The concept of the one-stop shop or ‘retail government store’, and specifically the one-stop welfare shop, had been discussed since the mid-1970s (Wettenhall and Kimber 1996:15–17). The idea of creating a one-stop shop appeared explicitly in the report of the Royal Commission on Australian Government Administration:

The commission sponsored an experiment with a one-stop shop, the Northwest One-stop Welfare (NOW) Centre in Coburg, Melbourne. It was unable to evaluate the experiment satisfactorily but it considered NOW to be an excellent example of intergovernmental cooperation (RCAGA 1976:163) and recommended that the experiment be continued for at least two years. It further recommended that the Commonwealth Government indicate its willingness to help establish other one-stop shop centres experimentally where local and regional organisations wished to sponsor them, and where the relevant state government was willing to participate.

There were also experiments at other levels: local government, the ‘shopfronts’ established in the Australian Capital Territory from 1987 and cases that crossed sectors and levels of government (Wettenhall and Kimber 1996).

One other echo of this interest was Prime Minister John Howard’s address at the official launch of Centrelink, which traced the one-stop shop idea to the 1970s (Howard 1997):

Politicians had already picked up the notion of a one-stop shop several years earlier. It emerged from Liberal Party proposals for public sector reform contained in the 1993 ‘Fightback’ policy statements that promised to ‘combine the Department of Social Security and the Commonwealth Employment Service and devolve responsibility for administering employment and training programs from government agencies to Local Employment Boards with an increased investment on [sic] the community’ (Hewson and Fischer 1991a:39).

There was explicit reference to creating a one-stop shopfront and further discussion of amalgamating the delivery functions of the DSS and the CES, citing an Australian National Audit Office (ANAO) report on the merits of closer working relationships between them (Hewson and Fischer 1991b). The Fightback documents painted a picture that resembled Centrelink—that is, collapsing the CES into the DSS network to create a one-stop shop for the unemployed and contracting out the job-placement function.

By the early 1990s, Australia’s public administration reforms had been consolidated and the first decade officially evaluated (TFMI 1993). There were unfinished agendas. The first was the evolutionary acceptance of market principles that were assuming centrality by the mid-1990s as competition and contestability became the currency of reform. The second was the acceptance of more flexible approaches to delivery systems for public services and the need to move beyond the traditional monolithic departmental structure.

New trends were appearing that influenced the direction Centrelink would eventually take. These included:

A pivotal influence on the DSS was exposure to the New Zealand system of decoupled policy (ministry) and implementation (agency), which occurred during the six-country (Australia, the United Kingdom, Canada, Ireland, New Zealand and the United States) annual meeting of social security departments in the mid-1990s. One view was that Centrelink was born on a visit to New Zealand, where Tony Blunn was exposed to delivery arrangements that went beyond the boundaries of what was thought possible. The experience produced a ‘sea change’ in terms of what was seen as operationally attainable.

Three ideas were now in circulation. The first was the growing realisation that service delivery could be a form of specialisation that went beyond arrangements internal to a department. The second was the acceptance of the focus on the ‘customer’ as a primary objective. The third was the movement away from the traditional DSS office layout by using open-planning principles in branch offices that encouraged better access for customers. The confluence of these complementary (and contending) ideas provided the opportunity for an organisation design that envisaged some combination of service delivery, purchaser–provider and integrated services.

One of the most interesting developments internationally at this time was the experiment with separating policy formulation from implementation, which included the creation of new executive agencies (Halligan 1998; Rowlands 2003). The diffusion of these ideas internationally was one of the strongest public sector trends of the 1990s, making the decade an age of ‘agencification’. The cases from New Zealand and the United Kingdom were well known to senior staff of the DSS and the central agencies.^[4]

In Australia, these ideas were not accepted either rapidly or enthusiastically. During the reform period, three positions were important: an attachment to traditional views regarding the handling of policy; the horizontal consolidation of departments; and the shift in attitudes in the 1990s towards separating policy and operations. Until the early 1990s, Australia was attached to combining policy and implementation roles within a department because it was thought essential to maintaining effective feedback from those delivering services. Of the anglophone countries, Australia remained closest to the tradition of maintaining integrated policy and administration. This tradition also reflected the firm belief that there must be close interaction between policy and execution and that integration within one department was the best means of securing this. Australia not only resisted moving in line with the United Kingdom and New Zealand, it followed a different approach in 1987, when it sought policy coordination and portfolio rationalisation through the horizontal consolidation of departments.

At the same time, however, the doctrinaire belief of the APS in the superiority of integrated policy and implementation was being eroded. The British executive agencies and New Zealand’s decoupling now seemed less unpalatable. In 1994, the option of experimenting with new-style agencies was raised at a conference by the former head of the Department of the Prime Minister and Cabinet, Michael Codd (1996:179):^[5]

Moreover, what Codd considered was a reasonable approximation to the arrangement was later instituted for the DSS and Centrelink (albeit in the context of questioning the merits of agencification, as befitted the ambivalence that still existed at that time). He argued that social security entitlements were defined clearly in the law and that the boundary between exercising administrative authority and policy work should be capable of being drawn with sufficient clarity without invoking an agency arrangement (Codd 1996:180). The benefits of agencification might therefore be realised without requiring a major organisational change.

There were other signs of change. Blunn, the head of the DSS, reflected on his conversion: ‘[W]e’d previously been trying to integrate service delivery into the policy structure so that it was sort of seamless because [the] view…was that if you separated service delivery from policy then you created the potential for failure. But in my view it wasn’t working’ (Interview).