The specific idea of creating Centrelink and the associated arrangements flowed from the two departmental secretaries most involved, Sandy Hollway (of the newly formed DEETYA) and Blunn (DSS). The duplication of the DSS and DEETYA networks was an obvious target for savings options for the Department of Finance. Blunn and Hollway recognised what was coming and moved quickly to achieve at least some control over events.

Blunn approached Hollway with the idea of creating a combined ‘shopfront’. According to Hollway, ‘it took us approximately 15 minutes to agree in principle that we wanted to head in this direction’ (Husock and Scott 1999a:3). In late March 1996, the secretaries arranged a meeting of their senior staff at which the third agenda item was ‘[c]o-operation in client service delivery, including property co-location and rationalisation; reciprocal placement of DEETYA and DSS information/desk officers in each department, and scope for [a] joint approach to realising savings’ (Husock and Scott 1999a:3). One participant commented after this meeting that Blunn and Hollway appeared to be ‘falling over one another in the struggle to appear to have the broader view and the grander vision’ (Interview with senior Centrelink official).

Blunn and Hollway, having anticipated the issue, used the available intellectual concepts (in the form of the agency or purchaser–provider models of government activity) to shape events. Those models were not at the forefront of the thinking that inspired this change; its basis was a ‘savings option’. As can be seen from the NCA report (1996), however, such thinking was not far away and was likely to have influenced the final reform.

The considerations that were influential included the following.

Conceptually, the DSS and DEETYA could be thought of as two organisations, each with shared ownership of a delivery organisation. There was a range of possible variations and substantial potential savings. There were also concerns, in Blunn’s view. First, he did not want to confuse or destroy roles. Second, he did not want to separate delivery and policy roles. There was, however, little choice. If the DSS and DEETYA did not take the initiative, others would. It was clear from subsequent events that the idea was then fed into the early decision-making processes of the new government, particularly its decisions focused on reducing expenditure.

One of the DSS division heads involved in the early process summarised it as follows:

It is generally accepted that had the DSS and DEETYA not come up with the proposal, the government would have taken action (as evidenced by the report of the NCA). Indeed, there was an element of satisfaction in Blunn’s (1997b) later comment, as part of a reflection on life in the first year of the new government, that he felt comfortable with the NCA. The government

An announcement was made in the 1996 budget of the decision to create the Commonwealth Services Delivery Agency (CSDA), later renamed Centrelink. The decision was consistent with the findings of the NCA, which reported to the government in June 1996. It recommended rationalising the delivery of income and employment services and raised the idea of opening the provision of those services to the private sector (NCA 1996:107).

A chapter in the NCA report entitled ‘Best practice in government activities’ (NCA 1996) listed a range of techniques that could be employed to improve the delivery of Commonwealth programs (a list similar to that proposed by Finance Minister Fahey in the new government’s initial weeks in office)—in particular, the discussion of ‘cross-program approaches’ that focused directly on DEETYA and the DSS. AusIndustry was cited as an example of a ‘cross-program approach to service delivery’, and ‘rationalising the delivery of income support and employment services’ was mentioned as the only potential case. The report observed that two departments were the major agencies delivering income support services:

There was seen to be significant scope to rationalise service delivery through developing a network of shopfront offices. There was also thought to ‘be scope in the long term to contract out the delivery of these services to the private sector’ (NCA 1996:107, 115).

In March 1996, the new Commonwealth Government was keen to address a substantial budget deficit. Some senior public servants believed that ministers thought that reducing APS staff numbers could generate a substantial proportion of the desired savings. An early decision (in April 1996) was made on strategies for securing a reduction in APS staffing.^[7] About this time, the government agreed that the Minister for Social Security, Jocelyn Newman, and the Minister for Employment, Education, Training and Youth Affairs, Senator Amanda Vanstone, should develop options for a one-stop shop for beneficiaries. In conjunction with the Minister for Finance, the two ministers were to bring forward a proposal on this matter for consideration by the government. The government also decided that the Minister for Employment, Education, Training and Youth Affairs should examine the scope for further privatisation of employment-placement services and report back to the government.

There were therefore two processes operating simultaneously, but the outcomes were shaped by two stipulations: the general brief from the two key secretaries (Blunn and Hollway) and the government’s expectation that initial savings of 10 per cent would be achieved. This required the merger of the employment side of the CES and the payment side of the DSS. The stage was set for the full development of proposals for a service delivery agency and associated changes to the CES and employment services.

Three models were initially offered: a super social security department, an independent statutory authority and a middle-ground service delivery agency. In terms of setting the agenda, the DSS preference was for social security to assume a broader delivery role that encompassed other departments. As this was unacceptable to DEETYA, the concept became one of policy coordination through a delivery agency. The third model was the preferred option, which derived from the original concept of a new agency emerging from the policy–delivery split.

The refining of the concept for implementation proceeded through two main stages: the initial approval by cabinet and the subsequent approval of the CSDA in the 1996 budget (before the public launch of the CSDA in September 1996). Again, two processes were operating: the first, influenced by considerations of rational design, was concerned with exploring the options for the development of a multipurpose CSDA. The basic concept was of an agency with two major clients but with the potential to serve others. There was also a ‘blue-sky’ dimension that reflected the open-ended potential of a large agency serving multiple policy departments. A range of possibilities could be envisaged for a multipurpose, one-stop shop. Ross Divett, the subsequent first deputy CEO of Centrelink, eventually incorporated the concept of an agency that could provide services in the draft legislation. In addition, there were the new governance arrangements under which it would operate. The other key element of the delivery agency refined at this time was a separate agency with a chief executive and an executive board of management (which would offer the government opportunities to appoint members with private sector experience).

The second process, paying heed to bureaucratic politics, addressed the question of pinning down the concept in a meaningful way for final approval by the government and laying the basis for the implementation of a viable organisation. This required recognition of the constraints on choices and the need to balance political expectations and the interests of departments.

The actors involved in these processes extended across the relevant departments and central agencies, with key secretaries and members of task groups most important. Of strategic importance were the head of the DSS, Tony Blunn, and a new leader of the task group, Carmen Zanetti, who was appointed fresh from exposure to the US environment during study at Harvard University.

The original concept envisaged a policy–delivery split that would produce a large agency with several small, high-level policy departments (each with perhaps 150 staff). The small policy department reflected organisational thinking at that time in New Zealand and the United Kingdom. Other models, such as the Treasury–Australian Taxation Office relationship had also been considered.

Blunn, however, had second thoughts about being left with a small department confined to policy and sought a clear demarcation of the interface between the two new organisations. Blunn’s reformulation was to add a middle area between policy formation and delivery that essentially represented program management or operational policy. This meant that the DSS would keep product design of delivery and control of the program—retaining as much as three-quarters of the program area. Under this revised concept, Blunn envisaged that the department and the agency would compete in the middle ground and the challenge would therefore be to ‘get the interface working properly’ (Interview). The use of purchaser–provider relationships was to be a core element in this process.

This compromise over the distribution of responsibilities reflected the dogma of the Commonwealth system, but had long-term consequences for operations and relationships between client departments and Centrelink.