Managing from the Centre

The role of central agencies continues to be an issue after reform, just as it was before. The three central departments - Department of the Prime Minister and Cabinet (DPMC), Treasury, and the State Services Commission (SSC) - have divested most of the central controls they previously exercised. Central agencies no longer select the accommodation to be occupied by departments, nor do they dictate the personnel or other management actions taken by departments and other entities. Despite these changes, the question of whether the centre should exert any control at all irritates relations between the central agencies and some core departments. Some department managers complain that Treasury is too powerful, others that SSC is too meddlesome. The chorus of complaints that decentralisation has not gone far enough leaves one wondering exactly what the role of the centre should be in managing the State sector.

I take the view that there is a continuing need for strong policy and managerial leadership from the centre. The central departments must prescribe and enforce accountability requirements, and they must deal with matters that cross departmental lines as well as with issues that individual departments cannot handle on their own. The transformed State sector has as much need for policy guidance as it had before reform and an even greater need than before to assure that matters affecting the collective interest are given due attention. Some central functions are carried out differently than they once were, and relations with line departments have changed, but it would be naïve to argue that managerial freedom is incompatible with central direction.

When the State sector reforms were introduced, some departmental managers pleaded for central guidance to assist their transition. They had to make big changes quickly and wanted detailed instructions on how to proceed. Treasury, however, issued minimal guidance and prodded departments to find their own way. Now that the reforms have been successfully implemented, some departments want to be left alone. They regard any substantive involvement by central agencies as intrusive and as an infringement of their freedom to manage. But despite the complaints, one should not lose sight of how far devolution has progressed. There is no comparison of today's central management with that practised less than a decade ago. Moreover, with greater freedom to manage comes the need for greater management of accountability processes and performance monitoring at the centre of government.

In arguing that the central departments should have a strong role in New Zealand management, I do not claim that these institutions are in perfect working order. Some adjustments should be made in these departments, but each should continue to have approximately the same sphere of responsibility. DPMC's main assignment should be strategic management and policy coordination, Treasury's financial management and policy, and advice on economic policy, and SSC's employment of chief executives and advice on the machinery of government. Because these portfolios overlap, it is important that the work of the three departments be closely coordinated. Senior officials from the three departments meet on a regular basis, and consultation among middle and lower officials is frequent and friendly. I see no advantage in formalising relationships that appear to work well on an informal basis.

Department of the Prime Minister and Cabinet (DPMC)

DPMC is a small department with a large responsibility. It handles the flow of papers between the Prime Minister, the Cabinet and its committees, and the departments. Cabinet generally meets weekly, with a full agenda of the documents to be considered or noted. A substantial portion of DPMC's staff resources goes into managing this paper flow, leaving in my judgment insufficient opportunity for strategic policy development. Treasury appropriately considers policy from a financial perspective, and SSC does so from an organisational and human resources view. DPMC's vital niche is to look at policy in terms of the Government's strategic interests. A modest augmentation of its staff would be appropriate, provided that the additional resources were earmarked for broad policy issues with medium term (and longer) consequences. Care would have to be taken to deter DPMC from drawing any new resources it might obtain into the deadline-sensitive chores that now dominate its workload.

In chapter 5, I urge Ministers to promote strategic planning in their departments. It would be appropriate for drafts of these plans to be reviewed by the three central agencies in the context of Strategic Result Areas (SRAs) and Key Result Areas (KRAs). DPMC would have lead responsibility for coordinating this activity and obtaining input from the Prime Minister and Cabinet. This role might be assigned to a section in DPMC, staffed by a small corps of professionals, some of whom would be seconded from other departments and possibly from nongovernmental organisations. This staff would support departmental planning efforts and would be drawn into Cabinet Committee work when strategic issues were on the table.

In recommending an expanded policy capacity, I am mindful that DPMC's staff should continue to be small and unobtrusive. But it is important to strengthen the institution whose primary responsibility is to define the course that policy will take in the future. Inevitably, the Prime Minister's interest will have a strong bearing on how DPMC operates and what it does. A "steady as she goes" Prime Minister will deploy staff differently than might a Prime Minister seeking to reshape national policy. Nevertheless, even a Prime Minister bent on maintaining the status quo faces strategic decisions.


