The paradox of greater NGO accountability A case study of Amnesty Ireland

Description
Despite mounting public, governmental and corporate interest in issues of non-governmental organisation (NGO)
accountability, there are few academic studies investigating the emergence of accountability mechanisms in specific
advocacy NGO settings. Drawing on the theoretical constructs of hierarchical and holistic accountability, this paper
addresses this research gap by investigating recent developments in accountability practices at the Irish section of
the human rights advocacy NGO Amnesty International

The paradox of greater NGO accountability: A case study
of Amnesty Ireland
Brendan O’Dwyer
a,
*
, Je?rey Unerman
b,1
a
University of Amsterdam Business School, University of Amsterdam, Roetersstraat 11, 1018 WB Amsterdam, The Netherlands
b
School of Management, Royal Holloway, University of London, Egham, Surrey TW20 0EX, United Kingdom
Abstract
Despite mounting public, governmental and corporate interest in issues of non-governmental organisation (NGO)
accountability, there are few academic studies investigating the emergence of accountability mechanisms in speci?c
advocacy NGO settings. Drawing on the theoretical constructs of hierarchical and holistic accountability, this paper
addresses this research gap by investigating recent developments in accountability practices at the Irish section of
the human rights advocacy NGO Amnesty International. Through analysis of a series of in-depth interviews with man-
agers in Amnesty Ireland, supported by extensive documentary scrutiny, this study examines reasons why Amnesty’s
historical reliance on internal forms of accountability has been augmented with a range of ad hoc external accountabil-
ity mechanisms. The study reveals that while managers favoured the development of holistic accountability mechanisms
exhibiting accountability to a wide range of stakeholders, a hierarchical conception of accountability privileging a nar-
row range of (potentially) powerful stakeholders, has begun to dominate external accountability discourse and practice.
It was widely perceived that this trend could, somewhat paradoxically, prove counterproductive to the achievement of
Amnesty’s mission. Resolving this paradox will, at a minimum, involve Amnesty Ireland’s leadership becoming more
open to and more knowledgeable about a broader, more holistic accountability conception. The paper considers the
possible implications of these ?ndings for the development of NGO holistic accountability practice more generally.
Ó 2008 Elsevier Ltd. All rights reserved.
Introduction
Despite the growing signi?cance of non-govern-
mental organisations (NGOs) within the expanding
civil society sector (Ebrahim, 2005) and increasing
calls for greater NGO accountability (Ebrahim,
2005; Gray, Bebbington, & Collison, 2006; Jordan
& Van Tuijl, 2006; Unerman & O’Dwyer, 2006a;
Unerman & O’Dwyer, 2006b), there is little aca-
demic research investigating issues of accountability
within speci?c NGO settings (Hopwood, 2005;
Unerman &O’Dwyer, 2006a). While a fewinsightful
studies into aspects of accountability within welfare-
provision and development NGOs have recently
0361-3682/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2008.02.002
*
Corresponding author. Tel.: +31 20 525 4260.
E-mail addresses: [email protected] (B. O’Dwyer),
[email protected] (J. Unerman).
1
Tel.: +44 1784 443478.
www.elsevier.com/locate/aos
Available online at www.sciencedirect.com
Accounting, Organizations and Society 33 (2008) 801–824
been undertaken in the social accountability litera-
ture (see, for example: Dixon, Ritchie, & Siwale,
2006; Goddard & Assad, 2006), there is a lack of
such studies examining accountability mechanisms
and practices within advocacy NGOs. Academic
research into NGO accountability has also lagged
behind rapidly developing public, governmental
and business interest in this area (see, for example,
ActionAid International, 2006; Blagescu, De Las
Casas, & Lloyd, 2005; Blagescu, Lloyd, Dombrow-
ski, Kadosh, & Oatham, 2006; Edwards & Fowler,
2002; International Council on Human Rights Pol-
icy (ICHRP), 2003; United Nations, 2006).
Two key related issues, however, have emerged
within the embryonic NGO accountability litera-
ture. Firstly, some studies have recognised the
potential for inappropriate accountability mecha-
nisms to damage, rather than enhance the social
and environmental bene?ts that many NGOs seek
to realise through their development and/or advo-
cacy activities (Dixon et al., 2006; Ebrahim, 2005;
Ebrahim, 2003a; Ebrahim, 2003b; Najam, 1996;
Unerman &O’Dwyer, 2006b). Secondly, the emerg-
ing dominance of upward hierarchical accountabil-
ity to donors at the possible expense of more holistic
accountability to a broader range of stakeholders,
especially bene?ciaries, has created concerns that
NGOs’ accountability priorities are being distorted
(Blagescu et al., 2005; Ebrahim, 2005; Ebrahim,
2003a, b; O’Dwyer and Unerman, 2007).
