Categorising Travel

The way we once chose to categorise travel may have stultified policy-makers' ambitions, even shaped policy organisations in the past, but right now, a fresh look at travel categories could be really helpful.

Fifty years ago, four stage zonally-based demand modelling started with a 'trip generation' sub-model. This would be based on guidance from Oi & Shuldiner, authors of the key text-book on the subject. Typically, demand would be split into different categories of trip purpose: home-based work (then meaning a commute to or from home), home-based education (journeys to school/college) and for the rest maybe home-based 'other' and 'non-home based'. Refinements could be thrown in: employers' and personal business travel, shopping/day out, visiting friends and relatives, for example. Trip rates would be developed for each purpose using local household-interview data.

Questions of modal choice and trip destination (and hence length) would ideally be tackled in stages 2 and 3. But to keep things simple, trips could be accorded a purpose and a travel mode in the first, trip generation, stage. Simpler still, whole demand models would just focus on an individual travel mode.

This would mean that however attractive any transport mode became, the number of trips by that mode wouldn't change. A handy assumption to keep forecasting for road schemes simple, such as in the Department's 1970s Regional Highway Traffic Modelling exercise.

Of course, over subsequent years, much was to improve, but the essential idea that policies for each of the main travel modes could be set apart from one another seemed to pervade the DfT for decades. It is still reflected in models which show that the nation could build a high-speed rail network based on projections that presume few travellers would switch from other modes to use it. Likewise with LRT systems, despite all the evidence to the contrary. The expression of consumer choice in this synthetic world is all but lost.

So, let's take another classifier of travel: trip length. To ensure that demand models reflect the real world, with its preponderance of short distance journeys, an impedance function had to be invented to weigh against choosing remoter destinations in trip distribution models based on gravity principles. In fairness, this simply reflects the lack of appeal for spending more (time and money) getting there. Yet this function suggests a further unchangeability in our travel behaviour: this is the way we divide up our 'travel budgets' (not that I have ever met anyone who admits to having such a thing). And in practice, people can choose differently, as they might in response to Coronavirus. With working at home say 3 days/week and visiting the office just once, the whole equation of where to live shifts in favour of longer, but less frequent, commutes.

Still, taking the current distribution of trip lengths, what we know is that it is heavily skewed: we make lots of short trips (65% are under 5 miles). But these account for only 18% of total travel mileage (NTS data, 2017). When it comes to policy setting for today's challenge of carbon reduction, it's the mileage that matters but somehow, it's the 2/3rds of trips that are short – amenable to a switch from car use to active travel modes and bus – that are presumed to be where the policy solution lies.

As a Transport for Quality of Life study carried out for Friends of the Earth last year showed, it's the fewer, longer, car trips that generate the most CO₂. True, active travel and bus can provide good alternatives to car for the 37% of our travel made up from trips under 10 miles (measured in trip-miles). For the 38% of car trip-miles that are accumulated by journeys of over 25 miles, it is to (electrified) rail that we must look for a lower carbon alternative. Yes, there is a critical medium distance category of 10-25 miles that accounts for 25 % of all travel, and which is where, realistically, electrically powered cars could be a good solution, alongside other options. The categorisation of most value for policy-makers, it appears, is trip length, not trip purpose.

But for those trying to manage over-stretched road networks and public transport constrained by social distancing, right now, transport planners' travel purpose definitions need an update. A three-way classification suggests itself: essential travel (mainly but not limited to emergency vehicles and "essential workers" journeys to work): important (but not essential) travel; and discretionary. Hard to explain and tough to enforce, no doubt, but needed if we are to avoid severe road network congestion while public transport capacity remains seriously reduced (and for essential travel only).

Right now, travel by car has been given a free rein in England, with Wales and Scotland to follow. The 'very English summer' is set to continue and spread westwards and northwards.