Are Predictable Traffic Jams a ‘Tragedy of the Commons?’

Predictable traffic jams or ‘dumb mobility’ is hampering progress towards a better mobility, and will only be reduced when people take greater collective responsibility for mobility outcomes. This article considers whether the road network is a ‘commons’, and if so if dumb mobility represents a ‘tragedy’ of the commons. It suggests that the ‘commons model’ may be a useful framework for thinking about the challenge of getting people to take greater collective responsibility for mobility outcomes.

In a 1968 article in Science called ‘The Tragedy of the Commons’, Garrett Hardin provided a fascinating and very readable insight into the nature of mankind to breed and consume itself into extinction. Mankind depends on certain common resources for survival. A feature of a ‘commons’ is that individuals can consume the common resource without restriction.

Hardin explains that as population expands beyond some point, rational (self-interested) consumption of the common resource by each individual can be expected to result in excess consumption by the whole of the community, thereby destroying the common resource: a tragedy. Hardin called for ‘mutual coercion mutually agreed upon’ as a mechanism for preserving the resource – in effect removing it from being a commons.

In 2009, Elinor Ostrom received the Nobel Prize for Economic Sciences for demonstrating that ordinary people can create the sort of mutual coercion that Hardin spoke of. In her 1990 book, Governing the Commons, Ostrom explains that at the heart of the tragedy of the commons and similar models is the problem of the free rider – the one who gets the benefit of the collective action, but does not contribute to the joint effort.

Is the road network a commons?

The road network is somewhat different to the types of common pool resources that Hardin’s and Ostrom’s work explores. As well as not being a natural resource, a key difference is that road networks are (or were) expandable, so upon finding the shared resource full, more lanes could be constructed – not the case with grazing land, fish, or an aquifer.

Also, putting too much demand on the road network does not necessarily damage it for future use in the way that natural resources suffer from over-use.

However, once the limits of road network expansion are reached (either physically or fiscally) the resources seem more similar. And if the resource is framed as a thoroughfare, with the benefit derived being unimpeded passage on a daily basis, then an argument could be made that its function as an effective thoroughfare during peak periods on certain days of the week can be permanently damaged by predictable traffic jams.

In this way, the road network can be seen as a commons, and dumb mobility can be seen as a tragedy of that commons.

Governing the road network commons

Hardin speaks of inevitability. However Ostrom’s research identified a number of examples where sustainable self-managed shared resources existed, and had existed over long periods of time without becoming tragedies. She analyzed case studies of these situations and proposed a framework for evaluating options for alternative structures in different situations.

She proposed a small change in the theoretical model that explained a workable alternative, and presented a very real example of such a solution in operation. In essence, the users of a common resource (in this case a fishery in Alanya, Turkey) devised a solution that allotted fishing sites to local fishers without the usual methods of privatization or central (read government) regulation.

Ostrom goes on about examples that have worked and ones that have not: “The differences may have to do with factors internal to a given group. The participants may simply have no capacity to communicate with one another, no way to develop trust, and no sense that they must share a common future. Alternatively, powerful individuals who stand to gain from the current situation, while others lose, may block efforts by the less powerful to change the rules of the game. Such groups may need some form of external assistance to break out of the perverse logic of their situation.”

“The differences may also have to do with factors outside the domain of those affected. Some participants do not have the autonomy to change their own institutional structures and are prevented from making constructive changes by external authorities who are indifferent to the perversities of the commons dilemma, or may even stand to gain from it.”

As people sit in traffic, suffering from dumb mobility, to what extent are they participants in a tragedy of the commons in which they have “no capacity to communicate with one another, no way to develop trust (that they will not cheat), and no sense of their common future”?

To what extent are they bound to stay in this situation as a result of ‘powerful individuals or external authorities who stand to gain from the tragedy, or are indifferent to it’?

Further, is it possible in 2015 that the tools are beginning to exist that could help them communicate with one another and develop trust? If so, could a new paradigm of road network management be devised by the users (rather than tolling authorities or central government regulators) that would result in better outcomes?

What do you think of the idea that the road network is a commons? Is dumb mobility a tragedy of that commons? And do you think road users could devise a better management system? Please share your opinions in our comment section.