Focusing on these two emergent and signi?cant
issues, the aim of this paper is to enhance and dee-
pen our understanding of the role and in?uence of
accountability (in its various forms) in speci?c
NGO contexts.
2
Speci?cally, it aims to contribute
towards the advancement of academic insights
into NGO accountability practice by analysing
factors driving, and problems experienced during
the emergence of, external accountability as a
key managerial concern within the Irish section
of the human rights advocacy NGO, Amnesty
International. Using a conceptual framework
drawing on the concepts of hierarchical and holis-
tic accountability,
3
this study investigates, primar-
ily through an analysis of a series of in-depth
interviews with Amnesty Ireland managers, the
pressures these managers perceive for greater
accountability to certain stakeholders, the broad
nature of the forms of accountability they are
embracing to address these perceived accountabil-
ity demands, and the possible impact that imple-
mentation of these forms of accountability may
have on the realisation of Amnesty’s human rights
mission in practice.
Amnesty Ireland has been chosen as the focus
for this study because it forms part of a wider
organisation, Amnesty International, that is cur-
rently assessing the nature of its own accountabil-
ity more explicitly than ever before. Amnesty
International has recently signed up to a global
International NGO (INGO) accountability charter
with ten other INGOs in order to demonstrate its
commitment to transparency and accountability
(International Non-Governmental Organisations
(INGO), 2006). This has evolved in response to
increased public scrutiny INGOs are experiencing
given widespread concerns about their increasing
power and in?uence (ICHRP, 2003; Unerman &
O’Dwyer, 2006a, b). Amnesty International’s com-
mitment to accountability is also re?ected and
2
There is little consensus on how to de?ne and classify non-
governmental organisations (NGOs) (Vakil, 1997). NGOs
cover a huge range of institutions encompassing multinational
corporations, voluntary associations, credit unions, farmer’s co-
operatives, consumer groups, religious organisations, and trade
unions (Hudson, 1999). Vakil (1997) suggests that NGOs can be
distinguished by their essential organisational attributes. These
comprise their orientation – the types of activities they engage in
– and their level of operation – at international, national or local
community level. He identi?es six orientation categories:
welfare, development, development education, networking,
research and advocacy. In this paper, we focus on a national
section of an international human rights advocacy-oriented
NGO (Amnesty International) which has signi?cant potential
to in?uence public policy making.
3
Within the embryonic NGO accountability literature, hier-
archical accountability is seen as embracing narrow, short-term,
accountability to powerful NGO patrons, favouring the use of
quantitative measures designed to assess speci?c aspects of
NGO ‘performance’ (Edwards & Hulme, 2002a; Najam, 1996).
Conversely, holistic accountability is seen as broadening the
accountability of an NGO to encompass accountability for its
wider impacts, including its actual and potential impacts on a
range of less powerful stakeholder groups. This can, for
example, include accountability for the achievement of an
NGO’s mission in bringing about long term structural change
(Edwards & Hulme, 2002a).
802 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
reinforced in its latest strategic plan, where
‘accountability’ is identi?ed as one of the three
fundamental concepts that will guide the scope of
Amnesty’s future work. This is articulated in terms
of holding both human rights abuse perpetrators
and Amnesty ‘themselves’ to account.
This study reveals that a broad recognition
within Amnesty Ireland of the need for greater
external accountability to enhance its mission
delivery has, in practice and paradoxically,
resulted in the promotion of speci?c hierarchical
accountability mechanisms and practices that are
seen by many in the organisation as threatening
the e?ectiveness of its mission attainment.
To address its objectives, the remainder of the
paper is organised as follows. We ?rstly expand
on the concepts of hierarchical and holistic
accountability, and speci?cally relate these to
the NGO setting we are examining. After provid-
ing some background contextual material on
Amnesty Ireland, and then discussing the
research methods employed to collect and ana-
lyse the evidence gathered, the analytical part
of the paper commences by exploring recent
trends in Amnesty Ireland’s accountability, which
have seen a historical focus on internal forms of
accountability for the veracity of its public
pronouncements augmented with ad hoc external
forms of accountability motivated by a desire to
demonstrate to a range of stakeholders how well
Amnesty ful?ls its human rights mission. The
analysis continues by exploring why, despite a
desire to be holistically accountable to a broad
range of stakeholders, the external forms of
accountability at Amnesty Ireland have come to
be dominated by concerns to adopt hierarchical
accountability mechanisms focused on demon-
strating e?ectiveness in a narrow range of
performance measures. The remaining section of
the narrative then examines the perceived
potential for this hierarchical accountability
trend to prove counterproductive in practice to
the short, medium and/or long term realisation
of Amnesty’s mission. The ?nal section summa-
rises and concludes on the ?ndings and explains
their possible implications for the development
of holistic NGO accountability practices more
generally.
