There have always been three critical issues to be addressed in establishing working performance agreement, assessment and pay systems for public servants undertaking day-to-day activities associated with regulatory programs or service delivery. The first is identifying useful performance indicators. The second is applying those indicators in ways that encourage rather than discourage flexibility and initiative. The third and more fundamental problem is establishing a ‘clear line of sight’ between the government’s goals, the agency’s goals, the line area’s goals and the individual’s personal performance agreement and indicators. As Figure 2 suggests, these are meant to cascade, with ‘a focus on managing performance to meet the Government’s required outcomes’.^[45]

Maintaining a clear line of sight to ‘government’s required outcomes’ as agreed between the secretary and the minister should in fact have the effect of clarifying those ministerial policy goals that should concern public servants and those political goals that should fall outside individual performance criteria (always assuming the ministers in question are among those who can ‘distinguish between their own political aspirations and the duties of a Secretary’^[46]). Staff were to be assisted in making these distinctions by corporate planning and governance statements (set to the side in the Figure 2 but meant to help shape the flow) articulating the APS Values and behaviours desired of them in meeting their performance requirements. Take two such statements from agencies responsible for driving performance as the system was evolving: the Departments of Finance in 2001–02 and Employment and Workplace Relations in 2002–03. (New Finance Valued Behaviours, it should be noted, were announced on 1 July 2003 some time after a change of Departmental Secretary following the 2001 election. The New Finance Valued Behaviours (Figure 3) were said in the annual report of that year to ‘more closely align Finance values with APS values’. The fact that the APS Values changed following the departure of the secretary—who then moved on to DEWR—is an indication of the importance an agency’s senior management can have in the conduct of performance management as plans cascade down to agency employees.)

Figure 2

Figure 3

Even setting aside elements of ‘advertising-speak’ in the Finance vision statement, it is certainly made clear that, as the Management Advisory Committee advised, ‘ministers are the key client’.^[49] Whatever public servants do, it is in the service of the minister-as-client or minister-as-customer—a metaphor that does little to clarify the special relationship between the administrative and political arms of government that is part of the Westminster tradition.^[50] Ethics and integrity appear to have a role in the conduct of public servants, certainly, but their particular application to ministerial relations is not evident from the documentation provided, apart from what can be inferred from the allusion to the Public Service Act. The overall ‘takeaway message’ from early corporate guidance of this sort was ambiguous at best, and could prove more likely to encourage than discourage a very broad view of client responsiveness, particularly when supported by individual performance agreements incorporating key performance indicators requiring officers to ‘meet their minister’s requirements’.

For public servants involved in delivering programs, then, particularly those working outside Canberra (the majority of all public servants), the line of sight was more likely to be directly to those outcome statements directly relevant to their program than to the minister’s direct requirements. This also applies to many public servants exercising regulatory oversight, where decision making can involve the exercise of a discretion, as the Public Service Commissioner has argued:

Agency outcome statements advise on the ‘policy outcome sought’ through program administration or through any discretion to be exercised by the regulatory decision-maker. Take, for example, the outcome statements of DMIA for the period 2001–02. Outcome 1 is specified as ‘Contributing to Australia's society and its economic advancement through the lawful and orderly entry and stay of people’ and output component 1.2.2 (Protection Visas (Onshore)) is ‘To ensure that Australia efficiently and effectively fulfils its international obligation not to return, directly or indirectly, refugees to their place of persecution’.^[52] Now take the Prime Minister’s broad vision of the same function, also in 2001:

In the political space between the Prime Minister’s statement—which set the tone for the 2001 election—and formal agency output indicators a whole series of silent assumptions might be made by public servants. The Prime Minister’s words were not an authoritative statement of Government election policy, but they were ‘the [2001 election] campaign slogan, the television message for the last weeks, the full page ads being prepared for the newspapers, the posters being printed for distribution from one end of Australian to the other’.^[54] No doubt many DIMIA staff, particularly at more junior levels, would be considerably more familiar with these very widely circulated statements than they would be with their departmental outcome statement. In fact their subsequent behaviour suggests that not only was this the case, but that agency senior staff also took no steps to remind them of the difference between the two. According to the report of the Inquiry into the Circumstances of the Vivian Alvarez Matter (Comrie Report),