Please note that this article expresses the opinions of the author and does not reflect the views of Move Forward.



  • drwool
    7. October 2015 at 22:35

    Yes, it makes sense to view the road network as a commons.

    I don’t know about devising a better management system for this commons. It’s hard to imagine what that might be. But a fairly obvious way to alleviate the problem of congestion is making it much easier for individuals to access both historic and real-time information about traffic patterns and alternatives. Our smartphones can already map routes for us. Imagine an app that can map routes while taking into account up-to-the-minute data about traffic flow, as well as incorporating smart forecasts about how that flow will change over the next 5, 10, 30 minutes based on pattern matching and historical data. It should also be aware of alternative modes of transportation, including mass transit schedules and instantaneous knowledge of buses & trains running early or late. I should be able to give my smartphone a desired destination and have it use all of this information to suggest the best routes for driving, biking, or walking with estimated transit times for each, along with mass transit options and their corresponding expected arrival times.

    Most people want to get to their destinations as quickly and easily as possible, so simply providing this information would tend to alter peoples’ behavior in ways that would alleviate congestion, without imposing a top-down system of management.

  • Pallavi Reddy
    8. October 2015 at 13:28

    Hi David, thank you for sharing your insights and for taking the time to elaborate. I agree that smart mobility apps could contribute to solving traffic congestion issues, but don’t you think we need more or different approaches to resolve this? What other ways do you think could help?

  • GmanFunkFlex5000
    8. October 2015 at 15:22

    Picture your typical traffic jam. A road is operational until it gets over burdened or some abnormality decreases the amount of lanes, constricting the flow of cars. In my experience, this is only part of what causes traffic jams, and the majority of what makes traffic come to a halt is how people behave in traffic jams (switching lanes, trying to out maneuver each other for an extra 10 ft forward, etc.). This could be one of the first and easiest problems for autonomous driving to fix, taking over the driver seats during these very straightforward bits of driving. This will remove the temptation for drivers to act out (causing more delays) and save the frustration of driving through a traffic jam, all while reducing the time it takes to get through the jam.

  • Paul Minett
    8. October 2015 at 22:40

    I think you describe a great system that would balance out the amount of congestion across the whole road network, and it will work best where the road network includes lots of alternative routes. However, there will come a point at which this process completely saturates the whole network and it comes to a complete stop everywhere. At that time perhaps we will agree that the underlying problem was not a lack of information, but a failure of the management system. Yes, it is hard to imagine what a better management system might look like, because if it was obvious we would already be doing it. Since it is hard to imagine, it also should not be a foregone conclusion that the result would be an imposed top-down system.

  • Paul Minett
    8. October 2015 at 23:01

    I agree with you that autonomous driving might keep everything moving a little more smoothly. And with V2I and I2V managing the flows, we should be able to completely fill every last inch of road space. But I disagree with your assertion that it is how people behave in traffic jams is what causes the majority of traffic jams. Once the volume of traffic reaches a certain level, the interactions between vehicles becomes more critical, and you are right that behavior will drive instances of stoppage. I think the underlying cause is the volume of traffic, given the amount of road. If there were fewer vehicles, the behavior would not be such an issue. The question is, could we create a management system that would achieve a persistent reduction in the volume of traffic?

  • Eric Verhulst
    11. October 2015 at 11:13

    I would not say that a road is a “commons”. Price and demand can regulate that, provided the real price is calculated, not a subsidised one. However, the way we use roads is rather dumb. There would be no traffic congestions, or at least a lot less if the space would be used a lot better. The average vehicle occupation is 1.3 and cars take up a lot more space than 1.3 persons (+ safety distances). Pool lanes do not solve that problem. Hence, a simple scheme is to share vehicles/cost/rides. How often do people in the same neighbourhood not travel along the same road to more or less the same destination? If by using e.g. an App, we can bring the average to 2.6, then on average we will need twice as less roads, cars and the cost will drop drastically. Fuel costs pr person will drop by a factor 2 as well. Mobility as a Service is a commons, not the roads. The doom scenario’s forget that we humans always have come up with better solutions for better using the resources.

