The DSS/DFaCS purchased most Centrelink services and only this department could buy these same services. In fact, most of the services DFaCS purchased or funded were bought from Centrelink. While there was nothing in the legislation requiring Centrelink to undertake work for the DSS/DFaCS, there was a clear expectation that it would continue to do so and there was a provision in the legislation that allowed the minister to instruct it to do so.

Formal agreements under the CSDA Act needed to be put in place to authorise Centrelink to undertake work for the DSS. The first agreement between the DSS and Centrelink, the Strategic Partnership Agreement (SPA), was drafted quickly by a small team. The document, a 10-page ‘core agreement’ with some 84 pages of various ‘protocols’ and ‘memoranda’, largely set out what the DSS formally required and what Centrelink would deliver during 1997–98. The SPA also provided for committees at operational and senior management levels to facilitate the two organisations in conducting their business. The document was succinct compared with later such agreements.

At the end of the first year, a new agreement was signed with the name Business Partnership Agreement 1998–2001, reflecting the view of the relationship of the new departmental secretary, Dr David Rosalky. This was intended as a three-year agreement, with an expectation that some modification would be undertaken if necessary at the end of each year. The level of specification, however, had now expanded substantially and the agreements grew accordingly (for details, see Rowlands 2003).

A major point to bear in mind when analysing the DSS/DFaCS–Centrelink relationship was the near monopoly–monopsony arrangement. Equally important was that even though there was an influx of staff to Centrelink initially from other organisations and an attempt by Centrelink to define itself as ‘not the DSS’, the two organisations were once one. That had a strong practical effect on continuing relationships at a personal level between officers who formerly worked together and at an aggregate level between organisational units and the organisations as a whole. The main dimensions are examined later in this chapter.

Centrelink’s second most important purchaser–provider relationship was with DEWR, later renamed the Department of Employment, Workplace Relations and Small Business (DEWRSB). The department took a different stance to DFaCS in its relationship with Centrelink, especially on Centrelink’s performance. DEWRSB’s key objective in 1999 was building an effective partnership between purchaser and provider. Centrelink was acknowledged as being committed to this approach but it was difficult to achieve with the provider having a monopoly over most services delivered and when issues existed over the pricing of these services.

The position of DEWR was argued to have been different to that of DFaCS because about $1 billion worth of services were bought in the market based on service delivery arrangements that were almost entirely outsourced. Employment services were being delivered through 200 providers in the Job Network and services were also bought from a network of community work coordinators and providers for the delivery of Work for the Dole projects, the Return to Work Program and community sport programs.

In each case, these arrangements were a product of a competitive tender that was mostly price competitive. There was therefore a competitive market, in which the price of services was known and contracts allowed people and organisations to lose or gain business according to their performance against standards. The contrast with the Centrelink situation was that the department was negotiating with a government monopoly without generally having an alternative (Senior departmental official).

From DEWRSB’s point of view, the main obstacle to a viable relationship with Centrelink was differences about the costs of services being purchased. The department regarded Centrelink’s role as crucial to the implementation of the government’s mutual obligation policy for the unemployed. In 1999, at the time of an official but private review of Centrelink, there was concern about how policy might founder on the quality of the management. The DEWRSB position appeared to question Centrelink’s commitment and effectiveness in implementing its services.

Centrelink’s interaction with the unemployed had been (as with the DSS before it) more demanding than that with other recipients of income support. Recipients in the pensioner and family payment categories tended to interact only at the time they lodged a claim for a payment and at various review points, perhaps a year or so apart. In contrast, the number of personal interactions with job seekers was much higher. As well, the prominence of the mutual obligation policy in public debate helped increase the relative importance of the services being purchased by DEWRSB in comparison with those purchased by DFaCS.

DEWRSB’s experience with the Job Network and purchaser–provider arrangements through managing labour market programs led it to require strong, explicit contestability for specific components of service delivery, and to moving the Centrelink operation towards a contestable environment. It was apparent to DEWRSB, however, that Centrelink did not fully appreciate the implications of a purchaser-provider approach in terms of being responsive to the purchaser and consistent in service delivery (Rowlands 2002). This produced frustration when the department had contractual obligations with its Job Network service providers.

One difficulty in this relationship, which was also common to that between the DSS and DEETYA, was uncertainty about who was the client. For the DSS, the unemployed person was the client, while DEETYA related to industry. This problem had been identified by Moran (1996:19) in vocational education and training: ‘An emphasis on competition within vocation[al] education and training has meant that the definition of “the client” has been a difficult and controversial point to resolve…at least as far as governments are concerned, the principal client is industry.’ The DEWRSB view was influenced partly by the perspective of the private organisations it served. Its more robust view than DFaCS’, assuming DEWRSB to be a successor to DEETYA, could also reflect an organisation that had lost a regional network and an operational arm (ie CES) (Rowlands 2002).