Without formal confirmation, what a minister says in a media interview, or even in an election advertisement, should not, of itself, carry authority. On the contrary, departmental outcome statements, which have been agreed with the minister, should provide the policy guidance. But to the extent that there was no clear connection between the emphases being adopted in ministerial public commentary and in agency documentation, what public servants may be expected to do can only be a matter for conjecture. In the absence of guidance from senior management to the contrary, this will generally involve assuming that either the prime ministerial media commentary or the agency outcome statement or both involve considerable spin for the consumption of particular audiences, but that on the whole it is better to mouth the agency’s words (on the assumption that this is what they are there for) and obey the Prime Minister’s signals (on the assumption that this is what both the Prime Minister and the organisation want). In the end, the exercise of regulatory oversight was considered by Comrie ‘unreasonable and therefore, by implication, unlawful’.^[56]

The report of the Palmer Inquiry, which also looked at the conduct of regulatory decision making at DMIA, argued that ongoing reliance on implicit rather than explicit direction fosters the development of an ‘assumption culture’ in which, because some things are assumed to be right, others must be assumed to be wrong:

Such a culture is defensive because in fact it cannot articulate a fully formed defence. The rationale for its behaviour is somewhere in the difference between the Prime Minister’s construction of the prerogatives of government and the organisation’s understanding of its international responsibilities. Defensive behaviour is simply a means of turning the attack away from those inside the agency ramparts and towards those outside them: either those outsiders who are still numbered among the politically correct (which links the agency’s behaviour with generalised responsiveness to government) or those who are uninformed regarding process issues (which covers practically everyone outside the agency itself).

When public servants are caught between implicit political direction and explicit organisational directives, and when agency leadership does not extricate them by saying something unambiguous, they will tend to put their heads down and focus only on their own individual performance indicators. The result in DIMIA’s case was diagnosed as an ‘environment in which people are unwilling to accept ownership of matters beyond their immediate responsibilities, regardless of the importance of the matter and the obvious need for continuity in its management’.^[58] Even in more benign environments than DIMIA, individual performance indicators can distort the clear line of sight to organisational objectives given the tendency of the measurable to drive out the important. Thus, for example, team leaders at the Australian Taxation Office (ATO) reported that ‘part of our message to our staff is, do our work, do it well and pass back to the other business lines what is theirs, because we’re not funded for it and all it does is make our performance look bad’.^[59] While this practice is clearly consistent with their low-level performance indicators, it is not consistent with ‘the overall ATO value of providing a responsive and integrated service’.^[60] In the case of DIMIA, immigration policing and detention were in separate areas of the department, and for the public servants concerned they were in separate performance agreements. Had those individuals felt their first duty was to contribute to ‘the lawful and orderly entry and stay of people’ consistent with outcome 1, they may have identified some process issues, but in fact according to Palmer, ‘the predominant, and often sole, emphasis [was] on the achievement of quantitative yardsticks rather than qualitative measures’.^[61] Due process lapsed into simple processes, applied against a background of high profile media commentary from the Government on ‘terrorists against the rest’.

The capacity of low-level performance indicators to re-introduce a process-driven approach to implementation and break the line of sight to the high-level agency outcomes they are notionally intended to deliver was also identified by the ‘Independent Committee of Review of Breaches and Penalties in the Social Security System’ in relation to service delivery at Centrelink (considered in more detail in Chapter 6). On the face of it, the problem lay with the low-level performance indicators themselves. In this case indicators of timeliness in handling reports on jobseeker activity appeared to have undermined other service delivery criteria, like attention to the regulatory environment and the application of procedural fairness:

This extract from the findings of the Independent Review would appear to report a commonplace conflict between efficiency and effectiveness (or efficiency and equity), or the cost and quality of service. But the decisions in question relate to the reduction or termination of welfare benefits for failing to meet certain work activity tests (referred to as ‘breaching’ in reference to the ‘breaching provisions’ of the Act). For welfare recipients, and for the integrity of the welfare system as a whole, breaching is a sensitive matter, and the Committee of Review was concerned that corners might be being cut in this area because of the existence of a quota of breaches to be identified and upheld. Even had sufficient resources been provided to deliver a fair process (a matter at issue), it was felt that the quota established a bias towards the finding of a breach. In this case, the assumption-based culture arose because of the displacement of agency standards by performance-criteria goals set through contractual arrangements with another agency for the delivery of welfare services.

