The very steadiness is supposed to be a sign that he is here for the long haul, and that he will not get over-excited, even as he makes it clear, both to ministers and the public administration, that he expects results. He has certainly made it clear that his stop-the-boats policy begins today - and, whatever, if anything, he has in mind, should be in demonstration soon. Abbott has already, unwisely, suggested that his public servants have indicated they are eager for the confrontation this will involve. Likewise, public servants are now at work on legislation to abolish carbon and mining taxes.

Abbott has also done a few signature things, not least announced name deflation for ministers and their agencies. It is, he says, designed to produce clear lines of authority in a back-to-basics government.

Bully for that, but he may have already missed a splendid opportunity, one he will probably never have again, even if he is to be re-elected. That's to seriously pare the size of cabinet, if not necessarily his ministry, and also the number of government departments.

The cabinet, the ministry, and the number of government departments have increased since the 1987 Hawke reforms to ministerial government, and, increasingly, prime ministers become hostage to the need to be able to reward as many followers with ministries, parliamentary secretaryships and so on as possible.

This may be thought, wrongly, to promote party harmony, or the divvy-up of the spoils of office, but it creates inefficiency and waste of the sort that Abbott has insisted his government will stop.

Even the Hawke reforms were compromises on the original intention. It had been planned to have a cabinet of 12, with 12 mega-departments. By the time the deals were made there were 17 members of cabinet, and 18 departments - two of which (Aboriginal affairs and veterans affairs) were not represented in cabinet. That was a considerable advance on 28 separate departments, but at the cabinet level still meant a very large table.

The first Hawke cabinet had 13 members, and was possibly the most effective cabinet Australia has ever had. There were 14 outer ministers, each with a department, but called into cabinet only when they were invited. The second Hawke ministry was no bigger.

The third ministry was expanded to 30, and despite the organisation had 17 in cabinet. Hawke's fourth ministry then began to include parliamentary secretaries, originally purely as a wheeze to reward more followers, if with minor duties.

Best to trim such matters early. Later on, all of the pressure is to increase, not decrease, often for subsidiary reasons such as gender, state or factional balance rather than efficiency.

Just which functions fit into which agencies does not matter greatly, although, as Hawke intended, and as Abbott says, it is best to have clear lines of authority and a minimum of overlap of responsibilities.

There will always be some overlap, and in some areas, a bit of a contest of ideas may even be desirable, but the creation of departments around the ministerial talent pool, or the expansion of agencies merely to maximise the number of jobs available, is bad policy, and almost invariably bad and inefficient management.

One reason why is that cabinets - certainly cabinets operating as boards - get more inefficient as they get bigger and more unwieldy. If everyone wants her or his two-bob's worth, they get over-long; if cabinets adapt to the public meeting approach they cease to be the engine room of higher government, becoming instead a chairman's club where the seniors get just a bit

of inside information minutes before it is broadcast.

Over-big cabinets typically produce two problems that become chronic sores, of a sort that Kevin Rudd, in particular, put on display.

Their very unwieldiness means that most actual government goes down to cabinet committee level - formal or informal - whereby decisions are made by ad hoc groups of a few ministers, or, worse, get settled between minders. In many such cases, ministers who are interested parties will find that they are not there when the decisions are taken, or at least argued.

With large cabinets, those who become annoyed at slowness or inaction - and typically prime ministers and treasurers are in the front line - lean more and more on their private office to push things along, manage the agenda, arrange co-ordination (especially of media) and even to manage the cabinet agenda.

Cabinet becomes even more dull and unwieldy - where ministers blather about broad tactics without strategy, or strategy without tactics, or where some ritual (but pre-arranged) debate occurs over conflicting priorities - such as mining and the environment. The result is fixed, but the advantage of having a meeting is that it can be claimed that the issues were thrashed out, everyone was given their say and everyone was committed to the result.

We will know details of the re-organisation of departments when the administrative arrangements orders are released. But it seems clear that two departments - climate change and resources and energy - are effectively folded into other agencies. It is not clear if Foreign Affairs and Trade has been re-divided. Yet there will still be 19 ministers in cabinet, and even with supposedly clear lines of authority it will take time to establish just which public servant follows which function, or vice versa.

We know, as yet, little of Abbott as grand chairman of a cabinet, let alone as invigilator-in-chief, although so far the signs are reasonably good. His models are, or ought to be, Hawke and Howard, both skilled chairmen, both inclined to give most ministers a fair degree of freedom, but both with the nous and the internal intelligence systems able to inform them if ministers were struggling, straying, or going silly.

He should eschew the micromanaging style of a Rudd, or the propensity to insist that every decision (or every media announcement) be cleared through and approved by his private office. Rudd's style began with a state-level tendency to think that everything had to be co-ordinated from the top, because half the ministers were dangerous, disloyal or dull. But Ruddism also tends to follow that style of manager who speaks, or worries, about matters such as focus, discipline and co-ordinated messages.

Abbott was a fairly effective, frank and relaxed minister with his office and departments. He ran a remarkably disciplined and generally loyal shadow cabinet. But the leadership task as prime minister is a definite step up, with a far greater tendency to be whacked by events or ambushes, sometimes from behind.

Methodical, calm and measured is a good start.

Jack Waterford is Editor-at-Large.