Agents or Stewards?

Contracting with
Voluntary Organisations
Jo Cribb

Introduction system is based. It suggests that better outcomes may

be achieved by a more discriminating application of
In order to escape a violent partner, take a disabled child
agency theory in the design of contractual relationships
swimming, get information about government services
between government agencies and community
or extinguish a rural house fire, New Zealanders will
organisations, and by the application of an alternative
generally interact with a voluntary organisation. In many
model - one founded on stewardship theory.
areas of government voluntary organisations play an
important delivery and operational role. A modest The agency basis of current contracting
estimate (based on data from only 14 government
practice
departments) suggests that at least $650 million of
public social services are delivered each year by voluntary Voluntary organisations have existed in New Zealand
organisations via government contracts (Community since the beginning of colonisation. Scholars conclude
and Voluntary Sector Working Party, 2001). that they have been dependent on state funding since
the earliest years of organised government. Prior to the
Agency theory was a foundational component in the
1990s, the most common method of state funding for
design of New Zealand’s state sector reforms in the 1980s
voluntary organisations was a system of grants and
and 90s. It has been strongly influential in shaping the
subsidies not usually attached to specific services. The
contractually-based working relationship between
early 1990s, a time of rising fiscal debt, saw a rethink of
government agencies as providers of funding for services what was the appropriate size and function of the state.
and the community organisations that deliver those New Zealand’s public management system, drawing on
services on their behalf. Contracting between economic and administrative theories, was restructured
government agencies and voluntary organisations can, (Boston et al., 1996). Several aspects of the restructuring
however, be fraught (Office of the Auditor General, altered how government agencies approached their
2003). Officials often find specifying what they wish to relationship with voluntary organisations. In particular,
be delivered in the prescribed manner for contracts a based on insights from agency theory, the assumptions
challenge. Contract outputs risk becoming checklists government agencies made about voluntary
for delivery and, as such, are not linked to the organisations changed.
effectiveness of the service or outcome desired and create
For agency theorists, social and political life can be
opportunities for ‘creaming’ the ‘easiest’ clients and tasks.
understood as a series of contracts. Principals delegate
From a voluntary sector perspective, small organisations
tasks, using contracts, to agents. Agents undertake work
often find the compliance costs of reporting onerous.
on principals’ behalf in return for rewards. The focus of
As a result, newly-established voluntary organisations
the theory is the contract governing the relationship
that provide services addressing emerging needs (such
between principal and agent and determining how the
as those of refugee communities) can be excluded from
Volume 2, Number 2 2006

contract can be made as efficient as possible (Eisenhardt,

funding and contracting processes.
1989). Agency theorists interpret broadly what
This article, which derives from the author’s doctoral constitutes a contract. Contracts may be ‘classical’: arm’s-
research at Victoria University of Wellington’s School length, formal and explicit. They may also be ‘relational’:
of Government (Cribb, 2005), raises questions about implicit, open-ended, incomplete and based on
the assumptions on which the current contracting obligations (Boston et al., 1996).

