Despite the DSS’s reputed emphasis on accurately interpreting and administering social security legislation, its May 1997 customer satisfaction survey reported that its lowest ratings for regional and counter staff overall were for ‘accuracy of advice/recording and consistency of information’ (63 per cent) and ‘decision-making/responsible for getting things done’ (62 per cent). By November, both of these had risen to 65 per cent. The telephone centre staffers were rated higher on both counts (69 per cent on both in May 1997, and 71 and 70 per cent, respectively, in November 1997).

Centrelink’s strong emphasis on satisfying its customers led to concern in client departments that insufficient effort was being made to achieve high levels of accuracy and to protect the integrity of government outlays. This was not just an issue for Centrelink clients, but was important for customers. Accurate decision making is an essential component of good customer service and is as important as friendly, helpful staff or good queue management. The customer satisfaction surveys indicated that the main concerns of customers regarding the payment process were ‘regularity and timing’ and accuracy of payments. In these areas, Centrelink had performed well, rating 89 per cent for regularity and timing and 78 per cent for accuracy in 2000 and 93 per cent and 79 per cent, respectively, in 2001. Nevertheless, there had been numerous occasions when the media had reported stories of Centrelink errors that highlighted problems related to the capabilities of staff, IT systems and management systems. Centrelink’s public reputation was affected by these incidents.

An ANAO report in 2000 highlighted issues of accuracy. It examined Centrelink’s assessment of new claims for the age pension and questioned the validity of the accuracy measures. The ANAO judged Centrelink’s performance against its own performance indicators, as well as gauging the accuracy of Centrelink’s own reporting on compliance. The audit did not question the reasonableness of the standard or its achievability. The ANAO (2001a:20) concluded that Centrelink ‘could not assure payment at the right rate, from the right date, to the right person with the right product for approximately half of new claims for the Age Pension assessed during the audit sample period’. It also asserted that Centrelink

It estimated the real error rate for new claims at 52.1 per cent, not, as reported by Centrelink to DFaCS and parliament, 97 per cent and 98 per cent accuracy in the two years between 1998 and 2000. The error rate estimated by the ANAO indicated a level of performance well below the 95 per cent accuracy standard agreed in Centrelink’s BPA with DFaCS.

In response, Centrelink redoubled its efforts to achieve the accuracy of decisions and processes expected by the government and its purchaser agencies. Centrelink scrutinised and tightened its quality-control processes under the general banner of ‘Getting it right’. This included independent checking, applying minimum standards, building decision support systems to assist Centrelink officers (although these, on the whole, were not successful), testing technical competencies of staff and delivering a national training strategy via satellite technology.

The ANAO audit also served to highlight the level of complexity of this and other welfare policies, such as the 200 rules governing applications for age pensions. In response, the Minister for Family and Community Services, Senator Vanstone, announced an initiative aimed at simplifying the existing administrative requirements and arrangements for new customers when they first accessed pensions, allowances and ancillary benefits (ANAO 2001a:24). The Rules Simplification Taskforce subsequently reported to the minister in August 2001 on simplifying Centrelink processes and guidelines. Its 20 recommendations included the creation of a customer account, streamlined processes, shorter forms, greater use of electronic information transfer and reduced duplication, and establishment of working groups to further investigate policy simplification (ANAO 2002b).

The ANAO report revisited age pensions in 2002. In the intervening period, DFaCS and Centrelink had developed a new BPA in which a distinction was made between correctness and accuracy: correctness related to decision-making processes within Centrelink’s control that were required for a correct payment outcome; and accuracy included this, but also recognised the obligation of customers to advise Centrelink of changes in circumstances that could affect payment entitlements. Correctness resulted from interactions between Centrelink staff and customers that led to ‘paying the right person, under the right program, at the right rate, for the right date, using the information supplied by the customer. Accuracy requires that these requirements are satisfied every time a payment is made to the customer’ (Centrelink 2002a:89).

The ANAO analysis confirmed a previous finding that no more than 3 per cent of the errors were attributable in whole or part to incorrect processing by Centrelink and concluded that 22 per cent of cases could be attributed to customer error—mostly customers not informing Centrelink about changes in their circumstances, as required by law (ANAO 2002b:17). The government approved a subsequent public relations program to encourage customers to notify Centrelink of changes.

A series of surveys of customers to ascertain their expectations noted the importance of accuracy. Customers generally felt Centrelink met customer expectations in ‘making accurate and timely payments. However, those who have experienced incorrect payments tended to feel an overall lack of trust towards Centrelink’ (DBM Consultants Pty Ltd 2001). Subsequent results showed that this continued to be a problem for Centrelink. In terms of how customers saw Centrelink, in 2001, only 62 per cent of customer survey respondents said ‘it can be relied upon to get things right’. Two years later, this view was held by 66 per cent of respondents. In contrast, customer satisfaction with the accuracy of the advice and information provided to them on their most recent contact was much higher. In 2002, it reached a new high of 80.4 per cent for CSCs and 88.6 per cent for call centres. The overall figure for 2004 was 86.2 per cent (Centrelink 2004a:108).^[8]

Centrelink’s ‘Getting it right’ strategy continued to be developed to improve payment correctness and eliminate preventable rework. Checklists of ‘must dos’ supported minimum standards by setting out the essential steps to be followed to improve accuracy in decision making. New assurance arrangements were developed and incorporated into new BPAs with DFaCS, the Department of Employment and Workplace Relations (DEWR) and the Department of Education, Science and Training (DEST). In an attempt to reduce errors and improve consistency, Centrelink was, during 2004, identifying members of its staff with high levels of specialised knowledge and developing specialised teams able to deal with some of the complex and detailed policy implementation tasks. As already noted, it was moving towards greater consolidation and rationalisation of functions and developing ‘centres of excellence’ for particular specialised services.

