Ryan Bourne is Chair in Public Understanding of Economics at the Cato Institute.

It is said that, during the 1960s, Milton Friedman was visiting China, where guides took him to a canal-building site. Shocked at the prevalence of men bearing shovels, Friedman asked why the project wasn’t utilising modern technologies, such as mechanical diggers.

“You don’t understand, Professor Friedman,” his host explained, “this is a job creation programme.” To which Friedman retorted, “Oh, I thought you were trying to build a canal. If you really want to create jobs, then by all means give your workers spoons, not shovels.”

That tale is beloved by economists because it contains an important truth: gross job creation is a poor metric to judge success when considering government-led infrastructure. We could “create jobs” through getting people to fill in holes.

What matters is the net value added of the output created, as determined in markets – i.e: by consumers and open trade. Using more workers less efficiently to produce a canal reduces the output’s net value, because labour is a cost of production.

This lesson sprang to mind last week during Boris Johnson’s speech to the Conservative Party Conference. As part of his ode to offshore wind, Johnson talked of the UK’s natural abundance of the stuff (the “Saudi Arabia” of wind) and his excitement at the technology (floating wind islands). But he also extolled the idea of a UK “green industrial revolution” that “in the next ten years will create hundreds of thousands if not millions of jobs.”

Any market-led or government-incentivised shift towards wind will see new jobs in the industry “created.” But this shouldn’t be the goal. To echo Friedman, “we thought you were trying to produce electricity, subject to mitigating climate change. If you wanted to create jobs, why not have people make the wind turbines by hand?” We should judge the desirability of a pro-wind energy policy framework, in other words, by its contribution to this social value added, not numbers employed in the sector.

“Gross job creation in wind and other renewable generation is clearly a cost in economic terms,” Professor Gordon Hughes of Edinburgh explained to me last week. “The higher the number [of jobs], the larger the subsidies required and the larger the damage to the rest of the economy.”

He views this outcome as an unacknowledged problem of wind power generally, which does require labour for operations and maintenance, particularly as turbines age. If we are purely looking at how to produce power most efficiently, then these jobs are a hindrance – an economic failure, not success. According to Hughes, talking of creating “millions of jobs” is a “shortcut to national impoverishment”.

Of course, the desire for climate change mitigation means policymakers reject the premise that we just want our energy sector to simply prize efficiency. Due to the “social costs” of carbon, they aren’t just concerned about traditional measured value added, but are explicitly willing to take an economic hit in targeting a broader conception of economic welfare that takes into account these CO2 externalities.

And that’s fine, in principle. Yet even then, “green jobs” shouldn’t be the aim. An economist would say we should try (albeit imperfectly) to price this social cost, and then let markets find the most efficient way of delivering power accounting for it. What that does not require is industrial strategies, picking winners, and seeing the green energy sector as some sort of jobs machine.

Indeed, simple logic would suggest that making energy less efficiently than socially necessary reduces net jobs across the whole economy because of its impact on energy prices. “Energy is a labour-extensive industry. It does not employ a lot of people” Richard Tol, the renowned climate economist, explains. “If the energy sector would start to employ many more people, retail energy prices would rise rapidly.”

Given every other activity uses power as an input, it surely doesn’t need to be said that “more expensive energy means less growth and so less job creation” elsewhere. Given the sizes of the “energy” and “non-energy” sectors too, “a large relative increase in employment in energy is easily offset by a small relative decrease in employment in the rest of the economy.” An explicit aim to “create jobs” in the wind industry, in other words, would be vastly outweighed by job losses elsewhere.

Note: none of my analysis here is passing judgment on the desirability of decarbonisation. Tol believes that given the energy framework of UK policy, wind power will probably be cheaper than coal or gas through the 2030s. What I am simply saying is that aiming for more employment in wind, rather than merely trying to deliver power cheaply subject to any climate goals, is a deeply economically destructive way of thinking.

So why is the Prime Minister talking of millions of green jobs? Well, unfortunately, many politicians have moved beyond simply wanting to set frameworks for energy or even climate goals, and their green credentials have become wrapped with their becoming re-inured to the idea of national industrial strategies.

Not content with allowing consumers and producers to find the best ways to allocate resources, the Conservative Government increasingly wants to decide which sectors the national economy specialises in, thinking the state will this time do a good job of picking winners. And Boris Johnson’s “Saudi Arabia of wind” suggests that he wants to try to use policy to actively push the UK towards exporting wind power.

Would that work? No, says Tol. “Electricity is much more expensive to transport than oil…Wind power in the UK is cheaper than, say, wind power in Italy – but wind power generated in the UK and transmitted to Italy cannot compete with wind power generated in Italy.” Exporting the end product is a non-starter.

What about manufacturing the parts? “Despite what he [Johnson] says, no one is going to manufacture wind turbines in the UK without massive subsidies – that game is long past,” Hughes concludes. So having the Government tilt the deck further to try to create a wind manufacturing export industry would not only drag resources away from other activities with higher net value added, but make us a hostage to technological fortunes.

As Hughes has previously written: there’s no guarantee technological progress is more likely to come in renewables than fossil fuels or nuclear (see, for example, fracking).

To return to the Friedman analogy: we might accept some more shovel than machine use for canal building, if there were greater environmental costs of using machines, though recognising this makes us poorer. It’s another thing to say that rather than building a canal as efficiently as possible subject to this, the national government should intervene to support canal building or our selling shovels around the world. Yet with his promise of a green revolution, that is precisely what the Prime Minister implies.