More recently, the so-called commission of audit tried to take the game further and Finance Minister Mathias Cormann has issued a paper titled Smaller and More Rational Government.

Interest in the dreary business of its administrative machinery is commendable, though a good deal of related material in both the audit commission's report and Cormann's paper is not so reassuring. Here, there's an unnerving sense of starting with an answer and working backwards, a willingness to follow the nostrums of false prophets like John Uhrig and inadequate attention to the practical working of certain proposals.

The audit commission tries its hand at what it calls ''principles of good government''. It's stretching it to see many of these as ''principles''. Most are dull cliches: for example, ''live within your means'', ''reduce complexity'' and ''protect the truly disadvantaged''. Get the picture?

The commission augments its ''principles'' by what it calls ''propositions'', one of which is that ''the need for independence alone does not justify the establishment of a new operational body''. Sometimes that may be so but it would have been more helpful to say that the need for independence from government is a critical requirement for the establishment of a body separate from a department.

So armed with a cack-handed ''proposition'', the audit commission wades into the Australian Public Service Commission and concludes that it should be abolished. It says the APSC could be displaced by ''a stronger role for secretaries, coordinated through the Secretaries Board''. This proposal is as vague as it is unworkable. It's hard to imagine a ''stronger role for secretaries''; they have it all now. And the idea that a meeting every couple of months of the 25 or so members of the Secretaries Board could coordinate anything is a joke.

Perhaps seeing the funny side, the audit commission says that ''alternatively'' the APSC's functions could be ''assumed'' by the Employment Department's secretary. It notes the department was responsible for public service workplace relations matters ''in the past'' but it doesn't go on to say it did the job badly, especially when audit commission member Peter Boxall, as the department's secretary, tried, in a flagrant abuse of the merit principle, to make accepting a Australian workplace agreement a precondition for appointment to the public service.

And that's the point. Apart from the sheer impracticality of expecting an already preoccupied departmental secretary to take on what is now the full-time job of a statutory officer, it would be doubly difficult for a secretary directly subservient to a minister to, at the same time, ensure the integrity of the system for merit-based appointments. As the audit commission rightly points out, this is the APSC's principal function and it should continue to perform it at a step removed from ministerial supervision.

Apparently, the government is yet to decide on the audit commission's recommendation on the APSC, but it should reject it. If it doesn't, it's to be hoped the Senate will have the good sense not to pass the necessary enabling legislation.

Indeed, there's a good case for the APSC role to be expanded to provide more common services to departments, including through the negotiation of a service-wide pay and conditions enterprise bargain that could save tens of millions of dollars. How could a fair-dinkum audit commission on the prowl for savings turn, as it has, a blind eye to this possibility while urging scrimping on petty cash via common support services for the ''cultural agencies''?

Cormann says his paper on smaller government owes a lot to the ''recommendations and reform principles'' of the audit commission. The document yearns for a ''leaner and more business-like administration''. It outlines the treatment, as it were, accorded 48 government agencies in the budget and the abolition of another 36 in a ''second phase of reform''. The paper claims its ''agenda is based upon methodical rethinking of the Commonwealth's responsibilities from first principles'', though there are reasons to doubt that.

For example, Cormann's paper says that Uhrig's report, commissioned by the Howard government, on the corporate governance of statutory authorities ''put clarity around the circumstances where it was appropriate to establish separate statutory authorities and boards''. That report, which has been thoroughly discredited, did no such thing. Indeed, Uhrig largely overlooked the very problem Cormann is worried about: government indiscipline in the creation and review of statutory authorities. Uhrig's notion that statutory authorities should only have governing boards when they can be ''given full powers to act'' is utterly wrong-headed, the desirability of using boards being basically related to the degree of independence required. That's why the ABC has a board. Uhrig's report deserves to be left in the rubbish bin of Commonwealth administrative history and used only as an example of what not to do.

Moreover, some of the decisions in Cormann's paper are backward steps.

For example, a couple of committees that advise on appointments to primary industry statutory authorities are to be abolished, with that advice being ''sourced from the department''. That's a very bad move. These committees were set up to foster more merit and less politics in statutory appointments, and their abolition is retrograde governance that could not possibly have been the product of ''methodical rethinking''.

Cormann's paper also points out the government has moved to abolish the Independent National Security Legislation Monitor. In his second reading speech on the enabling bill, that great red-tape warrior, the Prime Minister's parliamentary secretary, Josh Frydenberg, said ''multiple independent oversight mechanisms already exist'' to perform the role of the monitor. That's untrue. As the current monitor, Bret Walker, said in his 2014 annual report (his last): ''The INSLM is not aware of any other office, agency or level of government doing what Parliament required to be done by the INSLM Act.''

If the government had consulted Walker about the future of the monitor, which it did not, Frydenberg might not have fallen into the trap of misleading Parliament; another slip in ''methodical rethinking''. More worryingly, if other government bodies set for abolition, merger or modification are being treated in the same way as the monitor, there's a risk that decisions about them will be similarly ill-informed. Affected agencies should be consulted before they are abolished or changed, not just as a matter of decency and proper process but so the government can make fully informed decisions. While there's scope to reduce the number of government agencies, it's worth remembering that complexity in the machinery of government is a reflection of the complexity of government tasks and, often, a dedicated organisation will be a better way of skinning the cat.

To conclude, let's go back to the audit commission. Some of its views about the Defence Organisation provoked a thoughtful rejoinder from former senator Gary Humphries. He notes the commission's view that ''it is not clear defence headquarters in Canberra has the capacity to drive efficiency and better policy outcomes'' and asks if the diarchy (the joint management by the Defence Force chief and the department secretary) ''has … had its day''. He goes on to suggest ''a high-powered, independent commission, answerable to the Parliament, for setting the nation's strategy for national security including defence, with responsibility for setting the size and structure of the ADF and the nature and types of capabilities it uses''.

At times, the Defence Organisation may seem to be unable to help itself, but it's hard to think of a major institution in the country that, throughout its history, has come more or less successfully through bigger and more dramatic change. Even in the 1990s, under the diarchy, the department reduced its staff from about 45,000 to 15,000 and the Australian Defence Force reduced its full-time personnel from 75,000 to 50,000. At the same time, thousands of jobs were ''market-tested'', with average savings in excess of 30 per cent. The diarchy is not a problem; it's one of the Defence Organisation's strengths and separating the ADF from the department, as has been done in New Zealand, would create awful problems of coordination and inefficiency.

And it's hard to see how an ''independent commission'' running national security would do anything other than make matters worse. It would further coagulate and confuse an already over-heavy top management, and those with an attachment to democratic forms might prefer national security and defence to be the direct responsibility of ministers and governments.

If one were to assume that the audit commission is right about the capacity of Defence headquarters (and it's by no means clear that it is) to ''drive'' efficiency and better policy, what should be done? How about these for starters?

Anyway, that's enough about the machinery of government. It's important but it's as dry as a biscuit.

Paddy Gourley has held several senior public service roles, including deputy secretary of the Defence Department. pdg@home.netspeed.com.au