Mistakes are costly for Centrelink, administratively and publicly, because they affect the trust customers have in the organisation, impose additional work on staff and use resources better applied to helping customers. Eliminating preventable rework has become an important part of cost effectiveness in the agency. In its first full year, 1998, Centrelink’s customer survey reported that almost 25 per cent of respondents said the purpose of their most recent visit to a Centrelink office was to correct a mistake. By 2003, that percentage had dropped to 15 per cent (Senate CALC 2004:36). To find ways of reducing this further, Centrelink surveyed customers who had contacted Centrelink to have a mistake fixed in the previous four months to find out about the cause and nature of the mistake. Vardon reported the key results to the next Senate Estimates Committee: ‘13 per cent were not mistakes at all. Fifty-five per cent of the mistakes were assessed by one of our staff as Centrelink mistakes’ (Senate CALC 19 February 2004:41). For the remainder, 18 per cent of the mistakes were the result of errors on the part of Centrelink and the customer (so Centrelink was partly or wholly responsible for 73 per cent of errors), 25 per cent were wholly the result of customer error and 2 per cent were primarily third-party errors (DBM Consultants Pty Ltd 2003b:22).

Meanwhile, another report by DBM Consultants into service integration found from a random sample of customers that 14 per cent had contacted Centrelink because of a perception of a mistake, confirming Centrelink’s survey results of 15 per cent as reported to a senate committee in 2004.

In 2004, the issue hit the press with headlines such as ‘Millions of mistakes by Centrelink’ (Australian, 14 February 2004), ‘Centrelink hit by errors’ (AAP, 14 February 2004) and ‘Centrelink’s “victims” turn to food charity’ (Australian, 16 February 2004). According to a reporter, who obtained reports through freedom of information provisions, 1.13 million mistakes had been made over four months in 2003. Centrelink responded with a statement that its customer satisfaction level was at 86 per cent and its accuracy in payments was 96.7 per cent, with only about 20 per cent of the 3.3 per cent of errors directly attributable to Centrelink (AAP, 14 February 2004).

The problem for Centrelink was that the first report had taken the 14 per cent who had perceived a mistake from their random sample of 1065 customers in the Service Integration Survey and extrapolated them across the whole Centrelink service, indicating that more than 700 000 customers would be affected by payment errors (DBM Consultants Pty Ltd 2003b:24). Once multiple contacts to fix the mistake were added in, the media concluded that Centrelink had had to address 1.13 million mistakes in the previous four months. Inevitably, talkback radio had several days of lively debate and the reputation of Centrelink was damaged because it was not able to effectively communicate its side of the story.

Underlying this media scrutiny was a serious story: 14 per cent of all customers had experienced a mistake in the previous four months, equating (said the DBM consultants) to more than 700 000 customers, with some experiencing multiple mistakes. Almost one-third of the mistakes were seen as ‘very serious’ by the customers surveyed, with half the mistakes relating to problems with payments, 23 per cent to inconsistent advice relating to eligibility and 18 per cent to misplaced documentation. While customers attributed the majority of mistakes to the CSCs, most customers (65 per cent) attempted to have the problem fixed by contacting a call centre rather than visiting their CSC. Only 21 per cent initially approached their local Centrelink office. Because there was a range of problems that the call centres could not fix, however, more than one-third of customers were then required to go to their CSC anyway. In all, 22 per cent of customers had to make more than three attempts to have the mistake fixed before this was achieved (DBM Consultants Pty Ltd 2003b:39).

The number of mistakes being made—more than one million in the four months—created a significant and expensive workload for Centrelink, and concern and distress for customers. Customers and Centrelink staff suggested to the consultants a range of initiatives to change problem-creating practices and staff culture. A new emphasis on accurate documentation, increasing the scope of the changes call centres could make and training was instituted.

Centrelink’s customers had a much more positive image of Centrelink and awareness of the services it provided than the general community. Centrelink has struggled at times to win the public relations battle in Australia’s media. The 2004 media frenzy about Centrelink’s error rates was one example illustrating the vulnerability of the agency charged with administering a complex range of benefits programs and services. One chink in the armour was enough to open the floodgates on stories detailing alleged capricious and heartless treatment meted out by agency staff to deserving beneficiaries. It is difficult in the public arena for an agency to communicate the full background of the behaviour of recipients and privacy protocols make it difficult to explain the quality of the relationships they maintain with the agency.

Centrelink was also associated with examples of what was often seen in the community as unpopular, poor or rushed policymaking. The hurried implementation of the new Youth Allowance program in 1998 and new contract arrangements for the Job Network in 2002 and 2004 were such examples. The introduction of the family tax payments in 2004 again exposed Centrelink to the press, though there was some recognition that most of the confusion was due to rushed policymaking on the eve of an election rather than to any failure of implementation.