Exploring NGO accountability – hierarchical and
holistic accountability
Theoretical views on the stakeholders to whom
an organisation is considered responsible and
hence accountable range from a belief that organ-
isations are responsible and thus accountable only
to those stakeholders who can directly in?uence
the achievement of the organisation’s objectives
(Unerman & O’Dwyer, 2006b), through to a belief
that organisations have responsibilities and
accountabilities to all those whose life experiences
may be a?ected by the organisation’s activities
(Cooper & Owen, 2007). This section brie?y
explores these accountability conceptions, and
their related practices, as they apply to advocacy
NGOs such as Amnesty International. In particu-
lar, it draws on the embryonic NGO accountabil-
ity literature to develop and explain the
theoretical concepts of hierarchical accountability
as a form of accountability to a narrow range of
in?uential stakeholders, and holistic accountability
as a form of accountability to a much broader
range of stakeholders.
Hierarchical accountability in NGOs
Hierarchical accountability is narrowly func-
tional, short-term in orientation and favours
accountability to those stakeholders who control
access to key resources for both resource use
and immediate (campaign) impacts (Ebrahim,
2003a, b; Edwards & Hulme, 2002a, b). Some-
times referred to in the NGO accountability liter-
ature as ‘functional accountability’ (because of its
focus on accountability for selected narrowly
de?ned functions of NGOs), it conceives of
accountability as a form of external oversight
and control and encourages ‘‘rationalizations of
action – the constant giving and demanding of
reasons for [and results of] conduct” (Roberts,
2001, p. 1549). Roberts (1991) contends that this
can serve to instil a sense of anxiety or vulnerabil-
ity among organisational managers as they must
continually strive to demonstrate ‘performance’
(see also: Ebrahim, 2003a, b; Najam, 1996). Rob-
erts (1991, p. 366) further claims that the result-
ing ‘‘anxious preoccupation with how [they] are
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 803
seen [is] wholly antithetical to the creation of self
knowledge and the embrace of failure as an
opportunity for learning” which are central to
the ethos of many advocacy NGOs. Hence, it is
often claimed that hierarchical forms of account-
ability, through their denial of the scope for
learning and sharing, can have damaging e?ects
on the ability of NGOs to act as e?ective cata-
lysts for social change (Dixon et al., 2006; Ebra-
him, 2005, 2003b).
Among human rights NGOs, a strategic focus
on forms of hierarchical accountability has led to
a temptation to measure impacts mechanically, in
order to provide an impression of precision (ICH-
RP, 2003). Here NGO ‘performance’ tends to be
conceptualised around resource use, measuring
immediate impacts using short-term quantitative
targets, and a standardisation of indicators focus-
ing attention exclusively on individual projects,
campaigns or organisations (Edwards & Hulme,
2002a). Some contend that use of this concept
of ‘performance’ has restrained the imagination,
intuition and ?exibility which have been the keys
to some of the most successful human rights cam-
paigns (ICHRP, 2003). Furthermore, the ten-
dency of hierarchical accountability to prioritise
accountability to powerful patrons (upward
accountability) – usually comprising donors,
foundations, and governments – who have the
power to progress or retard the achievement of
an NGO’s key objectives has been argued to
result in a severe narrowing of NGO accountabil-
ity relations (Ebrahim, 2005, 2003a, 2003b;
Najam, 1996).
These potentially counterproductive impacts of
hierarchical, narrowly functional, forms of
accountability have led many scholars and NGOs
to seek greater examination of, and practice in,
broader and more holistic forms of accountability
which focus on longer term, sustainable campaign
impacts (Cronin & O’Regan, 2002; Ebrahim, 2005;
Lloyd, 2005; Najam, 1996). These forms of
accountability are motivated more by a sense of
obligation to mission attainment, rather than the
sense of anxiety regarding the power of external
others to deny this possible mission accomplish-
ment – which tends to characterise hierarchical
forms of accountability (Najam, 1996).
