road popular with tourists, shoppers and school groups. Removing the crossings in question would have led to around 300m between crossing facilities, encouraging pedestrians to dice with often fast moving traffic.

Many of the other proposals looked equally ill-thought through, so I stuck in a Freedom of Information Act request to TfL, asking for their evidence base in choosing the pedestrian crossings earmarked for removal. Specifically I asked for what, to me, seemed a sensible data set that ought to be part of any judgement as to the continued need for a pedestrian crossing – and which would not be particularly onerous to collect:

– the number of times each crossing was used by pedestrians on an average day;

– the distance from the nearest alternative pedestrian crossing;

– average traffic speeds at the crossing location;

– data relating to traffic congestion caused by existence of the crossing.

After it took two months for TfL to respond, I was expecting a pretty hefty evidence base. Alas no.

“I can confirm that we do not hold the information you require.”

So, what were the selections based on?

“These signals were selected following a review of traffic flows and historical accident data and, where the sites include pedestrian facilities, observed pedestrian demand levels.”

Oddly, this last point sounds remarkably like what I asked for, but never mind. You have to question how useful historic accident data is – in many cases the crossings will have been in place for decades and the nature/levels of pedestrian or traffic uses will have been completely altered in that timescale. Nor does it appear that there was any quantitative approach to considering which crossings actually had any serious effect on traffic flows. The lack of an evidence base certainly increases the perception that these crossings have been picked almost at random, hence examples outside schools or tourist attractions being amongst their number. Evidence-less policy making towards an evidence-less policy.

One small bit of good news, however. Apparently, further analysis

“may include the collection of some of the data you have requested.”

Good. But who knows if a completely different – and somewhat more sensible – list of traffic lights for removal would have been drawn up if some of that data had been used in the first place?