Attacking road congestion on multiple fronts is a must

Whenever I attend a conference on future mobility and autonomous cars, I hear experts in the technology space, especially those from telecommunications businesses, say that congestion will be a thing of the past and all our roads will perform at totally safe, free-flow levels with optimised spacing between vehicles, as well as intersection and lane merging controls to ensure free flow performance at all locations along a road¹. Vehicle-to-infrastructure communication will make the city function like a giant computer, with a central operating system that everything flows into.

Speculation?

As appealing as this prediction is, one wonders how much faith can be placed on this speculation. What is likely to occur, however, is that the information obtained, after being compiled and analysed by a central platform, will help maintain the traffic flow and direct traffic intelligently, but impact on levels of congestion in a directionally-unclear manner, and adapt the infrastructure to meet acute requirements. These assertions appear to fail to recognise the implications of evolving car-based business models focused on profits.

The role of government

There is a hint here about the possible role of government and whether user or provider side subsidy may be required (maybe less than currently outlaid on conventional public transport) if it ensures a business model that respects the needs of the city over the needs of the commercial operator.

With a fixed amount of road network capacity, if the popularity of the collaborative and connected economy actually takes hold, with or without vehicles being autonomous (and regardless of the fuel source, although this may impact favourably on the cost of car usage if electricity and hence attract even more ridership), we might expect congestion to continue to exist, albeit as more predictable (i.e., less due to random circumstances; lower uncertain travel time variability), and hence the case for efficient road user charging that accommodates congestion impacts will remain real. This will also be required as a mechanism to pay for the roads (as increased support for hypothecation mounts).

It is almost impossible to think of an argument that would result in fewer car kilometres on the roads under even a shared mobility model (which has the greatest prospects), especially if the claim is to also satisfy pent-up latent mobility demand and round-the-clock commercial imperatives. What might be expected is that intelligent mobility opportunities will buy a number of years of growth in automobility in such a way that it will take pressure off the need to invest in currently planned future road investment (which will a good thing), but will require imposing meaningful road-user charges.

Encouragingly, such reformed road-user charges will no longer be imposed on individuals (where the real sensitivity exists), but on mobility brokers who deliver the shared mobility services. Breaking the nexus between ownership and mobility should also break the resistance to supporting road pricing reform at the individual traveller and voter level.

It’s more than just cars!

In the debate on traffic congestion, we need to stop looking just at cars. What if there was a smart infrastructure that connected road, rail, and other modes of transportation in such a way that all of them worked together to serve the mobility needs of the modern urban population? It is in this context that the congestion debate should be positioned. As McCabe (2017) suggests, MaaS may well be a fraud if it is car-centric, and that if all of the arguments associated with car sharing are currently designed to get more people using cars without protecting the future of public transport, we are likely to see congestion getting worse, not better (after allowing for positive initiatives such as automated roads and reduced vehicle size in platoon settings).

Is public transport the saviour?

What may well be the saviour is ensuring that a public transport contribution to the mix is preserved as a basis of moving large numbers of travellers who prefer it to a car-based offer, and we have efficient road user charging and reformed public funding. If, as is suggested, car use will drop in price as we move to a sharing non ownership society with autonomous electric vehicles, the market will deliver higher levels of demand for such services (exactly what the technology and app developers are hoping for), and thus simple economics without efficient road user charging is likely to deliver congestion growth.

So will we end back at a position where the savings associated with selling a car and moving to a non-ownership model with relatively low costs of usage actually discourage travel by a car mode? (See the Appendix for the author’s actual situation in deciding on whether sharing and non-ownership is attractive.) The central theme of his paper is that even with significant sharing of vehicles that are not privately owned, we are unlikely ever to fully tame congestion through the smart technology initiative (Cars are still cars—geometrically unchanged by technology).

Road use pricing

Road pricing must come to the rescue as well as a new public transport funding model by government designed to switch control of subsidy to users and away from providers as a way of creating a competitive market for mobility contracts that support public transport. The greatest disruption may, in time, be created by pricing and funding reform.

The personal economics of vehicle sharing

There may be many car owners where the economics of sharing does not stack up until we have shared autonomous vehicles that have significantly lower user costs (i.e., no driver wages for example). All the while we have a driver in a shared car, the economics are not very attractive for regular car (centric) users. This is the dominant travel mode in most cities in Australia, and hence I anticipate that the exercise below reflects a circumstance of many current car owners and users.

Suppose I have bought a car for cash for $60,000 (which is a typical price for a quality car), and let me assume I keep it for 10 years and simplify the annual capital cost (depreciation) as $6,000, maintenance costs of $1,000 and parking costs of $300, and the car has a residual value of $5,000 after 10 years, which is reasonable.

The annual registration, insurance, maintenance, parking and fuel/toll bill is around $5,000 per annum, giving a total outlay of $11,000 per annum. There is also foregone interest by not investing elsewhere; however the shared vehicle user cost also has an opportunity cost, so it is reasonably financially neutral.

I now sell my Car (assume I only own one car) and enter the shared society and use Uber. Assume I used to drive my car to work 5 days a week and use the car on weekends, and that I intend to continue using a car-based model for these same trips.

If I use Uber for the same trips, then assuming each Uber trip is $20 (which is a basic charge and likely to be greater for some trips, even under a driverless scenario); the weekly cost is around $250 or the annual cost is $13,000, $2,000 greater than the ownership model. If we assume that Uber is used on average 5 days a week (which may be a mix of weekday and weekend), then the annual cost is $9,285 which is less than the ownership cost of $11,000. Clearly the comparison depends on the number of trips to be made under sharing and whether this might be less than under the private car ownership model. Regardless of the specific evidence, this simple exercise is a stark reminder that a sharing model may not be financially attractive to some car centric users. When we consider a view that for many, car ownership is not about cost, this adds another layer of issues to think through in the shared car society.

What is the sharing deal? Not currently very attractive! Some other points worth making are:

1. Even if we assume this possibility, and note the $2,000 difference for the 7 day activity, this is equivalent to 100 Uber trips switching out to PT if that is feasible.
2.  If a car owner was to take full advantage of the car share systems existing and proposed, they could rent out their own car (if they continue to own it) when they are not using it, and further recoup some of the costs of owning a car. The private car may then become nothing more than a rentable asset. (This hardly helps traffic congestion!)
3. Most people we suspect believe that it is less expensive to own a car than to take Uber. This is the issue of perceived costs associated with usage. So for which people would it be a good idea to actually sell their car and take Uber instead? Most people forget to include all the costs in this consideration, since they see a high Uber fee, while not considering the hidden costs of owning a car, which have been included in the exercise above.
4. This is only one market segment, but with dominant interest in the car it seems a very important segment to focus on.
5. Not included are some future possibilities such as selling or renting the garage at a private residence. Selling garages is quite common in London. One in central London recently sold for over $1,000,000.
6. One questions whether a household might sell the garage at home; however when deciding on a future house purchase, they may avoid having a garage, or may include one but rent it out just like a sublet (like airBnB but airG)? This may impact, however, the future design and scale of residential dwellings and lead to reduced purchase (and rental) prices.

Footnotes

1. The most recent conference I attended heard from the Chief Scientist of Telstra (Australia) who claimed that we will soon have 5G communications controlling autonomous vehicles that will be the basis of operating our road network with all communication built into the vehicles (no modification required to roads), resulting in the complete elimination of traffic congestion.