Back to top anchor
Working paper

Striking a Balance: Centralised and Decentralised Decisions in Government (WP 02/15)

Abstract

This paper identifies factors to be looked at when considering the extent to which decisions within government should be centralised or decentralised. In practice, the solution is almost always likely to involve a balance between centralised and decentralised decision-making. Nevertheless there are a number of common factors that are generally applicable to questions of centralisation and decentralisation. This paper identifies those factors in order to provide some guidance for decisions regarding the location of decision rights. Rather than being prescriptive, the paper simply presents the relevant issues for consideration. Centralisation (or decentralisation) is a complex and multi-dimensional issue. It is partly for this reason that the paper does not suggest any specific solutions. The solution in any particular case will involve tradeoffs between the factors identified in the paper as well as value judgements regarding the ranking of the various factors.

In determining the appropriate balance between centralised and decentralised decisions, various factors are in tension. Centralisation can help ensure uniform and consistent standards, minimise inequalities, avoid the duplication of services, allow for the achievement of economies of scale, and increase coherence and coordination. Decentralisation, on the other hand, can help enhance local autonomy and empowerment, encourage customisation and innovation, and increase participation.

Economic, social, managerial and constitutional perspectives offer a number of theoretical frameworks that are useful in considering issues of centralisation/decentralisation. This paper draws together ideas from across these perspectives. It concludes that the solution is likely to be characterised by a “tight/loose” pattern whereby there is “tight” or centralised control over the major objectives that is then joined by “loose” or decentralised discretion over the ways in which those objectives are achieved to varying degrees.

Acknowledgements

This paper was originally drafted for the Chief Executives Reference Group on Central Agencies. It has benefited from comments received by this group as well as comments received at a seminar held by the New Zealand State Services Commission on 25 June 2002. In addition, helpful comments have been received on earlier drafts of this paper from Richard Norman, David Galt, Simon Macpherson, Andrew Kibblewhite, Jonathan Ayto and Veronica Jacobsen.

Disclaimer

The views expressed in this Working Paper are those of the author(s) and do not necessarily reflect the views of the New Zealand Treasury. The paper is presented not as policy, but with a view to inform and stimulate wider debate.

1  Introduction

This paper identifies factors to be considered when thinking about the extent to which functions within government should be centralised or decentralised. The intention of the paper is to assist those considering questions about the extent to which functions such as human resource management, purchasing, or specific output delivery should be centralised, both across the public sector and within agencies. The paper does not suggest specific solutions, which will depend on the particular decision and circumstances in each individual case. Often the solution lies in a “tight/loose” fit between centralisation and decentralisation[1]. By this “tight/loose” fit I mean a pattern whereby there is “tight” or centralised direction and coordination over the major objectives and “loose” or decentralised discretion and autonomy over many of the ways to achieve those overall objectives.

This is a high level paper that attempts to highlight the relevant issues to consider when thinking about centralising or decentralising decision rights rather than anything more prescriptive. Centralisation (or decentralisation) is a complex and multidimensional issue. It is partly for this reason that the paper does not suggest any specific solutions. The solution in any particular case will involve tradeoffs between the factors identified in this paper and is likely to involve some value judgement regarding the ranking of the various factors. This paper aims to identify the relevant issues to consider when faced with a (de)centralisation question and in doing so, should provide a pathway through some of the complexities involved.

The paper seeks to answer the question: “According to what factors should authority be allocated?” Generalisations in this area can be dangerous and the solution is almost always likely to involve a balance between centralised and decentralised decision-making. Nevertheless there are a number of common factors that are generally applicable to (de)centralisation questions. This paper identifies those factors in order to provide some guidance for decisions regarding the level of centralisation or decentralisation.

In determining the degree of centralisation, various values are in tension. Advocates of centralisation often contend that it ensures uniform and consistent standards, minimises inequalities, avoids the duplication of services, allows for the achievement of economies of scale, and increases coherence and coordination. Advocates of decentralisation, on the other hand, point to the enhancement of local autonomy and empowerment, greater customisation and innovation, and increased participation as benefits of decentralisation (Boston, Martin, Pallot and Walsh 1996).

Economic, social, managerial and constitutional perspectives offer a number of theoretical frameworks that are useful in considering issues of (de)centralisation. This paper draws together arguments from across these perspectives and identifies common factors applicable to (de)centralisation questions.

Each of these factors points either to centralisation or decentralisation of decision-making, all other things being equal as shown in Tables 1 and 2. Resolving (de)centralisation questions requires tradeoffs to be made (such as that between the agency costs of inconsistent goals that increase as decentralisation increases, and the costs of gathering and processing information, that decrease as decentralisation increases). Different factors may dominate depending on the particular decision and circumstances.

