Renegade senator Jacqui Lambie may be leading a populist cause in support of her former comrades-in-arms in the Defence Force, but that does not mean she is wrong. Indeed, there are very good reasons to share her concerns about the government's (and its so-called independent tribunal's) decision on military pay. The need for budgetary restraint is not a sufficient reason to cut pay and conditions, nor is the fact that the government must find ways to pay for its decision to deploy more military in the Middle East.

Budgetary pressures are best addressed by disciplined priority-setting about what activities government should undertake, which ones might be contained and which ones allowed to expand. This should include reviews of who should receive what benefits and who should pay what taxes. Budgetary pressures will inevitably affect public sector employee numbers and their price, but these impacts are best made as a result of decisions about what activities governments choose to do or not do, not on crude judgments (prejudices) about the number of public servants and their remuneration.

If governments choose to undertake particular activities with defined effectiveness and quality, the question becomes how to ensure these are conducted with maximum efficiency. This will determine the resources needed in terms of number of employees (or the level of contracted services) and the skills involved, the price of those skills being determined by the labour market.

Fundamentally, it is the labour market, not budget restraint, that should determine the price of the skills required. Of course, the price must be affordable – but that is primarily a function of the activities governments want to perform being affordable.

How are these prices set in the labour market? Much has been said about the need for pay increases to be funded by productivity gains. Certainly, real pay increases can only be sustained by increases in productivity in the Australian economy, and Australia has been well served over the years by enterprise bargaining processes that have helped to identify and pursue productivity measures. But there is not a one-to-one relationship between wages and productivity in each enterprise. The price of labour will reflect changes in supply and demand, transferring higher real wages for skilled labour in the most productive enterprises to similarly skilled labour in other enterprises and industries. Indeed, if productivity improves sufficiently overall, real increases in wages will flow more widely, even to unskilled labour if demand is sufficient (as we have seen in Western Australia).

What does this mean for the public sector? Identifying and pursuing opportunities for productivity improvement remains sensible, but expectations that these can or should fully offset pay increases are just unrealistic – particularly after 25 years of enterprise bargaining in the Australian Public Service in some form or another. Genuine productivity gains are still available, particularly through constant investment in new technology, but, fundamentally, pay rates should reflect the supply and demand for the skills the public sector needs.

For the military, with continued significant activity, and with the increasing need for highly educated and skilled people, it is hard to see any case for reducing real wages unless the concerns expressed a few years ago about recruitment and retention have now disappeared. Moreover, not widely appreciated yet, the government's budget announcement about military superannuation involves a reduction on average of about three percentage points in employer contributions. That is, the military are facing a significant real reduction in pay and conditions despite increased demand for increased skills. Military pay and conditions are generous, particularly compared to those for the public service, but that is not the central issue: the key question is whether they are more than is necessary to attract and retain the people we need in our military, and I have yet to hear that case presented.

The public service has been warned it will be treated less generously than the military. With sharp reductions in employee numbers, there may be a case on supply and demand grounds for a tougher approach (it is particularly hard on these grounds to justify increasing the high pay rates at the very top, where the Remuneration Tribunal seems, in recent years, to have reverted to the discredited "comparative wage justice" approach). But considerable care is needed if our leaner public service is to be more highly skilled on average, and is not to lose the corporate knowledge that is crucial to its role in advising government. Some differentiation is needed, not just by classification level but also by qualifications and skills where these are in high demand elsewhere.

What is missing is some public document – from the Public Service Commission or its minister, Eric Abetz – about the basis for setting public service pay and conditions. Such a document needs to go beyond the rhetoric of budget restraint and productivity offsets, and give guidance on how future pay and conditions should be set and the evidence needed to make informed judgments. The shift away from each agency having authority to negotiate its own agreement is welcome recognition of the APS's shared labour market environment (and should go a lot further), but we need a more considered whole-of-government approach.

This might also clarify policy on superannuation. The push to remove super from enterprise agreements leaves a question mark that the Community and Public Sector Union is rightly concerned about. The current (open) scheme, the PSSap, formally involves employer contributions of 15.4 per cent but allows employees to reduce this to the statutory minimum (and take the difference in taxable income) or to supplement the contribution. The benchmark amount reflects historic employer contributions in Commonwealth schemes and is of the right order for people to achieve a retirement income that allows them to maintain living standards (most will need to contribute more, particularly if they have any periods with lower contributions, as few retired public servants will be eligible for much through the age pension).

Other than to cut public servants' pay and conditions, there is no policy reason to reduce the employer contribution rate. The statutory minimum is just that, and governments should continue to encourage people to save more. Only if total remuneration is excessive would there be a case to reduce the employer contribution and, in that case, a more sensible and transparent approach would be to reduce real wages, not the employer super contribution rate.

It is hard not to get the feeling that public sector pay and conditions policy is based on prejudice, not careful policy analysis.

Andrew Podger is a professor of public policy at the Australian National University, and a former departmental secretary and public service commissioner. andrew.podger@anu.edu.au