Transport as Utility

An earlier post introduced the MUCH Framework. The idea is that depending on the nature of good, it might be a M-type (Market), U-type (Utility), C-type (City) or H-type (Hub). That alone doesn’t imply whether the pubic or private sector should provide the good, though it does make some suggestions about whether the good requires public oversight to maximize welfare.

M-type goods, if they are excludable, are considered private goods. If they are not-excludable they are congesting or common pool goods.

U-type goods when excludable are club goods, but when they are not, they are considered public goods.

C-type and H-type goods are not only not rivalrous, more users make them more valuable. As discussed in a previous post, this situation has been called anti-rival.  Thus the traditional classification used below may be inadequate.


The economic definitions of goods discussed in a previous post, depending on excludability and rivalry are important because the case of transportation is quite complex and does not lend itself to an obvious answer of how it should be organized. Roads can be public (rural highways), private (limited access toll roads), club (the local streets in your subdivision), or congesting goods (urban arterials).

By the same token, transit may also be functionally public (a fare free transit system), private (a commuter bus service), club (a school bus), or congesting goods (a crowded campus shuttle).

(One could argue that any of these are potentially excludable if the service provider wanted to, but there are costs to charging, not simply in counting the fares, but also in the delays that charging imposes on travelers, especially when it is done on-board.)

This classification needs to correlated with the notion of natural monopoly – a service for which there are large economies of scale so that competition results in higher rather than lower costs. In that case, the fixed costs of providing the service dominate the variable costs, and so it makes sense for there to be one provider, so that the high fixed costs of infrastructure aren’t duplicated. The incumbent firm in a sector with high fixed costs rarely sees direct rivals. No second company will lay wires to compete with the incumbent electric utility or phone company. That is not to say there aren’t indirect rivals. Home solar and natural gas compete to some extent against the power company. Cable TV and satellite also compete.

Some monopolies are excludable and rivalrous. Electricity distribution comes to mind. An unregulated electric company could decide not to serve me unless I paid, and certainly any electrons flowing through my appliances will not flow simultaneously through someone else’s. Some monopolies are neither excludable nor rivalrous – national defense comes to mind. Some are rivalrous but not excludable (storm sewers). The rivalry though is often time limited, and off-peak there may be plenty of capacity (in fact, there must be, otherwise the excess storm water at the peak could not flow through the system). Some are excludable but essentially not rivalrous (cable TV).

In general roads have some aspect of natural monopoly, some more so than others. (Our agent based model of autonomous roads shows the complexity of the situation.) While city streets may be close parallel routes for travelers, there are usually a limited number of bridges over, or tunnels under, a body of water, creating a natural bottleneck. Expensive transit facilities (subways) also have strong monopoly characteristics.

Other services have monopoly by dint of regulation (I am not legally allowed to open up a new bus company (or dollar van), even if I wanted to, in most of the US.) These monopoly-preserving regulations have the nominal purpose of serving the public interest. And certainly there is an economy of scale to some extent, and advantages in ensuring that bus frequencies are maximally spread out across the hour, rather than clustered tightly at a single point, which might happen if there were two bus services under separate ownership on the same route.

And then there are network externalities – anti-rival goods where the value increases the more people on the same network. Transit is more valuable if more people use transit (frequencies and spatial coverage increases). This does not require common ownership or management of buses at some level, but does require some organization and similarity in the customer facing aspect. In London, buses are privately managed under public franchise agreements.

Roads as Utility

Roads are mostly a U-type good, with scale economies (except when it is congested – but not most of the time on most roads) but network diseconomies (once the network is mostly built out). In the US their anti-excludable aspects, which may have been important when cars were new in the early 20th century, are probably not relevant today.

This implies roads are a utility, which is  a natural monopoly (no one will go and build a competing road network in most places. In places lacking access or where congestion is a major feature, a competing road may work, but again, this is the exception rather than the rule).

This also implies that rates should be regulated somehow. However to be self-sustaining, rates should be higher than the cost of provision. For roads in the US, the rates have been assessed as motor fuel taxes since 1919 in Oregon, as well as tolls on turnpikes. These rates have been sufficient to cover the direct costs of operating and maintaining the backbone network, but have not been applied to the local or last mile networks. This is largely because of government ownership and access to general tax revenue. It is important to recognize that states vary in the nature of their DOTs. Some states manage local roads, others devolve that to counties. The last mile in other utilities is funded much the same as the backbone, with direct charges on users.

Examples in the US include electricity, natural gas, water and sewer, wireline telephone service, cable TV service. Some have better reputations than others. They are in general profitable and outages are scarce. They are also not terribly innovative.

In the post Minnesota Mobility, the idea of how a Roads Utility might work was illustrated. Rick Geddes has another idea of how private capital might be involved in maintaining roads, with private investment paying dividends to the public (and gaining revenue from road users).

Integration from last mile is an important operational question. Drivers don’t know who owns the road they are on, yet how they are funded varies (typically: local roads – property tax, state roads – gas tax). The advantage in principle of local control is that local officials know better local needs. The disadvantage is they lose economies of scale and efficient funding mechanisms. Counties are not usually enabled to implement user fees on roads, and if they did there with current technologies would be a major problem since gas taxes as user fees can be avoided by traveling somewhere else to fill up the tank, leading to a race to the bottom.

My water is metered, my wastewater is not, yet I get a bill from the local (government-owned) water and sewer utility telling me how much I owe for each service. If I consume less water I get billed less. If I produce less wastewater, bully on me, but I receive no discount. Instead of relying on a wastewater meter, they first assume what comes in, must go out. Second, they measure my land area, as a surrogate for storm-water runoff. Credits are available for environmentally good behavior such as rain gardens.

A Transportation Utility Fee could work similarly.

Transportation utility fees are a financing mechanism for transportation that treats the network as a utility and bills properties in proportion to their use, rather than their value as with the property tax. This connects the costs of maintaining the infrastructure more directly to the benefits received from mobility and access to the system. The fees are based on trips generated and vary with land use. 

Transit as Utility

Does transit fit the utility model? Fixed route services on exclusive rights-of-way, like rail or busways may. Services that ride on non-exclusive rights-of-way are not necessarily natural monopolies.  They do posses anti-rivalry and anti-excludability in thin markets (relatively low demand). While there is some desirable regulation of stops and stations to avoid poaching dynamics (see Curb Rights), there can also be competitive transit services, especially in thick markets – where demand is strong. The London model shows how private provision under public supervision can provide superior service while driving down costs and increasing ridership. Other examples are less sanguine.

There have been attempts to vertically dis-integrate railroads (e.g. The UK Network Rail, London Underground, Sweden). These attempts have been less successful (to various degrees, the degree varying in part by the author), because rail technologies are much more integrated than vehicles on roads.

Cars and trucks, we hope, have intelligence with the driver. Trains have intelligence in the tracks. Smart cars on dumb roads work. Dumb trains on smart tracks work. Dumb cars on dumb tracks do not. Smart cars on smart tracks can work, but seems to require a lot more coordination.


So the issues are complex, and there is not a universal answer (The Market! The State!). Each institution, along with other mixed structures (The Public Utility! The Trust! The QUANGO! The Regulatory Board!) have value as ways of managing publicly used infrastructure. Sadly our political discourse is not compatible with complexity.