Table 1 – Factors implying centralisation
Factor Examples
Decisions requiring knowledge of central information, that it is not practical to transfer, an understanding of the needs of the system overall, or knowledge of a number of different organizations and non-localised tradeoffs Whole of government resource allocation
The need for accountability for key judgements to remain clearly with the democratically elected government Key judgements about public policy values, trade-offs between organisations, broad resourcing and taxation issues
The need for the government to manage processes closely in order to manage risk more directly Management of an international relations crisis
Decisions where it is possible to achieve economies of scale New Zealand Debt Management Office’s overnight cash “sweep”[2]
Decisions where it is possible to achieve economies of scope “Brokerage” roles, budget-setting functions, spreading “best practice”
The need for coherence and coordination across the public sector Leadership on core public policy values
The need for uniform and consistent national standards and guidelines Benefit entitlement and eligibility rules, tax administration and rulings
Issues where it is difficult to specify the contract or provide suitable incentives for a subordinate to carry out government objectives Defence

Often, the solution to questions of (de)centralisation is likely to involve a combination of centralised and decentralised functions. This solution is likely to be characterised by a “tight/loose” pattern whereby there is “tight” or centralised control over the major objectives that is then joined by “loose” or decentralised discretion over the ways in which those objectives are achieved to varying degrees.

Table 2 - Factors implying decentralisation
Factor Examples
Localised non-transferable (or transferable only at a prohibitively high cost) information, knowledge and relationships Judgements about supplementary benefit payments
Issues requiring customisation, innovation and a flexibility to respond to localised conditions Business development initiatives
Situations where there is a span of control problem Specific functions are decentralised from doctors to nurses (and from nurses to nurse aides) as doctors are unable to make all the decisions relating to the care of their patients
Issues where community involvement, local empowerment and local participation are important Community employment schemes
The need to avoid an unnecessary concentration of power and the risk of abuse of power by distancing a decision from political or central bureaucratic control Specific functions may be decentralised to limit the power of Ministers to other entities such as the Police Complaints Authority and the Commerce Commission (decisions under the Commerce and Fair Trading Acts)

This paper discusses questions of (de)centralisation in more detail, providing some background on the various perspectives, (economic, social, managerial and constitutional), the theoretical frameworks within each of these perspectives, and the arguments that can be drawn from them. From these arguments common factors relating to questions of (de)centralisation are identified.

Notes

  • [1]The concept of a “tight/loose” fit between centralisation and decentralisation originated with Peters and Waterman, 1984.
  • [2]The New Zealand Debt Management Office (DMO) is the debt manager of the central government of New Zealand. DMO manage the risk associated with the Government’s fixed income portfolio. The account balances of all New Zealand dollar Crown and departmental bank accounts are consolidated at the end of each banking day. The net balance is then moved to the Crown settlement account at the Reserve Bank of New Zealand. This practice (called the overnight cash ‘sweep’) makes use of economies of scale to help achieve efficient cash management.

2  What is decentralisation?

There is much debate concerning the definition of “decentralisation” (or “centralisation” as the converse), especially with regard to how its definition differs from that of “devolution” or “delegation”. For the purposes of this paper decentralisation is taken to mean the locating of decision-making rights or responsibilities away from the centre, whether this be to a Crown entity from a Minister, to a department from a central agency or from a department’s head office to a regional office. Typically decentralisation involves the spreading out of decision-making authority from a smaller to a larger number of actors (Pollitt, Birchall and Putman 1998). The discussion in this paper concerning the allocation of decision-making rights applies also to the decentralisation of operations, resources and the allocation of rights more generally. These rights might include: the right to initiate or propose, the right to ratify, the right to veto, the right to set conditions, the right to determine in the event of a dispute, the right to deal with exceptions, the right to allocate resources and so on.

Centralisation (and decentralisation) should, for the purposes of this paper, be interpreted in a broad sense. That is, centralisation does not necessarily refer solely to decision-making by central agencies or ministers. For example, centralisation may occur at a local level or at a regional level. The location of decision rights (centralised or decentralised) is not intended to be a linear concept for the purposes of this paper. Rather, it may be multi-dimensional with the solution to (de)centralisation questions likely to lie in a combination of centralised and decentralised decision rights.

The issue of centralisation versus decentralisation is particularly relevant in the context of the new communications and information technologies becoming available, as well as in light of the Review of the Centre report[3]. Improved communications and information technologies change the costs of information and of information exchange. This may in some cases mean it is easier to centralise decision-making, as information is more easily transferred to the centre, similarly, in some cases it may make it easier to decentralise decision-making.

This study aims to provide a building block for thinking about centralisation or decentralisation decisions coming out of the Review of the Centre (particularly in the areas of human resources, central agency leadership and e-government).

Notes

  • [3]“The Review of the Centre” report presents the findings and recommendations of a review of the centre of the New Zealand State sector conducted over a four-month period in 2001 by a Ministerial Advisory Group.

3  Factors influencing (de)centralisation

Economic, social, managerial and constitutional perspectives offer a number of useful insights when considering questions of centralisation or decentralisation. These perspectives are briefly described in the Appendix. As a summary of the four perspectives we might say an economic perspective maximises effectiveness and efficiency while a social perspective seeks, in addition, to enhance social well-being through mechanisms such as participation. A constitutional perspective introduces the notions of democracy, separation of powers and the rule of law, thus, in considering centralising or decentralising decision rights, the current legal and constitutional frameworks must be taken into account. A managerial perspective emphasises giving managers the authority to make decisions and holding them accountable. This paper draws together approaches from across these perspectives to identify the relevant issues to consider when thinking about centralising or decentralising decision rights.

Centralisation (or decentralisation) is a complex and multidimensional issue. Without seeking to understate the complexities, this paper seeks to identify the issues and provide a pathway through some of them. The paper first discusses centralisation of decision-making rights and then decentralisation of decision-making rights. Sometimes, as in the case of information asymmetries, different sides of the same coin can lead in different directions.

3.1  Centralised information, knowledge and perspectives

There are a number of sources of information asymmetries that mean in some circumstances centralised decision-making can enable better, more timely and more appropriate decision-making. These information asymmetries are most prevalent in decisions relating to the interest of the nation as a whole, judgements that impact on the whole system and decisions regarding key public policy values and broad taxation and resourcing issues. Through the use of the information available at the centre (regarding government objectives, high-level direction, foreign policy, the availability of resources etc.), centralisation may permit a greater level of responsiveness to government objectives. It also means that decisions requiring such information (eg, an understanding of the needs of the system overall, or of different organisations) may be better taken at the centre.

3.2  Constitutional issues

In some circumstances there may be a constitutional imperative for elected representatives to be directly responsible and accountable for decisions, such as those relating to key government objectives and broad taxation and resourcing issues. This constitutional argument for centralised decision-making may be in part motivated by a concern to address the informational asymmetries that favour centralisation, as well as concerns about legitimacy, buy-in, accountability and incentives.

Centralisation of decision-making rights can be used to ensure accountability remains with Ministers who have the democratic mandate to undertake decisions. Such control over decision-making allows the government to more directly pursue its objectives and control the allocation of resources through the closer coupling of taxation and expenditure decisions. This may be of particular importance for decisions relating to key judgements about public policy values, tradeoffs between different organisations, and broad resourcing and taxation issues.

3.3  Risk management

Centralisation may also allow the government to manage processes more closely and thus manage risks more directly. Central risk management may be appropriate in many cases due to the information that is held at the centre. For example, the budget process may be managed at the centre as the centre has access to information spanning across the public sector and, as a result, is better able to manage risks. However, there are also incentive and control considerations regarding risk management. It may be the case that Ministers have better incentives to manage risk than departments and other decentralised decision-makers. This might explain, for example, why only the Minister of Finance has the authority to raise a loan on behalf of the Crown (see also the related discussion later in the paper on incentive problems).

Table 3 – Information, constitutional and risk management factors
Factors Examples
Decisions requiring knowledge of central information, that it is not practical to transfer, an understanding of the needs of the system overall, or knowledge of a number of different organisations and non-localised tradeoffs imply centralised decision-making. Whole of government resource allocation
The need for accountability for key judgements to remain clearly with the democratically elected government implies centralised decision-making. Key judgements about public policy values, tradeoffs between organisations, broad resourcing and taxation issues
The need for the government to manage processes closely in order to manage risk more directly implies centralised decision-making. Management of an international relations crisis

3.4  Economies of scale and scope

The achievement of economies of scale and scope may also suggest centralisation of some decision rights. Economies of scale can arise when similar operations are combined in order to achieve increased output without increasing fixed costs. For example, combining “back office” functions may be achieved in a decentralised way (eg, through shared services). However, the achievement of economies of scale often requires some centralisation of decision rights due to the level of coordination required across the sector. This is particularly the case when individual agencies have only weak incentives to look for efficiencies through collaboration (as an individual agency receives only a fraction of the total efficiency gain yet could bear the majority of the costs from attempting to set up collaborative arrangements).

Economies of scope, on the other hand, can arise as a result of combining different operations and functions. The gains from economies of scope are closely related to the information asymmetries that favour centralisation. For example, there are economies of scope in having a centralised budget process where information across the public sector is available in budget setting. Furthermore, there are economies of scope in combining the budget-setting process and macro-economic forecasting within the one organisation. The complementarities between economic and fiscal policy functions lead to the achievement of economies of scope.

It may, however, be possible to realise economies of scale and scope through careful structuring and contracting of activities, still allowing decentralised decision-making. For example, in some cases the benefits of economies of scale can be purchased from the market through outsourcing. Gupta and Eerola (1997) note that the ability to obtain economies of scale and scope may provide enhanced opportunities for innovation due to the ability to specialise and to better utilise technology.

Economies of scale and scope are closely related to each other and also to the achievement of coordination and consistency. For example, E-government initiatives can increase coordination across the sector by ensuring computer systems and information technology are compatible across departments. This consistency allows the achievement of economies of scope by enabling the use of an information network, and can also result in economies of scale as the central agency may have greater purchasing power when buying in large quantities.

3.5  Coordination and consistency

Centralisation can also aid in achieving a strong and unified culture with the centre conveying coherence and clarity of direction. This can be beneficial from an efficiency point of view, for example it is efficient to have the Crown Law Office[4] as an authoritative figure providing legal opinions. This may reduce the cost and time involved in determining the legal considerations relevant to a particular issue. Where the costs of decision-making are high, centralised policies can help avoid the costs of duplicating information and decision-making (Kerr, Claridge and Milicich 1998).

In addition, centralisation may provide advantages in ensuring uniform and consistent national standards, minimising regional and social inequalities, and reducing variability Aucoin and Bakvis 1988). In this way centralisation of certain decisions, such as benefit entitlements and eligibility criteria may increase equity.

Table 4 – Economies of scale and scope, coordination and consistency factors
Factors Examples
Decisions where it is possible to achieve economies of scale imply centralised decision-making. New Zealand Debt Management Office’s overnight cash “sweep”
Decisions where it is possible to achieve economies of scope imply centralised decision-making. “Brokerage” roles, budget-setting functions, spreading “best practice”
The need for coherence and coordination across the public sector implies centralised decision-making. Leadership on core public policy values
The need for uniform and consistent national standards and guidelines implies centralised decision-making. Benefit entitlement and eligibility rules, tax administration and rulings

Notes

  • [4]The Crown Law Office provides legal advice and representation to the New Zealand Government on matters affecting the Crown, and in particular, government departments.

3.6  Decentralised information, knowledge and relationships

The existence of information asymmetries mean that in some cases decentralised decision-making may enable more timely and better decisions to be made. That is, decentralisation may enable location-specific information, knowledge and relationships to be used to make the best decision. The quality of decision-making may be improved by moving the point of decision-making closer to those affected by it. Those with more accurate and up to date information can respond to client needs more effectively than those at the centre. In this way decentralisation may improve resource allocation in the public sector and better meet client needs. The presence of this “particular” knowledge is a prime rationale for the decentralisation of more responsibility for input choices to Chief Executives and away from central agencies (together with the need to increase accountability, discussed later in the paper).

Through the use of this location-specific information, decentralisation can foster greater diversity and choice, and encourage innovation and flexibility to respond to change. North (1990) notes that decentralised decision-making enables societies to explore alternative ways of solving problems In this way decentralisation may enhance adaptive efficiency and encourage greater innovation. Decentralisation can offer a greater opportunity to customise services and publicly provided goods to specific circumstances. This leads to the rationale, first expressed by Tiebout (1956) that decentralised decision-making enables the diversification and customisation of publicly provided goods and decisions in accordance with “local” preferences.

3.7  Incentive problems

Where location-specific information is required for decision-making this suggests decentralisation. However, this assumes close alignment of the interests of decision-makers at the rim with the objectives of the government (as mandated by voters). Thus, in considering the location of decision-making rights it is important to consider the incentives of decision-makers and whether those incentives can be aligned with policy objectives through mechanisms such as decision makers’ values, inherent incentives, performance monitoring, the desire to maintain reputation, or contractual arrangements (see also the related section later in the paper on decentralisation as a way to reduce the risk of the abuse of power, which discusses incentive problems associated with centralisation).

Decentralisation increases the risk that perverse motivations may be a problem and may increase the difficulty in aligning the interests of decision-makers with the objectives of the democratically elected government. Decentralisation of decision-making is closely related to the notion of giving managers and departments the freedom to manage those things they can manage well and holding them to account (Treasury 1987). This freedom to manage increases the importance of taking steps to ensure incentives are aligned through the use of performance monitoring, the establishment of a code of conduct and greater management of accountability processes by the centre (Aucoin 1995). In this way decentralisation of certain decision-making powers may also lead to an increase in the role of the centre in monitoring performance.

Incentive problems, together with the desire to manage risk, create a pressure to centralise decision-making. However, this may be costly. Centralising decision-making that requires localised information leads to two possible sources of costs: Either, the costs of transferring information (including reducing the likelihood of a timely decision); or, the costs of not transferring the relevant information (poor decisions).

Table 5 – Information and incentive factors
Factors Examples
Localised non-transferable (or transferable only at a prohibitively high cost) information, knowledge and relationships imply decentralisation. Judgements about supplementary benefit payments
Issues requiring customisation, innovation and a flexibility to respond to localised conditions imply decentralisation of decision-making control. Business development initiatives
Issues where it is difficult to specify the contract or provide suitable incentives for a subordinate to carry out government objectives imply centralised decision-making. Defence

3.8  Diseconomies of scale and scope

Decentralisation can avoid diseconomies of scale and scope. People have limited capacity, limited knowledge and time to acquire and evaluate new information in order to make good decisions (Kasper and Streit 1998). Thus, decentralisation may be used to avoid the concentration of an enormous load of decision-making on a few people and resources at the centre (Gorringe 1996). In some cases, this span of control problem may be able to be solved by shifting decision rights further from the centre, or by shifting decision rights to another agency the same distance from the centre.

3.9  Participation

There may be value in local participation in decision-making, whether or not that in itself leads to better decisions. Local and regional autonomy may be enhanced through decentralised decision-making, thus enabling a greater degree of power sharing, community involvement, and sense of empowerment. For this reason, decentralised decision-making may be more durable even when it is not the most “efficient” option when measured by narrower criteria. Decentralisation can be used to provide a flatter, less hierarchical and more empowering structure. This may lead to greater motivation and job satisfaction due to those further from the centre having a higher degree of freedom, discretion and control over their work (Child 1984).

3.10  Limiting power and providing independence

Decentralisation can also help avoid an unnecessary concentration of power and reduce direct control and dictated direction (at least from the centre). While Parliament (and Ministers) may decide the executive branch of government should undertake a function, they may also decide that it is not appropriate for Ministers or core central bureaucrats to exercise certain decision rights, and may provide a clear commitment to this by providing statutory independence (Tirole 1994). This may involve some decentralising of decision rights from Ministers, at the centre, to other organisations such as Crown Entities or other statutory bodies. In this way decentralisation of decision rights may provide for some separation of powers, thus reducing the risk of abuse of power. This may also be important in signalling independence from Ministers to the public. Providing this credible commitment to independence from political control is not necessarily the same as decentralisation, although it may often be achieved through some decentralisation of decision rights.

Table 6 – Diseconomies of scale and scope, participation and constitutional factors
Factors Examples
Situations where there is a span of control problem may imply decentralisation. Specific functions are decentralised from doctors to nurses (and from nurses to nurse aids) as doctors are unable to make all the decisions relating to the care of their patients
Issues where community involvement, local empowerment and local participation are important imply decentralised decision-making power. Community employment schemes
The need to avoid an unnecessary concentration of power and the risk of abuse of power by distancing a decision from political or central bureaucratic control may involve decentralisation. Specific functions may be decentralised to limit the power of Ministers to other entities such as the Police Complaints Authority and the Commerce Commission (decisions under the Commerce and Fair Trading Acts)

4  The allocation of decision rights

The above discussion leads us to a combination of centralised and decentralised decision rights characterised by a “tight/loose” fit with “tight” or centralised decision-making over issues that have “whole system” implications such as: major government objectives, key public policy values and broad taxation and resourcing decisions. “Loose” or decentralised decision-making may be appropriate, in contrast, over many other decisions in order to enable those with better information, knowledge and relationships with clients to make more appropriate and timely decisions.

Thus, a tight/loose pattern of decision-making is characterised by a clarity of objectives and values that is articulated and enforced by the centre. While, at the same time, those at the coal-face are given the freedom to manage (and the associated accountability) in order to ensure the best decisions. Scott (2001) discusses a similar notion, which he calls “strategic management”. He notes, “strategic management is about promoting direction, purpose and transformation from the management centre, while preserving the devolved management that is essential for quality and innovation in service delivery and also creativity in policy analysis.” (Scott 2001:321)

The presence of location-specific information and the need for customisation suggests decentralised discretion and control over how client needs are met. Centralised control over the overall objectives allows responsiveness to government objectives and ensures alignment with those objectives while decentralised control over how those objectives are achieved prevents the risk of decision-makers becoming detached from the needs of clients and enhances innovation and flexibility in service delivery.

Decentralised discretion and autonomy over the ways in which the government’s key objectives are achieved may help allocate decision-making rights to those with the comparative advantage to undertake them, thus preventing the concentration of an enormous load of decision-making on a few people and resources. For example, some decisions relating to service delivery should be allocated to those at the frontline while politicians undertake significant resource allocation decisions and determine the major objectives of the government. Decentralised decision-making does not eliminate the need for central coordination, indeed, it may strengthen that need in order to address the fragmentation resulting from decentralisation of authority to a larger number of decision-makers (Hart 1998).

Incentive problems must be taken into account when adopting a decentralised allocation of decision-making rights. In order to maintain centralised control over major objectives, increased performance monitoring, a combination of value/cultural alignment, and greater management of accountability processes by the centre may be necessary.

Decentralised discretion over the ways in which government objectives are achieved allows for greater participation and community involvement. Centralised control where national guidelines and standards are required enables “tight” control over government objectives. For example, there may be national standards relating to equal employment opportunities while other decisions regarding employment conditions are decentralised to fit the requirements of specific organisations and employees.

Through a tight/loose pattern of centralised and decentralised decision-making it may be possible to achieve the benefits of one approach without losing many of the benefits of the other. For example when decentralising decision-making rights a “tight/loose” approach suggests it may also be appropriate to use incentive mechanisms, contracts, guidelines, reporting requirements and performance monitoring. And when centralising decision-making rights it may be appropriate to adopt additional mechanisms such as consultation processes, surveys on results and reporting mechanisms. This enables a clarity of objectives and values to be articulated and enforced by the centre while managers with the location-specific knowledge are given the freedom to manage in order to make the best decisions.

4.1  Examples

The following examples show how a tight/loose pattern of decision-making might look in practice. These examples show that the solution often lies in a combination of centralised and decentralised decision rights, typically characterised by tight or centralised control over major objectives and loose or decentralised discretion and autonomy over how those objectives are achieved. This is not intended to be an exhaustive discussion of the issues relevant to each example but rather a useful way of considering the practical application of a tight/loose pattern of decision rights.

4.1.1  Human resources

One of the most important areas of information in this case is that relating to the performance of staff. On a daily basis this is most likely to be observed by “front line” managers. While some information relating to the performance of staff is likely to be transferable, it is important that the section managers have the rights and powers to respond to staffing issues as they arise. Within a department the responsibility for the day-to-day management of staff needs to rest with the person who has the relevant information.

If the Chief Executives of departments are to be held accountable for the actions of their staff then they should have the rights and responsibilities associated with being an employer. For example, they may have discretion over whom they employ and over many of the conditions of employment but not around some of core public sector competencies. Related to the issue of accountability is the need to ensure adequate incentives for Chief Executives. This requires performance monitoring, likely to be undertaken by central agencies.

There is a need for some public sector wide and economy wide standards, for example equal opportunities policies and safe working conditions. The setting and monitoring of such standards, therefore, may be centralised.

Economies of scale may be possible in areas such as advertising vacant positions (central agencies may provide a valuable brokerage role, for instance, in managing a vacancy web-site) and training of staff. However these benefits may be outweighed by the cost of obtaining the relevant information from departments.

Thus the solution is likely to result in a combination of centralised and decentralised functions, with day to day management of human resources and decisions regarding the allocation of resources (such as remuneration) decentralised to departments while central agencies set and monitor public sector wide standards (such as EEO policies).

4.1.2  Benefits payments

In this case there is a need for national guidelines outlining basic entitlements, including who is entitled to a benefit and the amount of the entitlement, in order to ensure national consistency, fiscal control and an equitable access to benefits. The setting and monitoring of such guidelines, therefore, is likely to be centralised.

However, front line staff need to decide whether an individual fits the eligibility criteria. Again this is necessary in order to ensure equitable access to benefits. Front line staff also need to have some decision-making rights in order to ensure timely decisions are made, particularly with regard to supplementary benefit payments and emergency assistance. Thus, while guidelines and standards may be set centrally, it is important that other decision making rights (such as the right to deal with some exceptions, the right to decide eligibility etc) relating to service delivery are decentralised to frontline staff in order to enable a timely and appropriate decision to be made.

Economies of scale may be possible in back office”functions such as organising and facilitating regular payments. Thus, these types of functions may be performed by a central office or by a shared services centre.

Again, the solution is likely to lie in a combination of centralised and decentralised functions, with front line staff having the discretion to make some decisions regarding eligibility while the setting of basic entitlements and eligibility guidelines and standards are set centrally and with some centralisation of back office functions.

4.2  Maori

There are two ways of looking at centralisation and decentralisation questions with regard to Maori. First, given that the Treaty of Waitangi is between Maori and the Crown, Maori need to be able to deal directly with the Crown (ie, Ministers and their direct advisors) on Treaty of Waitangi related issues.

Second, the tribal based nature of the Maori community and a recognition that the Treaty of Waitangi was signed by chiefs on behalf of their iwi, rather than on behalf of a Maori nation, suggest that enhancing the responsiveness of the public sector to Maori clients may require greater decentralisation. A more participative mode of decision-making may be required to empower Maori, enhance local accountability for Maori issues and ensure public services are more responsive to particular communities and cultural groups, including Maori. Decentralised decision-making can serve as a tool to ensure Maori clients are better informed about government services; ensure public agencies are more aware of the social, economic and cultural influences affecting Maori; ensure government services are more relevant to the needs of Maori; ensure Maori clients have a greater sense of ownership; and improve client access to services through the removal of social, physical and psychological impediments (Boston, Martin, Pallot and Walsh 1996).

Both perspectives (the first in favour of centralisation and the second in favour of decentralisation) are relevant, and in some cases, both may be relevant to the same issues.

4.3  Central agencies

The factors and underlying frameworks in this paper are particularly relevant with regard to the role of central agencies. The factors derived from the various perspectives help to answer key questions concerning the role of central agencies:

Which functions are most efficiently and effectively done at the centre?

How directive should central agencies be?

Should any of the functions currently undertaken by central agencies be decentralised?

Are there any additional functions that should be undertaken by central agencies?

5  Conclusion

The combination of factors identified in this paper, in general, leads to the conclusion that decisions relating to the government’s major objectives, including high-level service delivery goals, high-level fiscal goals, decisions regarding key public policy values, trade-offs between different organisations, and broad resource and taxation issues should be allocated close to the centre. In contrast, decisions relating to many of the ways in which those objectives are achieved should be decentralised to those with better information, knowledge and relationships with clients in order to increase participation, customisation and innovation, and flexibility and responsiveness to clients.

Thus, working through the factors outlined in this paper when considering centralising or decentralising decision rights tends to lead to a solution characterised by a “tight/loose” pattern of (de)centralisation. This tight/loose pattern embodies tight or centralised direction and coordination over the major objectives and loose or decentralised discretion and autonomy over many of the ways to achieve them.

This paper has not attempted to rank the common factors in order of importance, as the ranking is likely to differ depending on the particular circumstances. The final solution in any individual case is likely to involve some value judgement between the relative weights of the different factors. However, the paper does suggest that the solution in each case is almost certain to lie in some combination of centralised and decentralised functions and that this combination is likely to be characterised by a tight/loose pattern.

References

Aucoin, P and B Bakvis (1988) The Centralisation-Decentralisation Conundrum: Organisation and Management in the Canadian Government (South Halifax: The Institute for Research on Public Policy).

Aucoin, P (1995) The New Public Management: Canada in Comparative Perspective (Montreal: The Institute for Research on Public Policy).

Boston, J, J Martin, J Pallot, and P Walsh (1996) Public Management: The New Zealand Model (Auckland: Oxford University Press)

Child, J (1984) Organization (London: Harper and Row).

Durie, M (1988) “Social Policy Perspectives” in Martin, J and J Harper, Devolution and Accountability (Wellington: New Zealand Institute of Public Administration)

Eatwell, J, M Milgate and P Newman (1987) The New Palgrave: A Dictionary of Economics (London: Macmillan Press Ltd).

Gorringe, P (1996) “Government and Institutions.” in A Grimes, A Jones, R Procter and G Scobie eds Economics For Policy: Expanding the Boundaries (Wellington: Institute of Policy Studies).

Grimes, A, A Jones, R Procter and G Scobie (2001) Economics for Policy: Expanding the Boundaries (Wellington: Institute of Policy Studies).

Guerin, K (2002) “Subsidiarity: Implications for New Zealand”, Wellington, New Zealand Treasury, Working Paper 02/03.

Gupta, A and I Eerola (1997) “Knowledge creator vs knowledge broker: Corporate roles in technology development in diversified firms” in A Campbell and K Luchs Core Competency Based Strategy (London: International Thomson Business Press).

Kasper, W and M Streit (1998) Institutional Economics (Cheltenham: Edward Elgar)

Kerr S, M Claridge and D Milicich (1998) “Devolution and the New Zealand Resource Management Act. Wellington, New Zealand Treasury, Working Paper 98/7.

Hart, J (1998) “Central Agencies and Departments: Empowerment and Coordination”, in B Peters and D Savoie Taking Stock: Assessing Public Sector Reforms. (Montreal: Canadian Centre for Management Development, McGill-Queens University Press).

Hood C and M Jackson (1991) Administrative Argument. (Aldershot: Dartmouth Publishing).

North, D (1990) Institutions, Institutional Change and Economic Performance (Cambridge: Cambridge University Press).

Peters, TJ and R Waterman (1984) In Search of Excellence: Lessons from America’s Best-Run Companies (New York: Harper and Row).

Pollitt, C, J Birchall and K Putman (1998) Decentralising Public Service Management (Houndmills, Basingstoke, Hampshire: Macmillan Press Ltd).

Review of the Centre Ministerial Advisory Group (2001) “Report of the Advisory Group on the Review of the Centre”, Wellington, State Services Commission. .

Scott, G (2001) Public Management in New Zealand: Lessons and Challenges (Wellington: New Zealand Business Roundtable).

The Treasury (1987) Government Management: Brief to the Incoming Government 1987 (Wellington: New Zealand Treasury).

Tiebout, C (1956) “A pure theory of local expenditures” Journal of Political Economy 64: 416-424.

Tirole, J (1994) “The Internal Organisation of Government” Oxford Economic Papers 46 (4): 1-29.

Appendix – The perspectives

Economic, social, constitutional and managerial perspectives offer a number of useful insights into questions regarding the centralisation or decentralisation of decision-making rights. The issues discussed in this paper are drawn from across the perspectives briefly discussed below.

Economic Perspectives

From an economic viewpoint there are a number of theories that can assist in considering the issue of centralisation versus decentralisation. Significant examples include:

  • Neoclassical economic theory provides a number of tools that are of particular use in considering centralisation/decentralisation questions, including:
    • the use of production functions to determine the efficient level of activity;
    • the ability to achieve economies of scale and scope; and
    • the notion of comparative advantage.
  • A second theoretical framework is that of agency theory. Agency theory rests on the notion that social, political and economic interactions can be viewed as a series of explicit or implicit “contracts”. Agency theory assumes individuals are rational self-interested utility-maximisers. The interests of agents and principals are often likely to conflict. Furthermore, principal-agent relationships are complicated by incomplete information, asymmetric information and the “bounded rationality” of the parties.
  • Transaction cost economics is similar in many ways to agency theory but places greater emphasis on the optimal governance structures for various kinds of transactions, suggesting that rational agents will select governance structures that minimise the sum of production and transaction costs.
  • A final theoretical framework is that of public choice theory. Public Choice Theory seeks to explain how people will act in different institutional settings with different incentive structures. Public choice theory assumes human behaviour is dominated by self-interest, that is, individuals are rational utility-maximisers. For this reason suggestions that politicians are only concerned with societal well-being and their advisors guided by ethical principles are viewed with some scepticism under public choice theory.[5]

As fully developed and applied, for instance in the work of Peter Gorringe[6], these theories contain a much richer picture than these brief descriptions may suggest.

Managerial Perspective

Managerialism describes a body of theory and practice developed by managers, other practitioners and academics. This framework embraces a number of common beliefs. Most notable to the issue of (de)centralisation are the following:

  • the notion of effective management achieved by giving managers the authority, resources and freedom to manage;
  • an emphasis on this devolution of management control coupled with improved reporting, monitoring and accountability mechanisms; and
  • clarity of the role, responsibilities and objective is critically important to the success of a managerial approach[7].

Social Perspectives

Efficiency and economy are not the only criteria in deciding the location of decision-making rights. From a societal perspective the overriding objective is to enhance social well-being. A social viewpoint provides a number of additional considerations; three of which are particularly relevant to decentralisation questions[8]:

  • Participation refers to the level of public involvement in decision-making, resource allocation, and service delivery and the provision of mechanisms for effective local participation. From a societal perspective, social well-being is enhanced through community participation and empowerment. Thus participation has an intrinsic value beyond the achievement of efficiency and effectiveness[9].
  • Partnership refers to the partnership between the state and other providers, with, for example, the state ensuring regulations and standards of provision are met. The notion of partnership involves some joint setting of objectives and ongoing dialogue between government and other groups. This is important from a societal perspective as it represents a way in which participation can be extended while retaining some control over standards of provision.
  • Subsidiarity has at its core the notion of empowerment. The notion of subsidiarity is perhaps more relevant when considering a federal system of government. However, it deserves a mention here as it represents another perspective from which to consider questions of centralisation and decentralisation. Subsidiarity refers to the notion that decisions and actions should be taken at the lowest appropriate level[10].

A social perspective emphasises the benefits of participation and may therefore favour decentralisation even where it is not the most economically “efficient” option.

However a social perspective does not necessarily always favour greater decentralisation, for example, it may be the case that participation is best achieved through central organisation. Furthermore, from a social perspective, centralisation may increase equity through the provision of uniform and consistent national standards and the reduction of regional and social inequalities. Social well-being may, in this way, be enhanced through the reduction in variability and inequalities.

Constitutional Perspectives

A constitutional perspective places the emphasis on the allocation and control of public power, including accountability for the use of that power – that is, the allocation of some fundamental decision rights. There are a number of important features of our system of government that are particularly relevant to the allocation and control of public power, notably democracy, the separation of powers, the rule of law:

  • Democracy: As a democracy, key judgements about public policy values, trade-offs, resources and taxation should usually be made by representatives elected by and accountable to the people. The agencies of executive government should be accountable to Ministers and Parliament, as the elected representatives of the people;
  • Separation of Powers: At the same time, our system of government allocates important powers, and checks on the use of power to different branches of government – the judiciary, the executive and the legislature;
  • Rule of Law: The executive, and other branches of government are bound to obey the law. This is important in terms of shaping and constraining the allocation of decision rights, including the degree of centralisation for two reasons. The first is that Parliament is usually careful about the allocation and regulation of statutory powers, usually allocating them to a particular office holder or category of office holders, and providing some scope to delegate those powers to other persons. For example, the State Sector Act provides quite broad powers for chief executives to delegate powers to members of their departments (and for Ministers to delegate statutory powers to their chief executives). Any decentralisation has to be consistent with this allocation of decision rights. The second is that there are some well established principles governing how Parliament allocates certain types of powers, such as:
    • What should be covered by primary legislation, and what is appropriate to secondary legislation;
    • Who should have the power to make secondary legislation;
    • How coercive powers should be exercised.

Parliament (and Ministers) may also decide that while a function should be undertaken by the executive branch of government, with a responsible Minister, that it is not appropriate for Ministers to exercise certain decision rights, and may provide a clear commitment to this by providing statutory independence. This can be in either a departmental context (eg, the Commissioner of Inland Revenue in respect of individual taxpayers) or an arms length Crown entity context (eg, Commerce Commission). Providing this credible commitment to independence from political control is not necessarily the same as decentralisation, although may often be achieved through some decentralisation of decision rights.

New Zealand is a unitary and relatively centralised state – we do not have a federal system of government, and the role of local government is relatively limited. Parliament does, however, allocate some responsibilities in a decentralised way, notably to local government, but also to organisations like District Health Boards. This will usually reflect a desire to provide for local democracy or a regional basis for the organisation of some decision rights.

Notes

  • [5]For a more detailed description on these theories see Eatwell, Milgate and Newman, 1987
  • [6]Gorringe, 2001
  • [7]For more information on what is often referred to as ‘New Public Management’ see the doctrines outlined by Hood and Jackson (1991)
  • [8]Durie, 1988
  • [9]Boston, Martin, Pallot and Walsh, 1996
  • [10]Guerin, 2002