Holistic Accountability in NGOs
We use the term ‘holistic accountability’ to refer
to these broader forms of accountability for the
impacts that an NGO’s actions have, or can have,
on a broad range of other organisations, individu-
als and the environment (Edwards & Hulme,
2002a; Najam, 1996). For NGOs like Amnesty
International this expands the concept of ‘perfor-
mance’ articulated within hierarchical accountabil-
ity to embrace quantitative and qualitative
mechanisms concerned with signifying the long
term achievement of organisational mission and
the impact of this achievement in bringing about
structural change. Hence, holistic accountability
augments the short-term monitoring aspects of
accountability focused on isolated campaign
achievements/impacts favoured under hierarchical
accountability with mechanisms of accountability
for the ‘‘second and third order e?ects of NGO
actions” (Edwards & Hulme, 2002a, p. 195). Thus,
hierarchical accountability can be conceived of as
a subset within broader holistic accountability
practices.
Advocates of holistic forms of accountability
tend to recognise that every individual has a basic
right to participate in decisions on matters which
might impact upon them, irrespective of the power
which that individual holds in relation to others
(Unerman & Bennett, 2004). They also sympathise
with the view that moral rights for an individual to
participate in decisions over particular actions
become stronger as the potential impacts on the
life of that individual arising from the decision
become greater (Unerman & O’Dwyer, 2006b).
Thus within holistic accountability, in addition to
the key stakeholders recognised under hierarchical
accountability, the stakeholders to whom an NGO
might be considered accountable include the
groups on whose behalf the NGO advocates, along
with the individuals, communities and/or regions
directly and indirectly impacted by the NGO’s
advocacy activities (Ebrahim, 2005, 2003a; Najam,
1996). This leads to explicit consideration of multi-
ple stakeholder groups, with a signi?cant emphasis
being placed on downward accountability to ben-
e?ciaries (Dixon et al., 2006) in addition to upward
accountability to donors and governments. Issues
804 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
encompassing how responsive and aware NGOs
are of the actual needs of those whom they seek
to assist, as well as the extent of their openness
to involving bene?ciaries and/or partners in assess-
ing the nature and impact of their work, are
prioritised.
For human rights NGOs these impacts cannot,
however, be easily determined, as such ‘‘NGOs are
rarely able to control all (or even most) of the fac-
tors which in?uence the outcome of their work”
(Edwards & Hulme, 2002a, p. 195). As Slim
(2002) notes:
Accounting for the impact or outcome of
[humanitarian] NGO work can be uncertain,
is usually contested and can border on pure
speculation at times as NGOs try to track
cause and e?ect between their actions and
the personal, social, economic, environmen-
tal and political change around their projects
(para. 10).
A recent International Council on Human
Rights Policy (ICHRP) study on accountability
among human rights NGOs concluded that
these NGOs have generally failed to acknowledge
the implications their work has on broader
stakeholder groups. It called on human rights
NGOs to make themselves more aware of
the di?erent e?ects of their work (ICHRP,
2003, p. 135) and to embrace more holistic
accountability forms by broadening their concep-
tion of ‘‘impact assessment” beyond assessing
and reporting on:
. . .how many of the planned outputs of a
campaign were achieved – such as the num-
ber of reports or newspaper articles pub-
lished, [and to] measure their impact on the
behaviour of abusers of human rights or peo-
ple whose rights [a]re under assault (ICHRP,
2003, p. 98).
The report also highlighted a widespread ten-
dency among human rights NGOs to engage in
detailed dialogue with funders and members but
to remain blissfully unaware of the views of those
who were directly threatened by issues they cam-
paigned on:
Very few human rights NGOs seem to have
established procedures to determine how,
once contacted, they will create conditions
in which they will dialogue with, and be
answerable to, such bene?ciaries (ICHRP,
2003, p. 122).
