One of the few but crucial advantages that economists have over others in attempting to navigate the world is the recognition that the optimal amount of a ‘bad’ – such as pollution, crime, and obnoxious advertising – is rarely zero. This is at first a shocking statement – how could one dare suggest that it is best socially to have a positive number of muggings every year, noxious fumes from cars, and ads on YouTube? Yet, it is nothing more than the flip-side of a much more familiar idea: the optimal amount of a ‘good’ – a desirable thing such as carrot juice or Theodore Dreiser novels – is not infinity.

The central economic principle underlying both statements is called marginalism. It was independently developed by Carl Menger, William Stanley Jevons and Léon Walras in the 19^th century, and it posits that optimal economic decision-making happens at the margin. This means that people will be willing to pay to acquire units of a ‘good’ – or to avoid units of a ‘bad’ – up to the point where the cost of an additional unit – the marginal cost (MC) – exceeds the marginal benefit (MB) to be had from its consumption (or avoidance). How many units I acquire will be determined not by the total benefit I derive from consumption, but by the point at which the cost of one more unit is higher than my private benefit from having it.

Marginal thinking is at the heart of economic theory and practice, and it is taught – if not in verbal then in graphic and mathematical form – in one’s first or second ECON 101 lecture in university. However, like other established conclusions from economic doctrine – such as the principle of comparative advantage, or of opportunity cost – its influence outside the discipline leaves much to be desired. Nary a day passes by where one does not read political statements totally disregarding the MC = MB rule. When the infamous Red-baiter Joe McCarthy admonished that “even one Communist in the State Department was one Communist too many,” one could see that he had not studied economics – unless the corrupting effect of a single member of the Communist Party USA could be so great as to justify the vast resources required to find him in the haystack of State Department employees.

But one does not have to go back to the paranoid 1950s to find examples of the egregious butchering of good economics. The UK Government’s recent announcement that it plans to ban all petrol and diesel vehicles by the year 2040 is as shining an example of economic malpractice as one can muster.

One may first of all question – as others have done – the effectiveness of a measure which, even if it becomes law, will only be enforced a quarter-century hence. Beyond the private political benefits to accrue to the Tories – and the measure’s principal sponsor Michael Gove – from this ostensibly tree-hugging display, will the announcement have any impact, good or bad? After all, if an authoritarian – but, in the eyes of some, enlightened – government had in the 1980s proclaimed that, in the interest of capillary propriety, all mullets were to be banned after 2015, some surely would have cheered. But the actual consequences of the measure would have been limited, as mullets have long gone the way of other 80s mainstays such as the videocassette.

No one can predict the future, but it is likely that, by 2040, fossil fuels will have become a smaller share of all power generation, including automotive power, in the UK. But, if the broad trends in power generation and consumption are uncertain, the sequence of changes, their cost and allocation – where fossil fuels will first be superseded and to what extent – is virtually impossible to fathom. This is what makes Gove’s blanket ban on petrol and diesel such a bad idea.

Consider this: we know that cars powered by fossil fuels pollute, and that pollution is undesirable and has potentially rather significant economic and other consequences over the long term. We also know that we rely on cars for much of our ground transport, and that foregoing car use – or foregoing other goods by spending more money on an electric car – is costly. How do we balance the two? The economist’s magic all-purpose machine spits out the answer: MC = MB. But in order to implement the rule successfully we need to let markets work. In a free market, prices convey the relative costs and benefits from undertaking different activities, thereby facilitating efficient human action without requiring Olympic amounts of information.

Ah, you say, but the problem with environmental policy is that some – potentially very significant – costs are social not private, and thus they are often not captured by market prices. This objection is correct, but it does not invalidate the above reasoning nor does it suggest that banning fossil-fuel-powered vehicles is a sound idea. Instead, the efficient thing is to find a mechanism by which those social costs can be priced and included in people’s decision-making, so that they behave efficiently rather than overconsume because some of the cost is borne by others.

This outcome, as economists tirelessly argue, can efficiently be achieved by a cap-and-trade system – which I proposed in a recent IEA paper on post-Brexit electricity market regulation – or a carbon tax. It is far from straightforward to establish what the optimal amount of emissions is at any point in time – which is required by a cap-and-trade scheme – or what the social cost of pollution may be – needed to decide on the right amount of carbon tax. But either of those programmes will achieve the same pollution reduction at lower cost than a blunt measure such as a ban on particular emitters.

The most salient pollutant from fossil fuels, namely CO2, increases linearly with fuel consumption. This is true for both petrol- and diesel-powered cars, and it suggests that it may be desirable to curb fuel use by drivers. But the level at which it no longer pays to reduce fuel consumption – because MC > MB – is not only unknown but ever-changing, given oil price fluctuations, improving emissions reduction technology, changing driving and living patterns, and so on. Nobody knows whether zero will be the UK’s optimal amount of car emissions in 2040. Yet, that is what would be required for the Government’s proposal to be considered economically rational.

Elementary economic reasoning would quickly disabuse Michael Gove of the notion that banning petrol and diesel in any way makes sense. Instead, he might wish to consider an all-embracing emissions reduction target, which would automatically lead the most inefficient polluters – that is, the ones with the highest cost for the lowest benefit – to cut emissions first. This point is particularly salient for car use given the likely negative upstream effects of the measure: what is the point of banning petrol and diesel if more of the electricity required to power cars is then generated by dirty fuels? We could even end up with more emissions, at a higher cost, than good old-fashioned reliance on markets would have produced.

But don’t fret, Michael – thanks to capitalistic innovation you now have access to scores of online tools to learn marginalism in a few easy steps. I can suggest ieaTV and Cowen and Tabarrok’s marvelous Marginal Revolution videos. You have until 2040, but don’t wait too long!