  • Paul Minett
    12. October 2015 at 20:13

    Thanks Eric. I totally agree, that if we could use our vehicles and roads in a smarter way the result would be much better. What do you think we should be doing to bring about the changes you suggest? So far apps have not made that much difference.

    26. October 2015 at 13:37

    Sorry about length but this is a complex question, so like Caesar’s Gaul, in tres partes divisa est! [Part 1] Whether the roads are ‘commons’ seems a bit academic, but I suppose they could be called ‘renewable commons’ implying a certain maximum rate of delivering utility. So I defer to Ostrom, except that there need not be a ‘free rider’ just some who do not experience the costs they impose on others. Regarding ‘tools to communicate’, a few years ago I was involved in a proposal to formulate an agent-based system to trade road travel ‘slots’ so as to optimise social usage (it wasn’t funded). More recently, Professor Stephen Glaister CBE, director of the RAC Foundation (and by implication arch-motorist) recalled Reuben Smeed’s 50 year old proposal for road use pricing (“always 10 years away”) arguing that it’s time has finally come. These initiatives may well be overtaken by automation in the form of Cooperative Systems and ultimately (2035?) vehicle automation. On the other hand, if people all decide to travel at the same time, and the indications are that home working, virtual presence or shift-working will remain uncompetitive with diurnal office-based employment, plus automated vehicles make it easier rather than harder to get on the road, plus road ‘improvements’ and ‘pinch point’ treatments as planned by Highways England raise effective capacity and travel speeds leading to greater travel distances within the ‘time budget’, then there will be no let up in recurrent congestion. The term ‘recurrent’ reflects that congestion is not completely predictable. Typically, high capacity roads experience flow breakdown at regular ‘seed points’ like merges, but flow breakdown itself is probabilistic and can be triggered by short-term fluctuations in volume or the behaviour of a few drivers. Overall, as A J Miller pointed out back in 1969, a balance of toleration arises between delays, including their variability, and the perceived benefits of travel.

    26. October 2015 at 13:37

    [Part 2] As I understand it the difference between ‘complicated’ and ‘complex’ is that complicated means lots of stuff going on at the same time, while complex means there are interdependencies and subtleties leading to emergent behaviour. The practical difference is that a complicated issue can be addressed by a simple measure whereas a complex issue cannot. I suggest that complexity lies in the dependence of our society on relentless movement and high energy consumption, in the supply chain that feeds it, and in the reflexive relationship between provision and expectations, while the actual daily operation of road traffic is just complicated. About 10 years ago Professor Judith Petts of Birmingham University produced a diagram, reproduced by the Lawton Commission [], which explored the complex relationships between car use and other social and economic effects. However, congestion seems to me a simpler issue. One reason is its extreme sensitivity to volume, even without the catastrophic (in the mathematical sense) flow-breakdown effect. Pricing, metering, ‘intensification’ of living and working in cities, a massive hike in oil prices (tankers might turn into coral reefs!), severely taxing environmental pollution and resource usage, all these could reduce demand. This is why in my paper I cite transportation as an ‘easy’ target for promoting sustainability.

    26. October 2015 at 13:41

    [Part 3] Prediction might also work, although dynamic prediction is limited to around 30 minutes ahead. There is currently a joint HE/RWS project, though I don’t know its status. In principle, given the mass of real-time data, centralised traffic management could predict the risk of congestion, exacerbated by incidents, and apply control through speed limits, diversions or information. Problems with this are threefold. First, it amounts to increasing effective capacity, or at least perceived level of service, which could lead to traffic generation negating the benefit. Second, any effect on travel behaviour is only indirect. Third, any intervention to achieve a social optimum will reduce utility for some individuals, so is likely to be resisted. For example, road pricing and extra taxes would probably be resisted by residents of 4WD commuter villages, whose lifestyle is the result of previous biases in their favour which over the years have disadvantaged and displaced traditional rural inhabitants and poorer people reliant on public transport. Here we potentially run into those ‘powerful individuals …’, or just inertia and risk-averseness. Regarding information, as suggested by one commentator, the problem there is getting information about an uncertain future. Already, the AA, Inrix etc publish ‘heat-maps’ and plots of current hold-ups, but individuals responding dynamically to information could lead to chaos, which is why for example the Police often prefer to hold traffic in a jam on a motorway instead of diverting it onto an unsuitable network. Noting the trends in government policy over the past few decades, I think the simple answer will be what the government may possibly be working towards by its transformation of The Highways Agency into Highways England, namely privatisation of at least major roads. While, as in any such move, the details and mechanisms would be complex, the basic idea is uncomplicated.

