Depending on dodgy figures

Tweet

I’m not an expert on predicting traffic flows, and I won’t pretend for a moment to understand the detail of the modelling which has been used by the civil servants in Whitehall to arrive at its estimates of the likely increase in traffic over the Severn Bridges after the abolition of tolls.  So, it’s not expertise which makes me sceptical about their conclusions, it is, rather, the application of a simple test to the outcome of the calculations – ‘does this result look reasonable?’.  It’s an approach which mathematicians often use, and if the result does not look ‘reasonable’ then it’s often a sign that there’s something wrong – either with the initial data, or with the methodology applied to it.  Treating the model as an unchallengeable ‘black box’ isn’t enough; the fact that ‘the computer says’ something doesn’t make it right.

As I understand the numbers set out here, the conclusion is that in 2022, 24 million vehicles would use the bridges every year without tolls, but only 18 million with tolls, or, put another way, without tolls, traffic will increase by 42% but if tolls only slightly lower than today are retained, it will increase by only 4%.  One corollary of that is that 6 million journeys each year – or 1 in 4 of all potential journeys – will simply not take place at all if they cost a few pounds more than they might otherwise cost.  Now of course it is true that according to classical economics an increase in price results in a decrease in demand, and I don’t doubt that some people or businesses will be deterred from using any stretch of road by a toll.  But it just doesn’t feel at all reasonable to me that such a small (in relative terms) toll makes such a huge difference – I just don’t believe the figures and suspect the validity of the model being applied.  A marginal variation in cost ought to lead only to a marginal variation in traffic flows; and the assumption that cost is what drives the volume of traffic flows is itself somewhat dubious.

There’s another corollary as well.  If tolls of a few pounds are really so effective at deterring people from making journeys, then there is perhaps an easier solution to the problem of the Brynglas tunnels than building another motorway – we can deter a quarter of the potential journeys by imposing an extra toll.  In reality, that doesn’t feel reasonable to me either – I don’t believe that it would have that effect in practice.

Llanelli AM, Lee Waters sees these figures as a way in which the UK Government is trying to dictate transport policy in Wales, by forcing the Welsh Government to build extra capacity – the new M4 link – in order to cope with all the extra traffic.  I’m not sure that they are quite that devious, but I have a lot of sympathy with the underlying assumption that the government’s figures are produced in a way which supports its policy objectives – by which I mean that the assumptions built into the modelling are those likely to produce the preferred outcome.  I disagree with Lee’s suggestion, however, that the solution is to retain tolls, albeit at a lower level, in order to obviate the necessity for the new M4 link.  Not believing that the removal of tolls will directly lead to a given increase means that, logically, I also don’t believe that retaining them would prevent that level of increase.

If the problem is that dodgy figures are being used to justify a particular policy, the solution is not to use those same dodgy figures to justify an unpalatable alternative – it is to challenge the dodgy figures themselves and the basis on which they have been prepared.