However, a number of international develop-
ment NGOs have recently experimented with
and provided guidance on more holistic forms
of accountability (O’Dwyer, 2007). These include
ActionAid International whose system for
accountability, denoted ALPS (action, learning
and planning system), prioritises downward
accountability to bene?ciaries. ActionAid uses
annual participatory reviews in conjunction with
innovative engagement mechanisms such as the-
atre, people’s art and song in order to encourage
full participation by marginalised groups that
ActionAid are trying to help (ActionAid Interna-
tional, 2006). Social accounting (and auditing)
methods (Cronin & O’Regan, 2002) have also
been adopted by NGOs such as Oxfam (Daw-
son, 1998; Dey, 2007) and have been developed
for NGO usage by organisations such as Key-
stone (Keystone, 2006). These methods use stake-
holder focus groups and surveys in order to
gather the opinions, expectations and standards
required by stakeholder groups. They also
engage in partnerships with stakeholders to deli-
ver on common goals. Other participatory mech-
anisms such as ‘story telling’ whereby
bene?ciaries describe their perspectives in narra-
tive terms using their personal reference points
and means of interpretation have been used by
Voluntary Services Overseas (VSO) and Oxfam
(Cronin & O’Regan, 2002). The Humanitarian
Accountability Project (HAP), a partnership of
member agencies committed to making humani-
tarian action accountable to its intended bene?-
ciaries, has also produced guidance enabling
NGO accountability to bene?ciaries and commu-
nities. This provides advice on prioritising stake-
holder needs, consulting with bene?ciaries,
developing and monitoring complaint and
response mechanisms, and disseminating ?ndings
to bene?ciaries (HAP International, 2003).
Hence, examples exist to enable interested
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 805
human rights advocacy NGOs such as Amnesty
to experiment with more holistic accountability
forms.
Having broadly addressed the conceptions of
hierarchical and holistic accountability which pro-
vide the analytical focus for this paper, the follow-
ing section provides some brief information on
Amnesty International and Amnesty Ireland, in
order to contextualise the ensuing case analysis.
The case context: Amnesty International and
Amnesty Ireland
Amnesty International is a worldwide move-
ment of people who campaign for internationally
recognised human rights with the aim of ensuring
that every person enjoys all of the human rights
enshrined in the Universal Declaration of Human
Rights
4
and other international human rights stan-
dards. Its mission in pursuit of this vision is stated
as follows:
Amnesty International’s mission is to under-
take research and action focused on prevent-
ing and ending grave abuses of the rights to
physical and mental integrity, freedom of
conscience and expression, and freedom
from discrimination, within the context of
its work to promote all human rights
(Amnesty International, 2007).
It promotes itself as a democratic, self-govern-
ing movement answerable only to its worldwide
membership, with all policy decisions being taken
by elected bodies. The organisation also publicises
its impartiality and declares itself independent of
any government, political ideology, economic
interest or religion while not supporting or oppos-
ing any government, political system or the views
of the victims whose rights it seeks to protect. It
won the Nobel Peace prize in 1977 and is viewed
by many as setting the standards for the human
rights movement as a whole (Buchanan, 2002;
Candler, 2001; Ron, Ramos, & Rodgers, 2004).
It claims to have more than 2.2 million members,
supporters and subscribers in over 150 countries
and territories worldwide and professes to be the
world’s largest international voluntary organisa-
tion dealing with human rights. Amnesty Interna-
tional’s national sections and local volunteer
groups are primarily responsible for funding the
movement.
In 2001, Amnesty International’s worldwide
decision making body adopted a new mission by
reforming their historic mandate to embrace work
on economic, social and cultural rights, new
research methodologies, policy priority setting
and advocacy work in areas once considered to
be the primary domain of development NGOs
(Nelson & Dorsey, 2003). This mandate change
also abolished the ‘‘own country rule” which pre-
viously precluded nationally based Amnesty sec-
tions from campaigning on issues in their home
countries. The organisation also became more col-
laborative in its campaigning e?orts, in contrast to
its traditional approach where it had maintained
independence from other organisations (Amnesty
International, 2004; Nelson & Dorsey, 2003).
These developments, while criticised in some quar-
ters for potentially diluting Amnesty’s in?uence
(The Economist, 2007a; The Economist, 2007b;
Hopgood, 2006), re?ect recent trends evidencing
a growing interaction between development and
human rights NGOs emphasising a rights-based
approach to development. This has led to the dis-
course of human rights becoming a common con-
ceptual basis for these NGOs (Nelson & Dorsey,
2003). Amnesty’s involvement in the International
NGO (INGO) charter to promote greater account-
ability by INGOs is consistent with this strategic
focus and is partially fuelled by its desire to pre-
serve the huge trust in which Amnesty Interna-
tional is held among individuals worldwide, a
level of trust that outstrips the trust which many
states and business enterprises enjoy (Williamson,
2005a; Williamson, 2005b). However, despite this
commitment, Amnesty International’s annual
report has recently been criticised for failing to
provide information on the organisation and its
activities (Burall, Neligan, & Kovach, 2003) while
a recent evaluation of INGO accountability by the
UK charity One World Trust (Blagescu et al.,
2006, p. 39) noted that Amnesty International
4
Seehttp://www.un.org/Overview/rights.html. Accessed on
March 20, 2007.