11
 Two problems can arise from contractual relationships: reduce information asymmetries and ensure they did not
goal conflict, as the goals of the agents and their principals shirk. The development of contracting policies and
are usually different (Eisenhardt, 1989); and information practices was based on such assumptions (New Zealand
asymmetry, as the agent often has more information than Treasury, 1995).
the principal about the task at hand. Under such
Grant-based funding ceased for many voluntary
conditions, it is difficult and expensive for the principal
organisations in the early 1990s and was replaced with
to verify what the agent is doing. The focus of agency
contracts for the delivery of services. Government
theorists is on minimising the impact of these problems
agencies attempted to specify what services they
for the principal.
required, introduce monitoring regimes, and sanction
Agency theory, drawing on its neo-classical economic poorly performing organisations by not renewing their
roots, assumes that individuals are ‘rational, self- contracts. Clear objectives in contracts were designed
interested utility maximisers’. As such, the interests of to ensure voluntary organisation providers focused on
the principal and agent are bound to conflict because results. Reporting on the objectives would provide good
each party is assumed to be trying to maximise its information to government agencies about the provider
personal benefits (Eisenhardt, 1989). Even if principals and the quality of the service (Boston et al., 1996).
and agents have similar goals, assumptions about the
nature of individuals leads to the conclusion that agents Overlaying the introduction of contracting was the
will shirk by producing outputs at a higher cost than pressure to reduce government spending. Competition
required or producing outputs of a lower quality than between providers was encouraged. Voluntary
specified. This is termed moral hazard: a lack of effort organisations were often required to tender. Market
on the part of the agent. Agents may also claim that pressures were sought to ensure efficiency of service
they have skills and capacity to deliver that they actually delivery. A drive to increase transparency and
do not. Adverse selection may occur as principals contract accountability was also an important aspect of the
with agents who have misrepresented their abilities restructuring. Accountability was interpreted as
(Eisenhardt, 1989). answerability and took the form of formal reporting
against specified measures (Mulgan, 2003). Enhanced
Principals have a number of options available to them
accountability was seen as a way of developing more
to minimise the risk of their agents shirking, cheating
efficient and effective organisations and service delivery.
or operating with guile. They can provide incentives
for the agent to operate according to the principal’s By the mid-1990s, voluntary organisations with a
wishes, monitor the agent’s actions to ensure they are funding relationship with government had become
doing so, and sanction the agent if their performance is service providers. Many had a formal, ‘classical’ contract
not satisfactory. with government. What they were required to provide,
in what quantity, at what quality and price - all this was
Agency theory was influential in the restructuring of a
specified in the contract.
number of relationships within the public service.
Purchase agreements between ministers and chief
Voluntary organisations’ experience of
executives for the purchase of services from government
departments were introduced, as were performance
contracting
agreements between chief executives and the State Services The great majority of voluntary sector managers and
Commissioner. The contract model was also applied to board members surveyed in my doctoral research, which
the relationship between government agencies and focused on four community organisations of differing
voluntary organisations. The Treasury was a powerhouse types, focused their attention on their relationships with
Volume 2, Number 2 2006

of the state sector restructuring. Briefings written towards their clients. They felt they were most accountable to
the end of the restructuring period showed that Treasury their clients for the quality of the services they provided.
officials were concerned about the quality of services They generally identified their relationship with
delivered by voluntary organisations. Government, the government as a hindrance. Contracting and funding
papers concluded, needed to be vigilant about regulating agreements with government agencies were seen to be
and monitoring voluntary organisations’ delivery to driving down standards of care. Poorly designed

12
programmes, irrelevant performance measures and a Are the assumptions of agency theory
piecemeal approach to service provision were seen to valid?
detract from the quality of service they could provide
to their clients. At its core, agency theory predicts that agents will have
different aims from their government funders (the
Respondents’ perceptions of how they were viewed and principals), that they will operate opportunistically and
treated by government officials were uniform. They felt with guile to achieve their own aims, and that they will
that officials treated them as ‘poor cousins’. They shirk as much as possible. In this research it was found,
perceived that their skills were not recognised, their ideas however, that respondents had similar high-level goals
and suggestions not valued and their concerns not to officials. Officials and respondents both aimed to
considered valid. And they perceived that they were
achieve positive outcomes for recipients of their services.
being treated as agents, while, on the other hand, officials
Respondents saw that their primary accountability
were operating as if they were principals
obligation was to their clients, and they perceived
More specifically, respondents perceived that: themselves to be going the ‘extra mile’ at their own
* They were viewed as inferior, while officials saw expense to fulfil it. They saw themselves working for
themselves as experts. There was no negotiation or clients in the face of barriers created by the contracting
dialogue about the contents of the contract or process. Rather than shirking by attempting to provide
funding agreement. (Agency theory assumes that the less service than required, as agency theory predicted
principal has a superior position in regard to the they would do, respondents were actively fund-raising
agent and will work to maintain the power to increase levels of service provision, independently of
imbalance. Indeed, principals enact mechanisms to the funding provided by government officials.
control agents – such as financial incentives and International studies have drawn similar conclusions.
monitoring regimes – and ensure power For example, Rasmussen et al.’s 2003 study found that
asymmetries.) government officials and voluntary sector managers
* Officials did not trust them. They expected the shared a client focus. Indeed, agency theorists have
respondents to try to defraud the system or produce similarly come to accept that the assumption of goal
poor quality work. The integrity and expertise of conflict may not always be valid. Eisenhardt (1989)
respondents was not recognised. (One of the key tenets concedes that goal alignment may occur in highly
socialised or clan-orientated firms, or in situations where
of agency theory is the assumption that agents have
self-interest gives way to selfless behaviour. As goal
different interests from those of the principal and will
conflict decreases, so does the need for in-depth
seek to better themselves at the principal’s expense.)
monitoring. Agents will behave in a manner acceptable
* The main form of communication between to the principal regardless of the level of monitoring.
respondents and officials was monitoring reports.
The relationship was distant and paper-based. Agents or stewards?
(Agency theory promotes external monitoring as a
It may be timely to consider an alternative approach.
mechanism to control age
Stewardship theory takes as its starting point the
* Reporting focused on specific outputs that were assumption of goal alignment, and is increasingly seen
often irrelevant to service quality or organisational as an important framework for structuring relationships
performance. (Detailed pre-specification of (Block, 1996).
contracts is another tool agency theorists promote
The origins of the concept of stewardship are biblical.
to control agents.)
Volume 2, Number 2 2006