The impact of contestability and outsourcing on relations between the administrative and political arms of government is considered in more depth in Chapters 2 and 6. The point to be considered here in connection with individual performance management is that, for the public servants making decisions about the breaching of welfare recipients, contractual quotas on breaching were clearly consistent with public statements of the minister responsible for the portfolio administering the contract:

Given the public profile of statements of this nature,^[64] it is not surprising that public servants at Centrelink were focused on quotas for breaches, regardless of the fact that such quotas encouraged them, in the words of the Independent Committee of Review, ‘to effectively override or subvert legislation, policy instructions, or other standards and goals’. The agencies concerned subsequently pointed out that their employees had misunderstood the intention and nature of the contractual quotas. If it was a misunderstanding, it was certainly a pervasive one, and found by the Independent Committee of Review to be so. Were public servants listening to a dogwhistle that was not really there? If so, how did it become a performance indicator?

‘Assumption cultures’ are by their nature entrenched over time, and an agency’s leadership has a considerable capacity to either reinforce selected assumptions or to let them develop in default of any public guidance or, conversely, to provide such guidance and to ensure the system takes proper account of values, relationships, and behaviours as well as results. Take, for example, the case of a briefing note circulated to Centrelink staff that pursued ‘broad principles’ supported by the minister but remained altogether unknown to the minister until after an ‘internal briefing note’ was made available to the ABC:

Whatever the resonance of the ‘briefing note’ with the electorate, it is difficult to reconcile the instruction to tell clients ‘if you can't get a job in Sydney this year, you never will, and you're not serious about work’ with the requirement under the Section 10(1)(g) of the Public Service that ‘the APS delivers services fairly, effectively, impartially and courteously to the Australian public, and is sensitive to the diversity of the Australian public’. Nevertheless, although the actual author of the document is not known and its consistency with formal agency directives is ambiguous, it was evidently believed by some staff to have senior management support and was therefore accorded formal status as accepted operating policy.

Finally, take the more recent case of the Workplace Authority, an agency whose strategic objectives are to ensure that:

In this case the line of sight from employees to the agency’s strategic objectives lies through the behaviour modelled by the agency head. This behaviour is considered in Chapter 2. Its tendency to support the agency’s formal strategic objectives is patchy. While the Howard Government may be assumed to have had confidence in the Authority as a consequence of the director’s becoming ‘such an important component of the advertising’^[67] this was not the case with the Parliament more generally.^[68] And the controversy generated by the advertising was also unlikely to have made a positive contribution to widespread ‘confidence about making an agreement’. Nor did it produce confidence more generally in the professionalism of the APS; on the contrary, it ‘reinforce[d] the perception of a service giving too much weight to responsiveness over apolitical professionalism, and undermine[d] the confidence of the Opposition in the capacity of the service to support its policies and programs should it become the Government in the coming election’.^[69]

How are the employees of the Workplace Authority to understand their obligation to ‘provide straightforward information in response to … questions on all aspects of workplace relations?’^[70] How can they expect to be judged on whether their performance meets agency’s core standard of ‘professionalism’?^[71] No wonder there is some confusion (or cynicism) among public servants about the ‘line of sight’ from high-level agency strategic objectives to their own understanding of what is actually required of them by their senior managers and ministers. No wonder that, while public servants from both policy and service delivery agencies place a high value on regular performance feedback,^[72] they are overwhelmingly negative about the system of annual performance agreements complete with personal targets and indicators and, in the great majority of cases, with numerical rating scales associated with pay outcomes. Importantly, they seem to detect a gap between the theory and practice of system, between what it calls for and what it actually expects—a problem, according to the Australian National Audit Office (ANAO), that boils down to credibility:

In 2005–06, half of all APS employees responding to the State of the Service Employee Survey did not agree that the performance pay system in their agency operated fairly and consistently; and less than a quarter believed that the pay outcomes of performance assessments accurately reflected differences in individual performance (Table 1). These responses follow 2004 survey findings from the ANAO that many public servants were calling for more quantifiable performance indicators, as if that would simplify and bridge the gap between the theory of the system and their experience of its operation.^[74]

Note: Respondents who answered 'No' or 'Not sure' to question 68 were not asked this question. The percentage of relevant respondents who did not answer this question in 2003 was: a – 0.2%, b – 0.3%, c – 0.3%, d – 1.2%, e – 1.2%, f – 1.6%, g – 1.0%. The percentage of relevant respondents who did not answer this question in 2004 was: a – 1.2%, b – 1.0%, c – 1.2%, d – 1.6%, e – 1.2%, f – 1.3%, g – 1.2%. The percentage of relevant respondents who did not answer this question in 2005 was: a – 1.0%, b – 1.1%, c – 1.1%, d – 1.1%, e – 1.0%, f – 0.7%, g – 1.2%. The percentage of relevant respondents who did not answer this question in 2006 was: a – 0.7%, b – 0.8%, c – 1.5%, d – 1.0%, e – 1.1%, f – 1.0%, g – 1.1%.