806 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
did not have ‘‘institutionalised mechanisms
through which diverse external stakeholders
[could] engage in decision making processes that
a?ect them” and failed to ‘‘make speci?c commit-
ments to engaging stakeholders in evaluations of
[its] work” (Blagescu et al., 2006, p. 44).
The Irish Section of Amnesty International,
Amnesty Ireland, was o?cially formed in 1962
and currently has a membership base of over
20,000. Its structure comprises around 40 local
groups, with a directorate comprising the Director
and the Secretary General who are overseen by a
10 person executive committee. At the time of
the interviews, these two directors led a nine-per-
son management team comprising six managers
(managing issues such as research and campaigns,
fundraising, training and ?nance), two o?cers
(membership development and human resources)
and a ?nancial controller. The Secretary General
at the time of the research had been, until early
2003, chair of the Amnesty International Execu-
tive Committee for two years and chaired the
search committee that appointed the Amnesty
International Secretary General Irene Khan. Like
all sections, the Irish section cannot take any
action on issues that do not fall within the stated
vision and mission of Amnesty International.
Research method
As this study is motivated by a desire to
increase our understanding of the emergence,
enactment and impact of accountability mecha-
nisms within a particular NGO setting, qualitative
methods have been used in order to engage in
interpretation speci?c to the situation of Amnesty
Ireland. More speci?cally, the perceptions of man-
agers in the organisation have been used, in con-
junction with other qualitative sources, to
illuminate and interpret the nature and impact of
the evolution of accountability processes and prac-
tices within Amnesty Ireland.
The case evidence is primarily constructed from
11 in-depth interviews held over a period of 11
months (from May 2004 to March 2005) with
seven key members of Amnesty Ireland’s senior-
and middle-level management located in its head
o?ce in Dublin. At the time of the interviews the
management team contained eight individuals
who were designated ‘managers’ (including the
Director and the Secretary General) under the
Amnesty Ireland Management Team Structure
(extracted from the organisation chart provided
by Amnesty Ireland). We selected ?ve of these
managers for interviews on the basis of the direct
relevance of their roles to the accountability focus
of the study. These comprised the following: the
Secretary General and the Director (who jointly
head up the management team); the research and
campaigns manager; the fundraising manager –
both of whom had key roles in campaign develop-
ment and implementation; and the ?nance man-
ager. We also interviewed a recently departed
member of the Amnesty Ireland board who had
been highly active over a period of approximately
six years particularly in member level issues and
campaigning but who had recently stepped down
while maintaining an ongoing role in the organisa-
tion. He was an important informant as his active
presence on the board for a number of years meant
he had closely viewed some of the developments in
accountability we wished to study. Our ?nal inter-
view was with the treasurer as some issues pertain-
ing to functional types of accountability were
directly relevant to her role.
In our analysis, we categorised the Director,
Secretary General and campaigns manager as
‘senior’ managers given the nature of their roles
and their position in the Management Team Struc-
ture hierarchy. Using the same rationale, the fund-
raising manager and the ?nance manager were
classi?ed as ‘middle’ managers. The former board
member was also designated as a ‘middle’ manager
as the nature of the core work he undertook in
Amnesty Ireland was less strategic and more oper-
ational in nature. The treasurer worked mainly at
an operational level on a part-time basis and was
thus also designated a ‘middle’ manager.
In addition to the 11 interviews with these seven
managers, we also interviewed ?ve local group
members who had been highly active in the organi-
sation. While these members’ perspectives
informed our case narrative, the paper draws pri-
marily on the managerial perspectives obtained.
A number of other evidential sources were also
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 807
examined including: press releases; publicly avail-
able information on Amnesty Ireland; media
reporting on Amnesty Ireland and Amnesty
International; media articles by Amnesty Ireland
management and board members; and documents
on Amnesty’s mission and strategic objectives
(including its ‘2004–2010 Integrated Strategic
Plan’).
A series of semi-structured interviews were ini-
tially held over the period May to August 2004.