Treasury is a high-performance organisation that effectively leverages its power of the purse to influence a broad swath of public policies and management practices. Over a period of years, it has invested in upgrading the education and job skills of its managers and analysts, and these investments have paid off in Treasury's intellectual and organisational leadership of the reforms.

Some chief executives and senior managers have urged that Treasury take its own organisational medicine and split into two entities, one advising the government on economic and financial policy, the other managing the budget and accountability processes. I believe that both functions should continue to be operated by a unified Treasury. The government does not need a weaker Treasury; rather it needs a Treasury that is sufficiently powerful to maintain financial discipline and managerial accountability. It is highly unlikely that Treasury would be able to tender credible and defensible advice if it were not engaged on a continuing basis in operating the central financial system and overseeing the finances of departments and other entities.

Chief executives and senior departmental managers regard Treasury as the powerhouse of New Zealand government, and some feel inadequately matched against its analytical and organisational resources. But if there is an imbalance of power at the centre of government, the solution should be to strengthen the other organs - DPMC by augmenting its staff and role, and SSC by giving it a clearer focus - not to cut Treasury down to size.

It is difficult to disentangle discussion of Treasury's role from the tough stance it has taken on operating expenditure during the past decade. With support from the government, Treasury has been demanding before recommending any spending increases, even when departmental managers have thought that additional funds were justified by rising workloads. Department managers also argue that Treasury Vote analysts get involved in operational details that should be left to the discretion of the affected departments. These complaints may reflect disappointment over the allocations made in the budget rather than the process itself. Treasury has been demanding and persistent in applying downward pressure on resources; in view of the lack of objective allocation criteria - a matter discussed in chapter 6 - it probably has been somewhat arbitrary in constraining running costs. Allocational rules that do not adjust running costs for inflation, that impose across-the-board percentage cutbacks, as has been the case in several years (but not the most recent ones), or that do not take adequate account of rising workloads, may be especially onerous on small departments that have little room for manoeuvre.

Much of the abrasion in its relations with departments arises out of Treasury's role in defining output classes and in managing the process by which the government allocates resources among the various classes. These are core functions of budgeting, and there is nothing untoward in Treasury involvement. Nevertheless, many chief executives thought that Treasury had stepped over the line several years ago in demanding access to draft purchase agreements to assist in preparing the budget. Treasury took the position that in view of the fact that appropriations are made by output classes, it should be informed on the outputs each Minister intends to purchase; chief executives argued, however, that detailed output decisions were matters between themselves and their Vote Ministers. This issue is part of a much larger question: to what extent are relations between individual Ministers and chief executives (or their departments) of interest to the government? Throughout this report I take the position that the government has a legitimate interest in these matters. It may be the "third party," but it pays the bills and bears the risks. It makes sense, therefore, for Treasury to advise the government on these matters. Fulfilling this responsibility requires that Treasury has adequate information on which to base its advice. In this case Treasury retreated and no longer asks for draft purchase agreements.

In recent years, Treasury has taken steps to clarify its relationship with departments. These year-round relationships go beyond the interactions that occur in compiling the annual budget and the supplementary estimates, and they pertain to the ongoing work and performance of each department. The relationship is specified through two instruments:

  • Core performance expectations common to all departments; and
  • An annual relationship letter setting out the matters Treasury will review in assessing a department's performance and an annual feedback letter reviewing performance on the matters specified in the relationship letter.

The core performance expectations set out the mutual obligation of Treasury and departments to maintain open and trusting channels of communication, purchase and ownership requirements of all departments, and statutory obligations. The list is quite extensive, and is revised each year to incorporate new developments.

The relationship letter is a confidential communication from Treasury to the chief executive specifying the matters to be given special attention during the year. This letter is drafted by Treasury and finalised in negotiations with the affected department. The feedback letter comments on the department's performance in those matters specified in the relationship letter. In effect, it serves as Treasury's audit on assessment of the department's performance in selected areas. A copy of the feedback letter is provided to the State Services Commissioner as input to his assessment of chief executive performance.