More broadly, public servants responding to both the ANAO and the State of the Service Employee Surveys did not see access to performance-related pay as a significant factor in improving their productivity as promised by NPM theorists.^[75] In fact, when those who felt their productivity had improved over the year were asked by the Australian Public Service Commission to rate the contribution of performance-related pay to that improvement they ranked it at the bottom of the 16 factors from which they were asked to choose.^[76]

There are no consistent trends in how different classifications of employees respond to having their performance broken down into indicators and assessed, although the 2005–06 State of the Service Report found that overall employees in executive level classifications were most negative.^[77] This is consistent with evidence that longer-serving, experienced public servants find the system more distasteful than new public servants.^[78] There is, however, considerable variation between agencies in employees’ views on the operation of individual performance assessment systems. This suggests that the credibility gap may be more evident in some agencies than others and is consistent with the scope for variation between agencies in what people believe themselves to be explicitly and implicitly expected to do to win performance pay; how transparently assessment processes are considered to work; and the extent to which assessment is linked to financial outcomes through rating scales. Among large agencies (for which statistically reliable employee responses are available), employee assessments of agency assessment arrangements varied markedly against a range of criteria. State of the Service reports indicate that:

It is not possible to push the analysis further and identify those agencies whose employees exhibit high levels of scepticism in relation to their performance assessment systems; neither does the State of the Service Report provide data which can be used to establish whether there are correlations between employee views of performance assessment systems and challenges they have experienced in dealing with their own portfolio ministers and their ministerial advisers.

It is, however, possible to look generally at whether employees of large agencies experienced performance assessment and pay systems as supportive of the range of values they are legally required to uphold under the Public Service Act. In 2006, between 23 and 55 per cent of employees in large agencies agreed that the performance pay system in their agency contributed to a workplace that upholds the APS Values.^[80] Conversely, 62 per cent of all employees surveyed did not agree and 22 per cent actively disagreed. This is at least indicative of a view that no matter how clearly the ethical framework calls for balance between the APS Values, the performance system itself does not generally appear to reinforce it. Of course these may not always be the APS Values relating to direct interaction with ministers and their advisers, but in a system of ‘managing for results’ the views of ministers and advisers will often work in the background. For some policy advisers, performance criteria may have been assumed to call not only for an understanding of the government’s policy priorities and the use of its preferred discourse but also for the anticipation of its unspoken preferences. For some public servants who deliver services and administer regulations, performance criteria may be assumed to call not only for a focus on low-level indicators but also for a rule-bound and punitive approach to groups that fall outside ‘mainstream Australia’.

So, there is much ambivalence among public servants over the credibility of performance pay regimes. Some believe that they are there to signal an over-zealous approach to ‘Meet[ing my] Minister’s requirements for my area’s Ministerial and Parliamentary Business’; at the other end, some believe that they are simply a means of offering retention bonuses in disguise. No doubt agency systems and practices are equally variable. What employees do know is that in 2005 the Council of the Order of Australia saw no inconsistency in awarding an Australian honour—an Officer in the Order of Australia—to the Secretary who presided for seven years over ‘practices [that] have been in operation for a long time and seem to have given rise to an immigration detention culture that, in the opinion of the Inquiry, constrains thinking, flexibility and initiative and concentrates on functions, process and quantitative measurement to the detriment of the achievement of policy outcomes’.^[81] The appointment was for ‘service to the community through contributions to Australia’s international relations and to major public policy development including domestic security, border systems, immigration, multicultural affairs and Indigenous service delivery’.^[82] At the same time, the Government appointed Mr Farmer to one of Australia’s most important diplomatic posts (Indonesia) as ambassador; shortly after, it ushered in ‘a major organisational change process’ led by a new secretary.^[83]

Such an obvious disjunction helps explain public servants’ concerns about the credibility of decision making in performance assessment, at least in some agencies, and that in those cases informal messages can be more important than formal ones when performance is being assessed and rewards (and punishments) are being handed around. Ministerial appeals to the electorate constitute an important source of such informal messages, and being ‘informed and improved by ongoing discussions with political advisers’^[84] constitutes another. How these messages shape the behaviour of individual public servants will depend on the agencies in which they work—including the leadership in the agency and the quality of its performance management arrangements—as well as their own position in the food chain and their own susceptibility. The public service has been devolved, and agency-specific systems have grown considerably in their capacity to influence individual behaviour. This is the issue that is addressed in Chapter 4.