DRAFT. What is here at first is a copied paper - but I've lost the reference, which may be [3] below.
[2] https://www.peak-urban.org/publications/paradox-traffic-and-extra-cars-city-collective-behaviour
[3] https://royalsocietypublishing.org/doi/10.1098/rsos.201808
There we are, lots of us wishing to use the transport system at more or less the same time. In this wonderful modern world we have apps that help us pick a route that will be optimised—as far as the data goes—for our particular journey. That is, for our particular needs and chosen mode of travel.
Yet, if we all choose the 'best' option, we get in each others' way and the result is we all take longer. Solutions to this sort of issue result in us taking different options. These include but are not limited to: changing our choice of route, or of transport or when we attempt this travel. Factors we take as important are the length of time the travel takes, and some research shows that we switch to and from public transport under pretty limited conditions. Generally, we prefer to travel on personal transport though it is recognised that money is a significant limiting factor, from having a vehicle at all to choosing to use it. Deterrents to use —it was often said to me in China that the state would like you to own a car, but not to use it—are parking difficulty and cost, access issues (such as happens in 'clean' cities such as London), traffic jams and the general hassle of driving. balancing this are deterrent to use of public transport, the problems of luggage in any form, the physical demands, the end effects (to and from the system), the environment of public transport (noise, discomfort, perceptions of cleanliness, even mixing with others) and cost.
Often there is spare capacity on public transport (though many commuters would disagree, travelling at peak). It amazes me how little data has been collected on how we travel, given the fuss we make about it. Look up mode share data or modal share; I found two in the UK at [4]. The Mayor of London’s Transport Strategy 2018 set a target of 80% of all trips in the city to be made by walking, cycling or public transit by 2041, up from 65% at the time of the strategy’s release.2 This target was informed by analysis that established how many current car trips in London might feasibly be made by walking or cycling.
But persuading people to use alternatives, even walking and cycling requires the provision of convenient, efficient, affordable and appealing alternatives that travellers will choose to take. To achieve this, cities need to give space to – and prioritise – alternative forms of transport on their roads; invest in alternative transport infrastructure; ensure multi-modal network connectivity; and introduce schemes and incentives such as cycle hire and smart ticketing to make them an attractive first choice. [4]
The yellow/green diagram from [3] shows how demand affects travel time. It is assumed that all commutes are equal and an hour, that using a car costs less than the public system when traffic is low. The assumption further is that there are a million commuters making choices. Don't reject the diagram yet. Instead, read the chart as showing the point at which car density makes the journey equal to the public (bus, competing with cars directly) system cost and so when car numbers go beyond this one is (in blue) paying an opportunity cost for the use of a car over alternatives. Car sharing would make an immediate improvement, as the paper makes clear several times.
The appropriate chart notes are: Figure 1. Commuting time (vertical axis) as a function of the number of users (horizontal axis). In the top left panel, the commuting time of car users are divided into (i) a baseline cost, (ii) a cost induced by other drivers, which increases with more car users (traffic), and (iii) a cost induced by public transport users, which increases with more public transport users (so it decreases with more car users). The top right panel represents the cost of public transport users, and it is analogous to the car costs. The horizontal axis from the top two panels has opposite directions, as one measures car users and the other public transport users. Both cost models, combined with the same horizontal axis, form the model in the bottom panel. Considering a population of N = 1 000 000 individuals, the average cost, as a function of the number of car users, is the red parabola. The cost at equilibrium, 𝜇∘, is observed when both modes’ travel cost is the same. However, the minimum attainable cost, 𝜇⋆, lower than the equilibrium cost. The system produces 𝜂(𝐶∘)=200000 extra cars in the city and a social inefficiency of 𝜒(𝐶∘)=0.9
The model emphasises the need for collective behaviour using a parsimonious approach with minimal assumptions that are feasible [121,122]. Even though each journey is different, we are trying to observe the emergence of a collective decision-making process to generalise from such individual aspects to a general, city-level observation of public transport use. Similarly, from an economics point of view, lower prices of a particular good are assumed to increase demand, although many individual aspects are ignored, favouring a collective observation. With a simple model to account for the costs of commuting induced by other users measured in time and, assuming a dynamic in which individuals chose the least costly method, we grouped the ‘desirable’ modes of transportation into a single group and measured the excess number of car users that would be observed. Under different scenarios, we observe a very inefficient social system. There could be more than 40% extra drivers in the city (and 40% extra cars) and commute 20% or more extra time than needed under the best social scenario. They pay a high price for congestion and the rush hour. This is an emergent collective pattern, where all individuals try to minimise their own travel time, but that might result in very inefficient social scenarios. The limit case is actually a paradoxical result. All individuals drive their own car, and all minimise their commuting time, but collectively arrive at the worst-case scenario, in which the average cost is the highest.
