How fungible are high-earning workers? This is the question posed by David Cameron's attempt - which might not succeed - to defend the City against the EU.

To see what I mean, let's assume that regulations did force the City to shrink, causing some highly-paid people to lose their jobs. How big a loss would this be? There are two extreme possibilities:

1. Those workers would remain unemployed, so the loss to the economy would be equal to their lost earnings.

2. Those workers are highly skilled and could find employment elsewhere, or set up their own businesses. In this case, the loss to the economy would be equal to the difference between their earnings in their first-best occupation (the City) and their earnings in their second-best - whatever other jobs they do. Such a loss is likely to be small. Put it this way. Financial intermediation in London accounts for just over 4% (pdf) of GDP; this exaggerates the contribution of the City as it includes bank tellers in Acton as well as genuine City workers. This means that if the City shrinks by half - and I know of no estimate as high as this - and workers lose half their income in moving to other jobs, we'll lose 1% of GDP.

The contrast between these two positions is, of course, not confined to the debate about the City. Samuel Brittan has long argued against subsidizing jobs in the arms industry on the grounds that workers who lose their jobs if the sector shrinks would be re-employed elsewhere. To obsess about protecting jobs in a particular industry is, he says, to commit the lump of labour fallacy.

How, then, can one believe that lost jobs in the City would be a big problem? One possibility is that City workers owe their high earnings to job-specific human capital or to organizational capital that would be lost if they changed industry, and so they would suffer a big loss of earnings*.

But even if this is true, couldn't such losses be offset - at the macro level - by increasing employment by using orthodox expansionary monetary or fiscal policy?

Herein, though, lies a curiosity. When miners were threatened with pit closure in the 80s, the Tories believed argument (2) over argument (1). And when they cut public sector jobs, they also believed - maybe they still do - in argument (2), thinking that public sector workers would find private sector employment. However, when they consider City workers, they seem to believe in argument (1). Miners and civil servants, it seems, are fungible, but City workers aren't.

Which is really queer. If City workers really are so much more skilled, ingenious and industrious than idle civil servants and beer-swilling bolshy miners, shouldn't they find it easier to get well-paid work elsewhere?

It seems, then, that the Tories are being a little inconsistent.

The kindest interpretation of this is that they are so averse even to second-order temporary losses of GDP that they are prepared to jeopardize our relations with Europe to avoid them. There are, though, less generous possibilities.