DEWRSB was, however, able to cite performance information to substantiate its views. DEWRSB had retained its own computing systems—unlike DFaCS, whose main IT facilities went to the new Centrelink—and was therefore less dependent on Centrelink for management information. In addition, DEWRSB had developed experience in working in a genuinely contestable environment—specifically with the Job Network—and was enthusiastic about seeking the competitive supply of Centrelink services. It had consciously removed work—the Community Support Program—from Centrelink and outsourced it to private providers (DEWRSB 2001). It had also arranged independent customer surveys of Centrelink performance, something DFaCS had never done. It seems likely that the more businesslike and formal approach DEWRSB took to managing its relationship with Centrelink helped the ANAO form a more positive view of the arrangements between the two organisations. The ANAO attributed this to more formal consultation between the organisations at various levels and better performance reporting and analysis (ANAO 1999b:85).

DEWR then took a more aggressive role than DFaCS in dealing with Centrelink and was suspicious of its potential monopoly position. According to Rosalky (2002:8–9), DEWR worked with central agencies, such as the Department of Prime Minister and Cabinet and DoFA to limit Centrelink’s independence.

From Centrelink’s point of view, much of the problem was that DFaCS and DEWR provided different instructions in relation to the unemployed (differences that were echoed at the ministerial level). According to Vardon (Interview), ‘They played football with my head as the ball’ because they could not reconcile the different policy positions.

DoFA did not have a direct relationship with Centrelink as it had with policy departments because Centrelink’s funding was appropriated overwhelmingly to departments which then purchased services from it. Nevertheless, DoFA’s view of Centrelink was important because as an agency it reflected and influenced the views of key ministers and therefore had a major influence on government agendas. One example of this power was its role in the annual budget process. It provided advice to ministers in cabinet’s Expenditure Review Committee on the proposals brought to it by individual ministers.

Second, DoFA had a strong role in public sector reform for many years. DoFA’s support for the creation of Centrelink was attracted by casting the arrangements in the purchaser–provider form as well as by providing an opportunity for savings in operating costs. The secretary of DoFA viewed the split as comprising ‘the effective re-engineering of DEETYA and DSS to become essentially policy departments purchasing services from other suppliers (either in the public sector or elsewhere)’ (Sedgwick 1996). There was, however, an unfinished agenda from DoFA’s point of view in terms of the full implementation of the model. Under the devolved financial management and accountability arrangements, DoFA was charged with ensuring that the government was obtaining value for money when purchasing services from Centrelink and it retained strong views about the agency and the nature of the institutional arrangements.

DoFA’s position was that there was a need for reform in two areas to enable Centrelink to deliver a more efficient and higher-quality service. Both of these areas were seen as significant purchase issues: the appropriateness of Centrelink’s governance arrangements and corporate form; and the financial arrangements that underpinned these structures. It was clear that DoFA did not see the 1997 arrangements as final. In particular, it was not convinced that representation of purchasers on the board of Centrelink conformed to best-practice governance arrangements. DoFA continued to see Centrelink as a business enterprise. It viewed Centrelink as an ‘income support business’ and part of an ‘income support industry’. Government ownership of Centrelink was seen as comparable with that of Employment National and Telstra and there was preference for a private sector model for delivery.

DoFA supported the position that a contestable environment and genuine purchasing of services produced improvements in quality and price reductions. Its view, however, was that Centrelink did not have genuine competitors. Departments had to use Centrelink to deliver income support services and Centrelink was legislatively required to deliver these services. The purchaser–provider model had not worked with DFaCS because it was not sufficiently at arm’s length.

From DoFA’s perspective, the crucial issue in setting a course for Centrelink’s development was how the government ultimately wanted income support to be delivered. Two broad options were envisaged. The first model was income support operations that were fully integrated into the core business of government, and represented the old DSS, the model the Howard Government had moved away from. The second was income support operations separate from government. If there was no middle position, the obvious conclusion was that stability would be achieved only by either moving back or completing the move away from government. What DoFA apparently could not accept was a hybrid. What it seemed to find particularly frustrating was that, while Centrelink was a separate agency, with a CEO appointed by a board in turn appointed by the government, it was not sufficiently distant from the government. The issue turned on the trade-off between control and efficiency. Locating operations at arm’s length was meant to facilitate improvements in efficiency and service quality through exposure to the market and best business practice at the expense of direct ministerial control.