It seems to me that the differential between the costs of private and public can be nudged to encourage people out of cars, probably by significant improvement of the public transit system, and by making it markedly less costly than the private alternatives. That would encourage car sharing, too, which can be enhanced by incentives such as access to bus lanes at peak periods. Car sharing can be forcibly encouraged by congestion charging and by banning some cars (e.g., odd numbers) on certain days of the week. Most of all, though, the transit system has to become an attractive alternative. The car is seen as more secure, more comfortable and faster. But then I think working from home with maybe one day a week in the office (for those to whom that applies) has a lot to recommend it. That would not reduce the commute problem to 20% but it might well reduce the traffic demand to less than half its previous value.
Other snippets from [3]: ..changing mobility behaviours is challenging and might require interventions such as congestion charges, dynamic tolling, parking bans and removing free parking to reduce single-occupancy vehicle commutes, increase public transport use and promote active mobility [115].
Economics acknowledges congestion as a negative externality that needs to be corrected to maximise social welfare. Solutions in this direction have implemented pricing congestion through road toll as an effective policy [62,63,119,120].
Convincing drivers of avoiding single-occupancy vehicle commutes, using public transport, cycling and walking cannot rely on a single policy but on increasing journey options, accessibility and time efficiency. A city needs fewer and shorter journeys, accessibility so that people can walk, cycling lanes and more public transport, so that the best transport mode for everyone is not to use a car.
Looking at alternatives such as walking and cycling one must always compare these two alternatives with the security, comfort, speed and costs of the car and the available public system. There are a load of induced costs introduced by other users, such as queues, delays, passenger drop-off and pick-up, busy trains, stairs, streets or cycle lanes. Thirdly, the cost that car drivers put on public transport users. My 15-minute walk into town is efficient because I have a route that has minimised the road crossings, for each one is a delay. This is quite a lot quicker than any bus service, mostly due to wait time, however reliable the service might be. Since buses are free for me, I could take any available bus along the route but I have noticed that the time gain would be surprisingly small and with loss of exercise gain.
The several choices often conflict and so there are several classes of cost: (i) the baseline cost (building a metro line, all sorts of infrastructure, having a car and a road route, a bike and a cycle lane, a pedestrian route and suitable coats and shoes); (ii) the running costs being all sorts of maintenance, ticketing, support, staffing — yes, these are relatively trivial for walking and cycling; (iii) the risk costs attached to the lack of security or the balance of risks attached to a choice of transport. Then we have the time-based costs, (paraphrased from [3]) which all apply to the commute, (iv) the baseline cost, the time the journey takes when unencumbered and the whole system is working properly. You might view this as the best possible time; (v) the time costs induced by other users, such as queues, delays, passenger drop-off and pick-up, busy trains, stairs, streets or cycle lanes. Last, (vi) the direct interaction time cost that any group imposes on another, but especially the cost car drivers put on bus users.
Each of these costs can be reduced and each intervention has an impact, not always for the better. A new metro line is expensive and intrusive (this applies to much of infrastructure) especially during construction. Land values around the new infrastructure undergoes change, not all of it good and there is rarely compensation. The new addition can add induced costs [(v) above] to the other modes of transport. It is inherently a mess and unintended consequences often occur. This generally means that small changes allow adjustments to occur as effects are seen, but humans in general seek the solution that best solves their individual problem. Pedestrian crossings or spaces for bicycle parking could reduce the number of cars and increase the social efficiency without increasing the average social cost (as some users will decide to use the public transport instead). Expanding the public system needs to be interconnected with cycles and walkers (and most of the public system users will become pedestrian for parts of their route; there needs to be space for queuing and efficient systems for the reduction of all queuing, always encouraging fewer cars and encouraging more capacity for walkers and cyclists. I am well aware that we would like to keep three categories separated and that we generally barely manage to separate into two categories. The cyclist is always a problem when caused to interact with traffic or with pedestrians.
DJS 20211125
See also the 15-minute city study, essay 290
Looking 20211207 for agreement with my opinion, I found https://www.nature.com/articles/d41586-021-00396-2