  • Pallavi Reddy
    28. October 2015 at 7:57

    Hi Nicholas, thank you for taking the time to elaborate and for sharing your perspective on this topic!
    I am curious to know more information on why you think that transportation is an easy target for promoting sustainability. It would be great if you could share a link to your paper.

  • Paul Minett
    28. October 2015 at 22:53

    The Lawton Commission report is interesting and relevant, thanks for providing the link. I could not find the diagram that you referred to – could you provide a page number reference?

  • Paul Minett
    28. October 2015 at 23:00

    Hi Pallavi. I think this link is for Nicholas’ paper:

  • Paul Minett
    28. October 2015 at 23:13

    Nicholas, I see your point that traffic congestion is simple to resolve – we just need to reduce demand. So I realise that perhaps I didn’t state the moot quite as well as I might have: ‘is reducing [traffic congestion by reducing] demand a complex or complicated issue’ is perhaps what the title should have said. And while the list of interventions and conditions you list ‘could’ reduce demand, my question is intended to drive at our observed failure to create a persistent reduction in demand, with the flow-on effect of reduced congestion, even in the light of a decent list of policy and environmental options.

  • Paul Minett
    28. October 2015 at 23:33

    Thanks Nicholas. Free riders get the benefit of a solution without contributing to the cost of the solution. In the theory for avoiding tragedies of the commons, the users participate in the management of the commons (thereby incurring a cost to themselves in time, effort, or money; depending on the solution). A free rider in this case would be one who manages to avoid contributing to the management, but gets the benefit of the solution – so they not only do not experience the costs they impose on others, but they also avoid participating in the costs of delivering the improvement.
    You say that ‘if people all decide to travel at the same time…..there will be no let up in recurrent congestion’ – as long as the vehicle occupancy rates do not also change. It would be more correct to say that ‘if people all decide to DRIVE at the same time….there will be no let up in recurrent congestion’. Your point really assumes the status quo of dumb mobility – that people continue to travel as drivers in their hordes.
    My reference to the the ‘commons model’ is really intended to explore the possibility of people choosing to travel as passengers more of the time as a contribution to a community led solution to congestion – it is possible that there are other mechanisms that could work such as tradable travel slots, reverse tolls, road pricing, etc.

  • Paul Minett
    28. October 2015 at 23:49

    Thanks Nicholas. You are right that prediction and privatisation are possible policy settings that could be used, but they are fraught, and certainly not uncomplicated.
    I agree with your assessment of prediction. It doesn’t change the underlying behaviours, and in essence just fills the system with as many vehicles as possible. And unlike predictions of the weather, a prediction of congestion can change the outcome. Having heard a prediction of congestion, and finding no congestion (because others have diverted due to the prediction) I might be less and less likely to trust future predictions of congestion.
    It is interesting that you raise privatisation as a solution. In Ostrom’s work Governing the Commons she outlines that there are two traditional approaches to solving the tragedy of the commons: one is central government control, the other is privatisation. Her thesis is that there is a third way, and it is for the users of the commons to manage it themselves. I have seen no evidence that privatisation solves the problem of excess travel demand. Certainly we can see that central government control doesn’t achieve the goal. If we accept that something has to change, should it automatically be towards privatisation, or could we give more than a passing glance to management by the users? Perhaps TRL is the type of place that could carry out a serious investigation into the possibilities?


Convergence of Sharing and Automation:  Need for Proactive Public Policy and Research Understanding By Susan Shaheen and Adam Cohen In recent years, on-demand passenger and courier services – known as Mobility on Demand (MOD) – have grown rapidly due to technology advancements; changing consumer patterns (both mobility and retail consumption); and... View Article