All interviewees were sent a broad outline of the
issues to be discussed prior to their interview. Each
interview was tape-recorded and fully transcribed
and lasted between 45 min and one and a half
hours. The interviews were designed to instigate
a conversation about how accountability was expe-
rienced within Amnesty Ireland and why it was
experienced in the manner outlined. No de?nition
of accountability was presented to interviewees
and questions were asked in a naturalistic manner
(Patton, 2002). Each interview commenced by
referring generally to increased demands for
NGO accountability. This was designed to insti-
gate the conversation and we allowed interviewees
to lead this discussion in the initial period of the
interview. This normally led interviewees to speak
speci?cally about Amnesty Ireland with the focus
placed primarily on their experiences of account-
ability within Amnesty Ireland. We probed on
the internal and external pressures driving
accountability in Amnesty (if they were seen to
exist) and the perceived uniqueness of Amnesty’s
accountability requirements. Detailed notes were
taken throughout each interview, and analysis of
these helped build up a pattern of issues for further
in-depth probing in the subsequent interviews. It
also provided a provisional emergent record of
analysis and interpretation prior to the formal pro-
cessing and interpretation of the evidence
gathered.
The subsequent analysis of the interview evi-
dence involved three connected, disciplined sub-
processes: data reduction; data display; and con-
clusion drawing/veri?cation (Huberman & Miles,
1994; O’Dwyer, 2004). Initially a post-interview
analysis of transcripts was conducted through a
close reading of all transcripts and accompanying
notes and tape-recorded re?ections on interviews
in order to search for underlying themes in the evi-
dence collected. A coding scheme was intuitively
developed in order to aid in the identi?cation of
the themes emanating from this analysis. These
initial themes were then compared and grouped
under broader overarching themes. They were
then summarised into an 80-page synthesis of the
main ?ndings which made extensive use of direct
quotations from the transcripts. The codes were
then collapsed into a core code for each overarch-
ing theme, and matrices summarising these codes
for each interviewee were developed. These dis-
plays aided in identifying cross case patterns in
the interview data (Huberman & Miles, 1994) with
the predominant codes becoming evident partially
by mapping the relative incidence of di?erent
codes. An examination of these matrices enabled
the recognition of regularities, patterns and expla-
nations in the evidence collected. A number of
seemingly contradictory ?ndings also emerged
from this analysis. A comprehensive mind map
was then devised to illustrate the various themes
and reveal their interrelationships. This was devel-
oped iteratively during subsequent readings of the
transcripts and ?eld notes and eventually served to
provide a detailed visual interpretation of the main
themes identi?ed. As part of this process, tran-
scripts were also re-visited where deemed
necessary.
During this analysis phase, in March 2005 four
members of the management team were re-inter-
viewed to clarify certain issues that had emerged
and to gain some elaboration on issues which
seemed to con?ict in the analysis. An initial draft
of this paper was then prepared and was sent to
Amnesty Ireland managers for comment. Subse-
quent drafts of the paper bene?ted from informal
feedback given by some interviewees who read
these drafts and attended presentations of the
paper at academic conferences.
The case ?ndings
This section proceeds to analyse the emerging
accountability issues in Amnesty Ireland as identi-
?ed from the empirical study. The ensuing narra-
tive is structured into four sub-sections dealing
808 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
with distinct but interrelated themes. The ?rst sub-
section explores recent trends in Amnesty’s
accountability which have seen a historical focus
on internal accountability mechanisms, aimed at
building and maintaining public trust in Amnesty
through the use of protocols to ensure the accu-
racy of its public pronouncements, broadened to
encompass external forms of accountability aimed
at demonstrating how well Amnesty has ful?lled
its human rights mission. Within Amnesty Ireland
we found a clear recognition of the desirability of
adopting holistic forms of external accountability.
However, as is discussed in the second sub-section,
despite a desire to be downwardly accountable
there were considerable obstacles to identifying
in advance, and engaging in an accountability dia-
logue with, many of the (potential) bene?ciaries of
Amnesty’s work. Conversely, upward accountabil-
ity to some (but by no means all) donors was con-
sidered less problematic and more achievable. This
has e?ectively resulted in a ‘crowding out’ of prac-
tically problematic downward forms of account-
ability with more realisable upward forms of
accountability. The third sub-section identi?es
that, consistent with the literature discussed above,
this resultant focus on upward accountability has
encouraged hierarchical forms of accountability.
These have manifested themselves in a strategic
preoccupation with narrow forms of performance
measurement. The fourth sub-section then exam-
ines perceptions regarding how these performance
measures can be, and are, counterproductive in
terms of focusing on some issues that work against
the more holistic achievement of Amnesty’s broad
human rights mission. Overall, this case material
demonstrates the highly complex and potentially
counterproductive nature of a particular account-
ability focus in the context of a high-pro?le inter-
national NGO.
Trends in the nature of Amnesty’s accountability –
from an internal to an external focus
Amnesty Ireland has historically focused much
of its accountability on supporting and reinforcing
the veracity of its pronouncements on campaign
issues:
There has always been almost an obsession
in Amnesty about control over the quality
of information if we are doing research.
