Positive externalities of cars - free seats on public transport

Question

I was in a discussion about externalities surrounding private and public transport.

Someone argued that by taking a car they are creating a positive externality by not taking up a seat on a bus or train. My gut reaction is no that's wrong. But that was because I was focussing on congestion, pollution, accidents.

However their argument seems to be right according to the "dictionary definition" of externality and viewing it in isolation.

What say you, economics community? Are they correct or is there something in the small print of externalities of consumption, when it comes to substitute goods?

Answer 1

In order to answer this question we first have to properly and rigorously define externality. Externalities are often vaguely defined as any effects on third parties but the correct definition of externality is more nuanced.

Mas-Colell Whinston Green (1995) Microeconomic theory states:

"Definition "11.B.1 An externality is present whenever the well-being of a consumer or the production possibilities of a firm are directly affected by the actions of another agent in the economy." .... "When we say 'directly,' we mean to exclude any effects that are mediated by prices.

Mas-Colell Whinston Green (1995) is the most widely used graduate microeconomic text in the world, hence I would consider it authoritative source for a definition of externality. Moreover, most (good) textbooks exclude effects mediated by prices as well (see Mankiw and Taylor (2014) Mankiw (2018) Principles of economics or Cowen & Tabarok Modern Principles of Economics).

Furthermore, your problem is missformulated. Removing congestion from public transport would not be positive externality but reduction in negative externality of consuming public transport (assuming there is any). For example, I do not know of any paper or book that would consider firm that chooses to pollute less (where pollution is negative externality) being positive externality. Reduction in negative externality is reduction in negative externality, not introduction of positive one.

Consequently, in your case the question can be reformulated as:

Is congesting public transport by using it creating negative externality?

Answer is no if we assume that otherwise the market is well functioning and prices are determined by demand and supply. The reason for that is that the negative effect of you congesting public transport by sitting there is already priced in your public transport ticket. So here the external negative effect is already fully internalized in the price of the ticket (assuming no further market failures).

---

Edit:

As pointed out in the comments this answer assumes that public transport actually prices its tickets using market prices. I interpreted the question as a hypothetical, but of course in real life it is empirical question if public transport actually uses market prices or not.

Of course, this is not always the case. In some countries public transport is not private, in some it is public-private partnership etc. Also, there is often difference between travelling locally (within city) and more broadly.

A lot of trains in the Europe actually do price their ticket based on the market demand. For example, you can see on the website Omio (price comparison site), that the bus route between Amsterdam and Frankfurt is priced differently on different days (presumably reflecting among other things congestion). You can also see the same for trains here. Hence for these buses and trains any 3rd party effect would be already priced in and there would not be any externality.

However, of course, empirically there are many cities where public transport prices are fixed, or determined not by market forces. In such cases price would clearly not price in the 3rd party effects save for special case where the prices is fixed at what would be market level.

---

PS: Even if you would not like my reformulation, the above shows that congesting public transport does not create externalities, because whatever the congestion is, it is priced in the the ticket, so by extension claiming you not using it creates positive externality would be a contradiction, given that we already showed in case of public transport the external effects are priced in the ticket.

Answer 2

Effects that follow from rivalry are usually not considered to be economic externalities. From Wikipedia:

In economics, a good is said to be rivalrous or a rival if its consumption by one consumer prevents simultaneous consumption by other consumers, or if consumption by one party reduces the ability of another party to consume it. A good is considered non-rivalrous or non-rival if, for any level of production, the cost of providing it to a marginal (additional) individual is zero.

E.g., a supplier being present on the market increases aggregate supply and may push down the equilibrium price. This is detrimental to other suppliers. However, this is not a negative externality, this is how the market is supposed to work (competition, rivalry); in case of perfect competition the equilibrium allocation is an efficient one.

Similarly, if you buy a piece of chocolate in the store, and that chocolate is no longer available to me, that is also not a negative externality, this is a normal function of the market. The same goes for seats on public transport (or in the theater).

Answer 3

Short answer: yes, it makes sense to call this an externality. (In the literature, these are called congestion externalities).

---

To elaborate:

While one can define ‘externality’ how one likes, the more interesting question is the following: will an efficient number of people choose to take the bus (as opposed to driving)?

To simplify matters, assume (for now) that the only relevant difference between the two options, as far as bystanders as concerned, is that you can take somebody’s seat by taking the bus. In other words, ignore the fact that taking the bus is probably better for the environment. Let's also assume that bus trips have a constant marginal cost, and that bus tickets are priced at this marginal cost. (This would be true, for example, under Bertrand competition with zero fixed costs; or if the buses are public and so priced to ensure zero profits.)

Under these assumptions (plus some technical assumptions I'm going to smuggle in...), one can show that too many people will take the bus.

Here’s why. Ideally, a person takes the bus if and only if the benefit of their taking the bus exceeds the total social cost. However, if we assume that they are selfish, they will take the bus if and only if the benefit of their taking the bus exceeds their private cost, which we've assumed is just the marginal cost of a bus trip. And this can lead to inefficiency.

In case this isn't clear, here's an example:

• Suppose that I get £2 of benefits from taking the bus (this is my 'willingness to pay').
• Suppose that taking the bus costs me £1, which is also the marginal cost of a trip.
• Finally, suppose that if I take the bus, I'll take Bob's seat, which he values at £10.

Under these assumptions, I'll take the bus (the £2 benefit exceeds the £1 cost). But I shouldn't, at least from an efficiency standpoint, since the total social cost (£1 + £10 = £11) exceeds the private benefit (£2).

By the way, this is exactly the reason why negative externalities lead to overproduction in general. (If you consult a textbook, you’ll find more formal versions of the argument above which explain the precise assumptions you need to make in order to make the argument work.) So whether or not you want to call this an 'externality', it functions just like an externality in the basic microeconomic model. To my mind, this justifies calling it an externality.

Of course, you'll also want to take account of pollution in your model. That's fine -- it can be handled in exactly the same framework. The only complication is that now you have an externality in the opposite direction (and so acknowledging this externality suggests that too few might take the bus). To see which externality is larger, you'll need to calculate which leads to larger (monetised) losses, which in turn requires summing willingness to avoid estimates across the third parties involved.

Edit: I want to emphasise that this problem does not arise if congestion costs are appropriately factored into market prices. The point of the story above is to show how they do not need to be factored into market prices (e.g. under perfect competition with constant marginal costs). It is this which can lead to the externality problem.