Stewards, as valued employees who are entrusted with

* Contracting was an economic tool used to try to running households, are mentioned in both the Old
maximise efficiency. Respondents perceived that and New Testaments. Stewards were seen as servants of
officials took a ‘take it or leave it’ approach focused someone or something greater than themselves, were
on competition between providers to drive down the committed to their work, and had the discretion to take
price of service delivery. risks on behalf of their masters.

13
 Proponents of stewardship theory argue that pro- When it had been identified that goals were shared, the
organisational and collective behaviours are of higher contracting process could begin. There would be a large
utility than the individualistic, self-serving behaviours investment of time at the beginning of any relationship.
assumed by agency theory (Dicke and Ott, 2002). Put
If, after discussions with a voluntary organisation, it were
simply, if the organisation does well, its members will
clear that the organisation had different goals from those
do well, so they invest their energy in their organisation’s
the officials were trying to achieve (or that it was not
success (Davis et al., 1997). People will thus put the
capable of delivering what was required), officials would
organisation’s needs first. They are not assumed to want
seek another provider. In practice, this might mean that
to shirk or act opportunistically, as agency theory has
organisations currently contracting with government
it. Instead, they are motivated to work to achieve the
would not be offered future contracts: a potentially
organisation’s goals (Block, 1996).
difficult and politically volatile situation.
Stewards, like others, have basic survival needs, such as
Once a relationship had been established, voluntary
a regular salary. The differences between a self-interested
organisations would be trusted to get on with the job.
agent and a steward are in how these survival needs are Performance measures focusing on measuring outcomes
met. Stewards realise that there is a trade-off between for clients would be jointly developed over time. Blanket
personal and organisational needs, and choose to work controls and ‘boilerplate’ contracts would not be used.
for organisational needs. By doing so, stewards assume Non-financial motivations would be acknowledged.
their personal needs will be met (Davis et al., 1997). Knowledge generated from the performance measures
Some theorists acknowledge that stewards may work would be used to modify service delivery. Dialogue
for altruistic reasons – unselfish concern and devotion between parties would focus on delivery problems, and
to others – without expected return. Dicke and Ott potential improvements or innovations. The
(2000) argue that altruistic motives best explain information asymmetries that are problematic in
voluntary organisations and their employees, as principal-agent relationships would still exist: the staff
employees are selected and socialised to care about and of voluntary organisations have more information about
serve clients. service delivery than officials. Under a stewardship
Less controlling organisation structures and mechanisms framework, such asymmetries would not be seen as a
are needed for stewards. Extending the autonomy of potential source of risk for principals (such as an avenue
stewards maximises the benefits of their behaviour. for inflating contract prices), but rather such knowledge
Informal and intrinsic accountability mechanisms, would be seen as expertise that should be incorporated
rather than ‘hard’ legalistic or mechanistic ones, are best into policy processes.
suited to stewards: the promulgation of professional The additional time made available to officials by the
standards, peer review, and mechanisms that build a reduction of detailed monitoring of contracts could be
sense of internal responsibility (Dicke and Ott, 2000). used by them to provide long-term strategic guidance,
Control can be counter-productive because it signals and research on effective service delivery and the
that the steward is not trusted with a level of discretion. attainment of outcomes. The stewards (voluntary
Such control will lower the motivation of a steward to organisations) would be freed from detailed reporting
work for the organisation (Davis et al., 1997). and the provision of government-designed piecemeal
If a stewardship approach were to be adopted to guide programmes to do what they do best – that is, deliver
the government-voluntary sector contracting in-depth services to clients. The principals (officials)
could support this through research, development and
relationship, a number of changes would occur.
investment in strategy.
Stewardship theorists would see current accountability
Volume 2, Number 2 2006

mechanisms, such as monitoring, audit and reporting, Officials could choose to contract better with fewer
as superficial. They would instead focus on ensuring organisations, given the resource-intensive nature of
that goals are shared. This would mean government establishing relationships. This would not be without
agencies would take the time to understand what the risk for the voluntary sector. Fewer organisations might
voluntary organisations were trying to achieve, how they receive funding. A trusted ‘inner circle’ could develop.
were doing it and where there was a congruence of goals. Under such circumstances, new organisations might be