As part of the process of vitalising the State sector, it would be sensible to both initiate periodic reviews of Treasury-department relations, and also enhance the current relationship management system. Any change in Treasury-departmental relations should be made only (a) as part of a review of the overall budget and financial management system and (b) in ways that do not weaken the government's control of expenditure totals.

State Services Commission (SSC).

Few departments have been as dramatically affected by the reforms as has the SSC. The Commission is a much smaller department than it was a decade ago, and most of its responsibilities as the employing authority have been terminated. SSC has acquired a critical role in the recruitment and evaluation of chief executives, and it has expanded its machinery of government activities. Despite these and other changes, some critics argue that SSC hardly has changed at all; others characterise it as an organisation in search of a mission. I part company with those who would terminate the Commission and urge instead that it be invigorated and focused on a small number of important responsibilities.

It is widely accepted that managerial control and accountability must reside in departments and other entities where the work of government is carried on, not in central agencies. But although the rationale for withdrawing SSC's operational control is convincing, the extent to which SSC hovers over departments has been greatly exaggerated. SSC does not intervene in daily departmental business, nor does it shadow everything they do. Critics have been slow to acknowledge SSC's progress in focusing on core activities. Some traces of its old culture survive, as do some of its old responsibilities, such as those pertaining to industrial relations in the educational sector. This particular task is assigned to SSC by law, so it has no choice but to devote some resources to collective bargaining.

SSC is responsible for four output classes: policy advice on governmental management; personnel management services; industrial relations; and management services. I would curtail SSC's involvement in industrial relations and personnel management, except to the extent it pertains to the employment of chief executives and senior managers or to broad management issues.

SSC is the appropriate institutional home for two important responsibilities that have to be discharged at the centre. One is to provide advice on the machinery of government; the other is to manage the employment of chief executives. Ministers and chief executives are responsible for the performance of their departments, but they are not well deployed to deal with general issues affecting all departments. SSC does and should continue to provide guidance on government-wide issues. The list of current issues on which its advice would be appropriate is quite long: developing concepts and techniques for specifying outputs and outcomes and for measuring performance; reviewing performance and purchase agreements and recommending government-wide standards; studying the potential impact of MMP on government; recommending changes in the structure or operations of certain departments; evaluating the EEO/good employer performance of departments; assessing the governance and management arrangements for Crown entities. Some of these matters also are considered by Treasury in the course of managing public finance, but the government needs advice that is not framed only by a financial perspective.

The second SSC role has irked some chief executives. SSC is the hub of the process by which chief executives are recruited, employed, and evaluated. Quite a few chief executives have argued that their performance should be the Minister's concern, not SSC's. In their view, if the Minister is satisfied with their performance, that should be sufficient. I argued earlier, however, that the government also is a party to this relationship. SSC represents the collective interest in negotiating terms of employment with chief executives and evaluating their performance. Perhaps SSC should limit its involvement in other personnel matters. But in view of concern over the future supply of senior managers it probably will have to be involved in this matter as well.

I argue in chapter 4 that the chief executive's performance should be evaluated in the light of the department's performance. This means that SSC must have sufficient resources, including information, to evaluate how well departments are doing. A full-blown review of each department's performance should be conducted at least once during the chief executive's term; these reviews should feed into decisions on reappointing chief executives. But SSC also must assess chief executive performance each year, and this responsibility has nudged it to monitor departmental operations somewhat more closely than chief executives would like.

SSC has taken steps to regularise its monitoring and assessment of departmental performance. It issues a standard letter to all departments outlining its performance expectations for the year ahead. The current version of the letter concentrates on progress in achieving departmental KRAs (especially those related to the ownership interest), matters affecting the collective interest, human resource management, and systems of management control. SSC also issues a statement of principles and protocols governing the relationship between it and departments.

Focusing on core responsibilities should allow some shrinkage and reallocation of staff. SSC should be cautious in taking on new assignments, especially those pertaining to the internal management of particular departments. It must be available, of course, for trouble-shooting assignments, but it should remain focused on core activities.

Last modified: