According to one narrative, a full meeting of the Cabinet is where the big decisions are made; but according to another it is no more than a rubber stamp. The truth lies somewhere in between, though where exactly depends on the issue – and government – of the day.

Certainly, there’s no doubting the symbolic significance of Cabinet rank – which is why Prime Ministers come under such pressure to expand its membership. Writing in the Independent, Alun Evans provides a brief history of Britain’s top table:

“When he was Prime Minister over half a century ago Harold Macmillan commissioned a new table around which the Cabinet should meet. Its distinctive coffin shape was intended to give Macmillan, as chairman, a line of sight to all attending the meeting. Macmillan had started out with a Cabinet of 17 but this increased over his seven-year term of office to 23.

“When David Cameron’s coalition Cabinet last met there were no fewer than 33 politicians at the table. Indeed the table itself has had to be extended to crush the full membership, together with the Cabinet Secretary and note takers, around it.

Though the number of full Cabinet members is limited by the Ministerial and Other Salaries Act 1975, a lesser category of membership has emerged:

“Attending Cabinet when not a full member is not new. In 1979, Norman Fowler, the transport minister, was not a full member of Margaret Thatcher’s Cabinet. But it was Gordon Brown who created a two-tier Cabinet in 2007, by having 23 full members and nine others who were invited to attend when relevant business was being discussed. Mr Cameron has continued the two-tier approach and now has 22 full members plus 11 who attend every meeting, not just those meetings when their business is on the agenda.”

Evans argues that this is too many and that, in an era when we’re struggling to cut government down to size, a more streamlined Cabinet would set a good example. This, he argues, could be achieved “quite easily”:

“The Deputy Prime Minister could be responsible for Communities and Local Government, Scotland, Wales and Northern Ireland; the Chancellor would, as now, oversee the Chief Secretary to the Treasury; the Foreign Secretary could oversee the International Development Secretary; the Business, Innovation and Skills Secretary could oversee the Culture Media and Sport Secretary; and an Environment, Energy and Climate Change Secretary could take over both Defra and DECC.”

This seems perfectly workable (though I’d get rid of the pointless business department altogether and combine culture, science and universities into a single department of giving-grants-for-clever-stuff). On the other hand, reducing the number of ministers does little in itself to shrink the state – indeed, it may be counter-productive.

As Tim Montgomerie explains in The Good Right, to reduce the size of government we must reduce the demand for government. That means government acting to solve problems as opposed to merely maintaining them.

Left to its own devices the machinery of government defaults to maintenance mode. Reform is only achieved if there is a dedicated minister driving it forward against the inertia and resistance of the usual vested interests. The real change-makers in this government – as opposed to the competent and not-so-competent administrators – are distinguished by their personal commitment and ability, but also by the freedom they have to concentrate on key reform objectives, the opportunity to operate across departmental boundaries and the visible authority granted to them from the top.

This is why we should make a virtue of the expanded two-tier Cabinet. The first tier would be composed of a limited number of Secretaries of State responsible for broad policy areas (as per Alun Evans’ scheme or something similar). But the second tier would be composed of ministers chosen to lead a focused, cross-departmental effort to deliver specific priorities for reform. There would be as many of these ministers as there were priorities – and their Cabinet rank would embody a government organised around policy goals rather than civil service structures.