14
excluded. Organisations receiving funding could also Moreover, government agencies need to be prepared to
find it difficult to maintain their unique qualities as share decision making. For the Community and
they developed a close relationship with government, Voluntary Sector Working Party, this meant providing
as predicted by neo-institutionalists (DiMaggio and genuine opportunities for organisations to comment on
Powell, 1983). policy and involvement in the early stages of policy
Of benefit to the voluntary sector, however, would be development. Respondents’ perceptions of the
the increased investment government agencies could contracting process as a ‘take it or leave it’ one suggest
make in building the capacity of voluntary organisations. that this is still a long way off.
In particular, officials might seek to improve the ability Time is also needed to develop shared expectations and
of voluntary organisations to monitor how well they sensible performance measures. Experimentation will
are doing in fulfilling their mission, so that they could be needed. However, as demonstrated by the example
self-regulate. Evidence indicates that these organisations of a contract in the recent Auditor General’s report on
have little capacity to monitor closely their own the government contracting processes, practices
performance.
currently do not result in shared expectations:
A number of conditions would need to be met if
There was no ‘meeting of minds’ on the part of the
stewardship theory were to effectively govern the
Ministry and [the voluntary organisation it
government-voluntary sector relationship. Voluntary
contracted with] as to the length and expected
organisations would need sophisticated systems of
outcomes of the contract. The parties were confused
performance management and self-criticism to ensure
over the length of the contract and the expected
that they were working towards the shared goals. This
outcomes during the term of the contract. (Office
author’s research and other studies have found that
of the Auditor General, 2003, p.59)
internal performance measurement is a weakness, and
that the level of self-regulation needed to ensure that
The way forward?
such organisations are working to enhance the public
good is not present. In theory, adopting a stewardship rather than an agency-
theory frame may address many of the limitations of
Both parties need to be interested in each other and looking
the current contracting regime. However, the reality is
for better ways to serve the public (that is, to have an
much more complex.
effective focus on outcomes). The author found that most
respondents believed that officials had little interest in them Stewardship theory is untested, particularly in voluntary
and that the contracting process was mainly about sector research, and the enthusiasm with which some
delivering outputs for clients rather than outcomes. of its proponents advocate this approach makes their
claims seem too good to be true. As Arthurs and Busentiz
Government needs to recognise that it puts up cultural
(2003, p.155) argue, ‘stewardship theory paints an
barriers to working with voluntary organisations. The
excessively rosy picture of the steward’.
report of the Community and Voluntary Sector Working
Party (2001) provided specific examples of such barriers Government agencies need to make voluntary
operating in the New Zealand context. These included organisations accountable for taxpayers’ money.
the speed at which policy makers expect the voluntary They need to be seen to be in control (Dicke, 2002).
sector to respond to draft documents. A number of As guardians of the public purse, public agencies
organisations reported that they were unable to take what can be considered a risk-averse approach.
participate in policy processes in a meaningful way. The mechanisms of external control, such as
Umbrella organisations, which seek feedback from their monitoring and reporting, can be seen to provide
Volume 2, Number 2 2006

constituencies before engaging with government the needed assurance to both ministers and the
agencies, saw themselves particularly hampered by the public that taxpayers’ money is being used effectively
short timeframes allowed. The speed and demands of (Davis et al., 1997). Being accountable for taxpayers’
the political cycle provide one example of a cultural dollars is one example of the different set of pressures
barrier for voluntary organisations that prioritise that officials as opposed to voluntary organisations
discussion and consensus. operate under.

