Unit 8 Economic dynamics: Financial and environmental crises

8.13 Destabilizing a carbon trap to promote green technologies

Ending our dependence on carbon-based internal combustion engines for our vehicles would make an important contribution to limiting the process of global warming. But accomplishing this confronts an important obstacle: in the absence of public policies to disrupt it, our carbon-based transport system is a stable equilibrium, at least in the short run. Attempts to escape it will be impeded by a series of negative feedback processes that reinforce the prevailing carbon-based equilibrium.

We can use models similar to those we have used to understand housing price bubbles and the disappearance of the Arctic sea ice to explore policies that will overcome negative feedback processes that sustain the ‘internal combustion engine’ stable equilibrium, and make rapid progress in reducing the CO₂ emissions of our transport systems. (This is another example of the versatility of the dynamic model with the S-shaped dynamics curve.)

A carbon trap: Lock-in to conventional carbon-based transport

Here are the assumptions in the model. There are two kinds of vehicle—powered by either conventional carbon-based internal combustion (c-vehicles) or electric batteries (e-vehicles). The costs of owning and using an e-vehicle (per kilometre travelled) are lower, the more other e-users there are. This is true for two reasons:

• If there are few e-vehicle users then it will not be profitable to build charging stations, so they will be few and far between, which would increase the time needed to travel to the charging station and possibly wait for an available charger. It also means that trip planning will be difficult resulting in the use of alternative means of transport for longer trips.
• Economies of scale and learning by doing in the production of the EVs mean that the more of them there are on the road, the lower will be the price at which they can be sold for a profit.

Costs of owning and operating a conventional vehicle, by contrast, do not depend on the number of other internal combustion vehicles on the road. These assumptions about costs are illustrated in Figure 8.31.

Figure 8.31 shows that if there are few EVs in use then the cost of operating one greatly exceeds the cost of having a conventional vehicle. But because costs of operating EVs fall as use expands, if everyone were driving EVs the situation would be reversed. We call the intersection of the two cost curves the break-even point because it indicates the extent of EV adoption such that someone shifting from carbon to electric would ‘break even’; that is, they would not incur higher costs by going electric.

The problem is that in most of the world most of the vehicles are powered by carbon-using internal combustion engines.

To the left of the break-even point, if people cared only about costs nobody would buy an EV. This is a carbon trap, another case of locking in to an undesirable outcome, like a poverty trap.

But if somehow enough people were to switch to electric so that the fraction of EVs was to the right of the break-even point (indicated by the intersection of the two lines at Z) in the figure, then the cost advantage would shift to electric vehicles, and even more people would switch. In many countries it appears that people are happy to switch to EVs if the costs are similar. This means that if there is no cost disadvantage to going electric, even a small amount of pro-green values would be enough to propel the remaining c-vehicle users to switch, resulting in a transition to a mostly EV transport system, thereby escaping the carbon trap. However, it would be hard for this to happen when a society is near zero on the left of Figure 8.31—EVs are virtually absent—and locked into carbon-based transport.

Green values as well as costs matter

A few people, we assume, care enough about reducing their emissions that they will go electric even if there are few others doing the same, so that (as shown in Figure 8.31), their cost will be much higher than if they had stuck with a conventional vehicle. And correspondingly, we assume that a few are sufficiently in favour of sticking with carbon that they would not buy an e-vehicle even if most others had gone electric, incurring higher costs than those driving e-vehicles (once others had done the same). Most people are somewhere in between these two extremes, willing in varying degrees to buy an electric vehicle as long as it is not much more expensive.

We assume that in any year, some given fraction of people buy a new car. They decide whether to buy an electric or a conventional vehicle based on the cost and also on how much they value reducing their own carbon footprint, on which people differ. The key idea is that a person will switch from conventional to electric if taking account of the relative costs and the individual’s own values about carbon, they prefer the electric alternative. So people switch if:

\[\text{Green values (preference for a reduced carbon footprint) > Cost disadvantage of the EV}\]

Using these assumptions, we can now build up an electric vehicle adoption dynamics curve (ADC) that is similar to the house price dynamics curve (PDC) and the sea ice environmental dynamics curve (EDC). In Figure 8.32, the adoption dynamics curve shows how many will adopt the electric vehicle alternative to carbon next year (period t + 1), for every level of e-vehicle drivers as a fraction of all drivers, this year (period t).

