This was published as Transportation Diversity as Insurance  on the Strong Towns blog on January 29, 2016

Diversity is insurance against risks. Diversification of investments is often suggested in personal finance. The claim is diversification reduces your potential downside loss. (Logically, it also reduces your upside gain.) The appropriate degree of diversification depends on your tolerance for risk.

More generally, there is a cost to insurance. It may turn out the risks you were insuring for did not happen. You might buy fire insurance, and never face a fire. In that event, the insurance premiums were money you could have spent elsewhere. For others – who lose homes, they get back more than they put in. On average, with a mutual insurance system, there is a slight loss due to administrative overhead, but the cost is worth the benefit if the large losses due to a fire, multiplied by their probability, are worse than the small but certain losses of the insurance premium. The probabilities of various events are well understood by insurers, and large risks are pooled, and then pooled again in a re-insurance market.

In transportation, we often talk about resilience. From a resilience perspective, diversity is a good thing. If one kind of thing fails, perhaps the other kinds of things will not.

Path diversity on networks has recognized and measurable value.  If one link fails (a bridge collapse, a sinkhole, a major crash, etc.) and is closed, others can pick up the burden. There may be stresses on the system which show up as congestion, but at least connectivity can be maintained.

Modal or technological diversity is useful. If the bus system’s drivers go out on strike, people can walk. If gasoline becomes expensive, people can take a train. If all cars are recalled because people are suddenly horrified they kill 35000 people a year, travelers can ride a bike. Again, this is limited by the amount of available capacity in the alternatives, and in practice most people cannot take the train in the US, at least not tomorrow, but it provides some type of technological redundancy.


So the question in transportation is whether network and modal insurance policies are worthwhile. We will never know for sure until after the fact. Yet it is not hard to imagine the kinds of risks that occur. We see link failures regularly, so know that path diversity is good. We see that link failures are worse on more fragile systems like limited access freeways and subways, and so the penalty we pay for faster average speed is lower reliability. We have seen technology failures, from the (policy-imposed) “oil shortages” of the 1970s and price shocks of the 2000s, to transit strikes, to vehicle recalls. One could imagine that Lithium, for example, will be the new scarce element in a world of battery-electric vehicles. If not that, something else. Scarcity is not eliminated just because oil is replaced.

All of which is to say don’t put all your eggs in one basket. We need different types of redundancy, choices in transportation modes and paths, multiple possible destinations that serve similar types of purposes, the ability to do things in person or remotely. The more alternatives we have, the less vulnerable we are to any particular type of failure. But we also cannot economically justify empty buses or empty roads because something might happen.

Slowly, the surface transportation sector is re-recognizing the advantages of connected streets networks, certainly for pedestrians and bicycles, but also for motorized vehicles. The advantages of multiple technologies are less clear. In the extreme, school buses can be used for evacuation, pools of buses can be moved from one city to another in the emergency (unless you left them to flood), and so on (though train cars are not as compatible between cities as they should be). We are slowly recognizing that the empty seats in a car on most trips are redundancy that can be exploited, with services like UberPool and LyftLine.

These are types of conversations that are seldom heard in surface transportation, but are common in aviation. When a flight is cancelled, an airline tries to route its passengers. It has a redundant network. The alternative paths (flights) are seldom as convenient, and may require a transfer — or worse for the airline — booking on a competitor, and will take longer, and passengers will grumble, but eventually the passenger gets to her destination. The airline will not generally rebook you on a bus or train (though I have been bused to a final destination when an airport was closed due to fog).

We have a lot of resilience now. There are low cost solutions to get more. These are conversations that Strong Towns should be having, thinking about investments that are flexible and adaptable, rather than brittle, and solutions that work under a range of conditions, not just a single set of assumptions.