15
 Another pressure originates from the three-year In sum, stewardship theory is based on assumptions that
political cycle and the associated changes in policy may portray more accurately than does agency theory
and programmes. Such change will make it difficult the government-voluntary sector contracting experience.
for officials to establish committed long-term But it is certainly not to be regarded as any sort of
relationships with voluntary organisations. A regime panacea, and if it were to be applied as slavishly as agency
built on the assumption that providers should be theory was in the late 1980s and early 90s, it too would
allowed a high level of discretion is unlikely to be give rise to all sorts of unintended, and often undesirable,
feasible in this context. Public accountability systems consequences. The least that can be said is that, given
can be seen as a trade-off between discretion and the importance of the contracting relationship for the
assumed innovation and efficiency, on one hand, and delivery of public services, there needs to be much more
the need for control on the other. Increased control exploration of the validity of the assumptions upon
comes at an increased cost: the resources needed to which government agencies and voluntary organisations
monitor behaviour, as well as reduced levels of enter into contractual relationships.
innovation and the performance improvements that
can result from such innovation. References
At the heart of the debate is the issue of how much Arthurs, J. and L. Busentiz (2003) ‘The boundaries and
discretion agencies can allow their providers: how far limitations of agency theory and stewardship theory in the
can they trust them? Finding the balance between control venture capitalist/entrepreneur relationship’,
and trust is the key question for any alliance, and there Entrepreneurship: Theory and Practice, Winter, pp.145-62.
are no easy and obvious answers. Both come at a cost to Block, P. (1996) Stewardship: choosing service over self-
the organisations involved. For a relationship based on interest, San Francisco: Berrett-Koehler Publishers Inc.
trust to develop, government agencies and voluntary
Boston, J., J. Martin, J. Pallot and P. Walsh (1996) Public
organisations will need to become familiar with each
Management: the New Zealand model, Auckland: Oxford
other. Trusting relationships are only formed between University Press.
actors who have established close bonds. The interests
of the parties must be aligned. As previous discussions Community and Voluntary Sector Working Party
of agency theory have shown, when each party is self- (2001) Communities and Government: potential for
interested and working towards different goals, control partnership: whakatopu whakaaro, Wellington: Ministry
becomes a central feature of the relationship, at the of Social Policy.
expense of, or as a substitute for, trust (de Leon, 2003). Cribb, J (2005) ‘The Accountability of Voluntary
For trust to develop in a relationship a sense of shared Organisations: implications for government funders’,
higher-level purpose is needed. Trust is also most likely PhD thesis, School of Government, Victoria University
to eventuate when parties have taken care and time in of Wellington.
choosing their partners. Davis, J., D. Schoorman and L. Donaldson (1997)
However, in trusting voluntary organisations, ‘Toward a stewardship theory of management’, Academy
government agencies will be implicitly accepting the risks of Management Review, 22, pp.20-47.
associated with being dependent on them for the delivery de Leon, L. (2003) ‘On acting responsibly in a disorderly
of services. Trust implicitly involves loosening world: individual ethics and administrative
mechanisms of control, and relying on voluntary responsibility’, in B.G. Peters and J. Pierre (eds),
organisations to deliver services of an acceptable quality. Handbook of Public Administration, London: Sage.
Such interdependence increases the vulnerability of
Dicke, L. (2002) ‘Ensuring accountability in human
Volume 2, Number 2 2006

government agencies (particularly to poor performance

by voluntary organisations) and increases the potential services contracting: can stewardship theory fit the bill?’,
for betrayal or harm from voluntary organisations (such American Review of Public Administration, 4.
as the potential for being defrauded). Understandably, Dicke, L. and J.S. Ott (2002) ‘A test: can stewardship
such risk is not readily acceptable to either public sector theory serve as a second conceptual foundation for
managers or politicians. accountability methods in contracted human services?’,

16
International Journal of Public Administration, 25,
pp.463-87.
DiMaggio, P. and W. Powell (1983) ‘The iron cage
revisited: institutional isomorphism and collective
rationality in organisation fields’, American Sociological
Review, 48, pp.147-60.
Eisenhardt, K. (1989) ‘Agency theory: an assessment and
review’, Academy of Management Review, 14, pp.57-74
Mulgan, R. (2003) Holding Power to Account:
accountability in modern democracies, New York: Palgrave
MacMillian.
New Zealand Treasury (1995) ‘The Voluntary Sector
in New Zealand’, unpublished papers released under
the Official Information Act, Wellington.
Office of the Auditor General (2003) Inquiry into Public
Funding of Organisations Associated with Donna Awatere
Huata MP, Wellington.
Ott, J.S. and L. Dicke (2000) ‘Important but largely
unanswered questions about accountability in
contracted human service’, International Journal of
Organisation Theory and Behaviour, 3, pp.283-317.
Rasmussen, K, D. Mallroy and J. Agarwal (2003) ‘The
ethical climate of government and non-profit
organisations’, Public Management Review, 5, pp.83-97.

Jo Cribb recently completed her PhD at

the School of Government, Victoria
University of Wellington, and has
worked in various capacities in the New
Zealand Public Service. She can be
contacted at [email protected]
Volume 2, Number 2 2006

17