In 1957, British prime minister Harold Macmillan could tell Britons they had never had it so good and still go on to win the next election. That would be impossible in Australia today. Politicians feel it necessary to tell us we are doing it tough, even though the evidence is precisely the reverse. For most Australians, unemployment and inflation are lower than in living memory, real wages are higher, health is improving, crime is down, we have more information and leisure options than ever, families can afford overseas holidays that once only the wealthiest enjoyed, food is getting cheaper, so are consumer goods.

The people for whom life is a struggle are on Newstart Allowance or in remote indigenous communities - but one doubts it is these voters whom Gillard had in mind.

In setting out her priorities for the future, she stated the obvious: ''to shape a future of prosperity and shared opportunity, we have to manage our economy well''. The government had identified ''five pillars'' to increasing jobs, prosperity and productivity: ''increasing skills, building a national culture of innovation, rolling out the national broadband network, investing in infrastructure, improving regulation, and leveraging our proximity to and knowledge of a rising Asia into a competitive advantage''. That's clearly six, not five. It seems an extra pillar was added to the speech at the last minute. It's anyone's guess which one.

Fortunately for Gillard, coverage of the speech skipped over these rhetorical flourishes and instead focused on her announcement that the election would be on September 14, 7½ months away. That long a lead time is unprecedented in Australian politics.

It is, however, nothing like the longest caretaker period in Australian history. The caretaker period only starts with the issue of the writs (documents under sections 12 and 32 of the constitution that tell officials to get cracking and organise an election). Gillard said she would ask the Governor-General to dissolve the House of Representatives and issue writs on August 12. About a month for the official election campaign and caretaker period is pretty standard these days.

What the early announcement of the date amounts to is the closest federal politics will get to a fixed term for Parliament. It carries all the advantages and disadvantages of fixed terms. The positives are certainty for planning and decision-making. The negative is the risk that, as the election date approaches, both government decisions and opposition promises become increasingly flaky and unachievable.

The election date itself will be of little concern for the public service, and the benefits of certainty are large for some areas. The Electoral Commission, for example, though it excels at last-minute organisation, will welcome the luxury of a more relaxed planning horizon than usual.

What is more likely to be of interest for the future of the federal public sector is the as-yet unrevealed list of savings which were also mentioned in the speech. Gillard said ''it is the wrong time to be spending without outlining long-term savings strategies which show what will be foregone in order to fund the new expenditure''. The government is looking for savings to fill the hole left by lower tax revenue as the minerals boom goes off the boil.

Her office circulated selected excerpts from the speech to some journalists in advance. This tactic raised expectations that the speech would outline specific cuts, which the media eagerly awaited. No such luck. Instead, Gillard kept her remarks general, saying only ''we are in an era when new structural calls on the budget need to be associated with new structural savings''. She did, though, hint at what might be involved by recalling past structural savings, including those to the dependent spouse rebate, tax concessions for ''golden handshakes'', tax concessions on superannuation for high-income earners and means-testing the private health insurance rebate.

The timetable for a more detailed list was left unclear. It was ''in the lead-up to and in the budget, we will announce substantial new structural savings that will maintain the sustainability of the budget''. That could be any time between now and May 14.

It is a three-S savings strategy: substantial, structural and sustainable. What does each mean in practice?

Substantial has to be interpreted in the context of the likely price tags for the new priorities that these savings are to fund: the national disability insurance scheme and the school reforms arising from the Gonski review. When fully implemented, they are likely to cost more than $12 billion and $6 billion a year respectively. Some estimates are higher. That is big money in anyone's budget. Just how much will fall on the Commonwealth and how much the states and territories is yet to be settled. However it pans out, the Commonwealth still faces a very large bill indeed.

Savings of the size needed to fund these programs cannot come from small nips and tucks here and there. An extra per cent or two on the efficiency dividend won't cut it.

To earn the tag ''substantial'', savings must be found in social security and welfare, health, education or defence - nothing else in the budget is big enough. The question is whether the government has the courage in an election year to make such savings.

Cutting other education spending to fund Gonski reforms would make sense, but the only substantial areas remaining are higher education or payments to private schools. Both have well-organised lobby groups that would protest loudly against cuts. Savings out of the current social security budget large enough to meet the costs of the NDIS would need to include significantly reduced entitlements, more means-testing or both. There is plenty of scope to make savings in health by reducing overlap between federal and state/territory governments, and by limiting growth in expensive new pharmaceuticals - but again, protests would be loud. We have already seen the strong fight by Defence, albeit conducted by proxy through statements from retired senior military officers, against cuts in that portfolio.

In light of the likely protests, substantial savings might come not on the expenditure side of the budget but through further measures to tackle tax concessions, especially those seen to favour the wealthy. Technically, these are revenue increases, but are sure to be presented as budget savings.

For savings to meet the second S, structural, many of the kind included in the 2012-13 Mid-Year Economic and Financial Outlook are off the table. Moving company tax payments from quarterly to monthly, as the outlook announced, can only happen once; and, in any case, the large savings ($5 billion) from this measure come in the first year of operation. They drop away to about $1 billion in later years. Moving lost super accounts to the Taxation Office after one year instead of five raised $555 million in the first year, but tails off to $34 million in 2015-16. That is characteristic of measures that change the timing of payments rather than make underlying structural changes - over the longer term, they dwindle away. The same applies even more so to the trick of shifting money between years to make the budget look good. It was a bad habit introduced by the previous Howard government and continued since, but it is simply an accounting trick, not a structural saving.

The final S, sustainable savings, means that asking the public service to do more work with less money is for once likely to be ruled out. That is unsustainable over the long term.

This is not to suggest the public service cannot be cut. Of course savings can be made, should a future government wish to make them. But for savings to be sustainable, the programs that public servants are asked to manage must be cut at the same time. There is scope to do this, especially with legacy programs set up to meet a need that no longer exists. Some have remained on the books for decades without proper scrutiny. Governments still need courage to take on the interest groups that they inevitably gather around them, but can do so if the will is there.

Although the Community and Public Sector Union seems to believe the size of the public service should grow with the population, there is no necessary link. The size of the public service depends on what it is asked to do. As noted by Public Service Commissioner Stephen Sedgwick in the overview to the latest State of the Service Report, ''the optimal size of the APS is difficult to establish a priori. It is best derived from the nature and scale of the activities the Australian government undertakes.''

Sustainable savings should improve Australia's government net worth, which we can now measure due to accrual budgeting. Accordingly, short-term measures to improve the government's cash position, while accruing large debts that have to be paid off in later years, are out.

Sustainability also means avoiding immediate savings with long-term costs, such as cutting policy units or statistical collections. These savings last only until a later government realises that information it needs for decisions is no longer available, and it then faces huge costs recreating the capacity.

At present, the budget is unsustainable. The mining boom in the 2000s provided a revenue boost that allowed the government to cut income tax and increase spending, but those days are over. There are things that can be done to address the problem: implement the Henry review of taxation; reduce regulation (which has budget as well as economic costs); conduct a fundamental review of programs which have outlived their usefulness and the proliferation of small programs; restructure federal arrangements so that different levels of government work more efficiently together. Even getting halfway there, perhaps as much as political reality might allow, would put the budget on a sustainable footing and leave future generations better off.

Stephen Bartos is the executive director of ACIL Tasman and a former senior public servant.

s.bartos@aciltasman.com.au