The initial task required the new organisation to bring together the staff of two departments with different expectations and cultures into one organisation, and to rationalise the delivery services of these two departments that were now responsible for overseeing the delivery of their policy program outcomes by one new agency. These departments had a long history of rivalry and turf warfare so the potential for tension, conflict and dispute was considerable.

The different cultures and expectations of each parent department would gradually lead to different relationships developing: one more contractual, the other more of a partnership or strategic alliance. Within this structure, Centrelink was subject to more than one minister, with a decision-making board appointed by and responsible to the minister of its largest client portfolio department, yet without control of its budget or staffing numbers. These were the prerogative of its client departments, which determined its outputs and performance measures through purchaser–provider agreements. In time, Centrelink would extend service provision to other clients, federal and state.

There was a range of issues that could create tensions. Centrelink was established as a customer-focused agency, which offered new levels of integrated customer service and was promoted as the human face of government, but it was also required to save money through the amalgamation of the two service networks of its parent departments. It would, as Sue Vardon put it, need to adopt a businesslike approach that was ‘harmonious with the aspirations of human service’ (Vardon 1998d).

As the Prime Minister (Howard 1997:2, 3) said when he launched Centrelink in September 1997:

Improving services for customers became a high priority. The initial emphasis was placed on improving staff attitudes and shopfront offices, and later on developing clearer options for customers and broadening technological access.

An agency through which such a large proportion of funding was transmitted and with such a large, and often economically vulnerable, clientele was always going to be of interest to politicians. Any weaknesses, errors or failures would be noticed and reported in parliament and the media sooner and louder than failures in smaller organisations with less vulnerable customers. Centrelink became an agency whose every move was scrutinised.

There were also possible tensions in being a public service agency. Minister Philip Ruddock (1996:7624) indicated that Centrelink would be a statutory agency with public service responsibilities and subject to the mainstream legislation for agencies as well as the Audit Act and finance regulations and directions. As a statutory agency, like the Australian Public Service Commission, Centrelink has operated under the Financial Management and Accountability (FMA) Act 1997, which brings with it a high level of accountability. The chairman of its board of management was the chief executive for the purpose of the FMA Act and was given specific powers in the legislation establishing Centrelink. This included direct accountability to the portfolio minister.

These accountability and reporting mechanisms were quite complex, with the CEO, appointed by and reporting to the board, also reporting directly to the minister (or ministers) of the client departments while also being responsible for delivery of programs for which their departmental secretaries were funded and accountable. Ministers had the power under Centrelink’s legislation to intervene directly in Centrelink’s affairs. Attendance at Senate Estimate Committee hearings was therefore a joint activity with officials from Centrelink and the Department of Family and Community Services (DFaCS) in attendance to respond to questions. There were also potential tensions for Centrelink as a public service agency staffed by people transferred from its parent departments, but also expected to be an agency able to operate in businesslike ways that were entrepreneurial and alert to expanding its business.

A further challenge was the separation of policy and service delivery, which was unusual in Australian public administration. Another example of where the two had been separated, the service delivery organisation of the Australian Taxation Office, had just one policy department, the Treasury. Centrelink began with two clients and had three after one year, with the expectation that this number would grow in time.

Governments invariably modified and amended social welfare and employment policies during their tenure so a degree of policy change was anticipated, but not the radical changes that occurred soon after Centrelink’s inception. The government undertook a wholesale review of social welfare, and the policy decisions taken in 2001 flowing from that review substantially changed how Centrelink needed to operate and significantly extended its core business. These changes galvanised Centrelink’s role in designing the implementation of the new policies. They also obliged the policy departments to take into account and modify policies when given feedback from Centrelink about how policy changes were affecting recipients or community service providers. Managing the policy/service-delivery balance was always likely to be difficult, even more so in times of a complete overhaul of social welfare and employment policies.

In time, these administrative issues and varying expectations would drive Centrelink to behave and develop in particular ways. There were opportunities for conflict between the different objectives set for the new organisation. Moving in one consistent direction was going to be difficult as there were several imperatives to be attended to. These were worked through in several stages.