Not so long ago, the UK's political classes used to rubbish industrial policy as "picking winners". But these days, things have changed. As Ha Joon Chang, an economist at the University of Cambridge recently observed, with Vince Cable in the lead and the TUC and CBI in the chorus, our policy wonks and politicians are all industrial activists now.

But if we are witnessing a change of faith, it's a conversion that rests on the antidote fallacy. Since 1945, the syllogism which justifies every major British policy regime change is that if strategy (a) is a dismal failure, then the opposite strategy (b) will deliver policy objectives. The definition of strategy (b) is always blurred, rhetorical and imaginary so that policy never engages with specific national conditions or variations in a specific sector.

The antidote fallacy gave us ill-considered nationalisation after 1945 as the antidote to the coal owners and then as balm for distressed sectors. After 1979, it gave us labour market deregulation and all the rest: an antidote to British corporatism. The result is a 30-year policy cycle where initial failures are attributed to the incomplete revolution, before the undeniable gap between promise and outcome drives policy makers to embrace the antidote.

So it is with industrial policy. Old aims of manufacturing competitiveness and export success are now to be delivered by a new investment bank, and policy targeted on priority sectors such as hi-tech, green technologies and the creative industries. The sectoral list is an old device which has been rehabilitated. So the novelty comes from proposals like Robert Skidelsky's for a national investment bank, which, among other things, would fund SMEs and green projects. But this is what I call "foie gras industrial policy", whose assumption is that if we choose and force feed the goose it will deliver what we want.

Enthusiasts cite national examples (currently Germany, Korea and China; previously Japan), where success is attributed to focused government ministries and supportive banks. Their emphasis is on generic policies and transferable results rather than on exceptional national conditions such as a trajectory of sustained development, intact supply chains which build technical and financial resources, plus government ministries with sectoral expertise.

With recent initiatives like the Green Investment Bank, the coalition is only half-heartedly doing what the industrial policy enthusiasts want. But there are fundamental questions about whether a more active state plus investment bank finance can do the trick. This is because the precondition for successful foie gras industrial policy is a healthy goose, but the British manufacturing sector looks more like Monty Python's dead parrot.

There has been no sustained increase in British manufacturing output since the 1970s and firms are now braced for another punishing downturn. Government departments do not understand the underlying structural problems about broken supply chains and the wrong ecology of ownership because ministries now lack the sectoral knowledge, which in the 1980s was the basis for intervention on British content in North Sea oil engineering.

Politicians are deceived by manufacturing success stories which do not bear close examination. We are building more cars, but 35% of the value of UK automotive output is imported. British-owned manufacturing firms employ, on average, 14 workers and will never export directly because they are trapped in a workshop sector; the factory sector has been sold to foreign owners whose branch factories are part of a global division of labour.

This is not an argument against industrial policy per se, but against the antidote fallacy which encourages the unreal expectation that an industrial policy focused on an aspirational economy can deliver the competitive success that has eluded British manufacturing since the war. This will end in disappointment and right now does not start the necessary debate in the UK about what kind of private sector we want and the policies that can get us there.

It would make much more sense to start with the foundational economy, which sustains the infrastructure of everyday life via utilities, food and supermarkets, health, education and welfare. The foundational economy currently employs nearly half the workforce distributed throughout the UK. For example, the foundational economy includes food processing: our largest manufacturing sector and a powerful lever for import substitution. And the foundational economy is all that's left across large areas of the north and west of the UK which need industrial policy in a way that the Thames Valley and the M11 corridor do not.

The UK needs a different kind of industrial policy. First, to stop the disorganisation of health and education provision in the name of competition which covers corporate looting of the contracting state. Second, to curb point-value calculations like those of the supermarkets who deliver shareholder value and low prices by capturing the margins of British food processors and producers. Out of this kind of industrial policy would come a learning which ensured we did more than vary Thatcher's mistakes.