The figure shows that there are a few ultra-green individuals who would buy an electric vehicle even if none were in use the previous year (the intercept of the ADC and the vertical axis). They value going green sufficiently to offset the cost disadvantage of being among the first EV users. Then if there were to be more adoptions of e-vehicles, this would reduce the costs, and more people would switch. So the curve rises at first slowly because there are relatively few of these ultra-green or nearly ultra-green individuals.

The fact that most people are in the middle, not being ultra-green (but not strongly pro-carbon either) means that if a substantial number of people were already driving e-vehicles, then a small increase in the number of e-vehicle users (and the resulting small reduction in the cost disadvantage of electric) would have a large effect on the number of people switching from c- to e-transport the next year.

This is why the dynamics curve will have a steep portion in the middle, giving it the familiar S-shape. As costs become more nearly equal, there are large numbers (the somewhat green) ready to switch, steepening the ADC. Beyond the break-even point, the ADC eventually flattens out as the only remaining c-vehicle users are committed to carbon even if EVs are cheaper.

We know from our analysis of the house price and sea ice dynamics, that in the figure both the carbon-based (B) and the e-vehicle-based (G) equilibria are stable, while point T, with a substantial fraction of both vehicle types, is an unstable equilibrium, the tipping point.

The break-even point on the horizontal axis (point Z) is to the right of the tipping point. This means that the population would tip to adopting EVs even before the cost disadvantage of EVs had been entirely eliminated (because we assume that most people have green enough values to buy an EV if it is not more expensive).

How a change in values or public policy can disrupt a ‘bad’ equilibrium

Given that most of the world today is locked into the carbon-based equilibrium B in Figure 8.32, how could a society move from there to the e-based alternative at point T?

Breaking a society out of the carbon trap could be accomplished in the same manner that the Swedes managed to shift from driving on the left to driving on the right: by a government mandate requiring a large enough fraction of the population to use e-vehicles rather than conventional ones, so that the proportion of e-vehicle users in period t exceeded the tipping point at point T. That is, the government can simply require the switch to electric vehicles as many are doing, for example, by banning the sale of conventional vehicles at a certain date, thereby replacing our model of private adoption by legal enforcement.

However, there are also ways to promote the transition to an electricity-based transport system through changing things that influence the choices that people make. This means shifting the adoption dynamics curve up. There are two ways that this could be done: reducing the relative costs of EVs and changing people’s values.

The relative costs of using an EV could be reduced, for example, by government subsidies for their purchase, or for research and development of better vehicles (and batteries) and the production of the vehicles, or the network of charging stations. The top panel of Figure 8.33 shows that by shifting downwards the costs of EVs, these policies would reduce the fraction of the population that would have to adopt EVs in order for them to no longer be more expensive. The break-even level of adoption shifts to the left. As shown in the lower panel of Figure 8.33, this reduction in the costs of EVs shifts upwards the ADC.

An increase in the value people place on reducing their carbon footprint will have the same effect on the tipping point. For a given level of EV adoption (and therefore a given level of relative cost of the EVs compared to conventional vehicles), more people will go electric than before due to their change in values. We show this directly in the ADC figure since we have drawn Figure 8.32 (and the top panel of Figure 8.33) for a given distribution of green values.

The upward shift in the ADC resulting from either a change in values or in relative costs has two effects, as shown in Figure 8.33. It moves the carbon-intensive equilibrium (the status quo) to the right, so it is a little less carbon intensive. And it also shrinks the distance between the new status quo (B′) and the tipping point (T′), so that it becomes more likely that some shock—such as a particularly attractive design of one of the EV models that boosts sales unusually—would push the level of EV use this year beyond the tipping point T′.

Further subsidies of the EVs would push the ADC further up, eventually eliminating the carbon-based equilibrium and setting off a green positive-feedback process resulting in the runaway adoption of EVs. In the model, the new equilibrium is stable and in the absence of large shocks would persist even if some of the subsidies were removed, or values became a little less green (shifting the ADC downwards, but not far enough to eliminate the EV-based equilibrium).