For example, if your research was being
commissioned by the International Secretar-
iat you would have to submit it to your team
leader in whatever programme you were
working. It would then have to go to the
head of programme, it would have to go to
the lawyers, it would have to go to the cam-
paigners and then if it was changed in any
way it would have to go back up through
the process again before it gets to the public
domain because if we screw up on one report
it’s not just that, it’s that next month when
we are producing a new report, the [Irish]
government will say ‘look, you got it wrong
on Turkey’ so your credibility, your reputa-
tion, the trust people place in you is almost
gone. You can come back from a ?nancial
crisis but if there is a collapse in trust or that
relationship then it is very hard to rebuild
that. (S)
5
In seeking to maintain the widely held credibil-
ity of its public pronouncements, Amnesty Inter-
national’s latest strategic plan reiterates the need
for internal accountability protocols by articulat-
ing an aim ‘‘to strengthen [internal] accountability
within a system that is less rule-based and provides
more ?exibility for action and growth”. This, it
claims, will involve ‘‘encourag[ing] ‘tough trust’”
in the form of ‘‘minimum rules with maximum
accountability” in order to ensure any public utter-
ances are fully defensible (Amnesty International,
2004, p. 17).
Reinforcing the importance of these internal
accountability protocols in ensuring Amnesty’s
pronouncements are based on accurate material,
the interviews revealed widespread apprehension
among senior managers that external critics
5
Interviewees are denoted throughout the narrative by the
letter S for senior managers and the letter M for middle
managers. We have not separately identi?ed each manager (for
example, S1; S2; S3 etc) because we agreed anonymity for each
interviewee and given the small number of managers at
Amnesty Ireland it may be possible for an informed person to
be able to identify speci?c quotes to speci?c managers if we
separated quotes by individual manager.
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 809
would, in future monitor Amnesty more carefully
and ‘hold’ it even more accountable for its public
pronouncements. Relations between Amnesty
and sections of the Irish government were strained
at the time of the interviews. Amnesty had publicly
condemned the Irish government’s immigration
law and policy as being out of step with interna-
tional human rights commitments (Holland,
2004; Love, 2004a; Love, 2004b), and the govern-
ment’s delay in ratifying the European Convention
on Human Rights was publicly denounced by
Amnesty as being ‘‘pathetic” (Cummins, 2002,
see also, Love, 2002). Furthermore, when Amnesty
researchers were refused access to Irish prisons for
a project on racism Amnesty accused the Irish gov-
ernment’s Department of Justice of being ‘‘the
most secretive and paranoid institution in the
[Irish] State” (Haughey, 2002). In the mainstream
print media Amnesty’s Director also explicitly
referred to perceived hostility towards Amnesty
by the government:
. . .it’s the ?rst time we have encountered
active hostility from a home government . . .
we have concerns that there is a sensitivity
in this government (Humphreys, 2004).
Moreover, the nature and content of Amnesty’s
anti-racism campaign was publicly criticised and
challenged by certain politicians and high pro?le
print journalists. One media commentator coun-
tered ‘‘that [while] there is racism in Ireland. . .it
is neither as prevalent nor as widespread as
Amnesty would seek to suggest” (Waters, 2001).
In common with many others in the op-ed and let-
ters pages of broadsheet newspapers, this com-
mentator lambasted Amnesty for their
advertising campaign which implicitly accused cer-
tain government ministers of tolerating racism
(Andrews, 2001; Haughey, 2001; Humphreys,
2004; Labanyi, 2004; Waters, 2001). Amnesty
remained publicly unrepentant about the cam-
paign (Donnellan, 2001; Haughey, 2001) despite
being censured by the Advertising Standards
Authority of Ireland (ASAI). Our interviews
revealed that this increased scrutiny and critique
led to a predominant concern that Amnesty should
pay even closer attention to the core tenets of its
mission and statute, focused on systematically
and impartially researching the facts of individual
cases and patterns of human rights abuse
(Amnesty International, 2007).
6
While this historical focus within Amnesty on
internal accountability protocols to ensure the
veracity of its public pronouncements helped to
counter many emerging external concerns, this
focus alone was now no longer deemed su?cient
to attract further funding and support. There
was a widely held perception among the managers
we interviewed that the rapidly evolving, more
competitively driven, donor dominated NGO
world now sought greater external accountability
for impacts in addition to accountability for pro-
nouncements. This led to a growing conviction
that this broader need for accountability required
the implementation of a range of external account-
ability mechanisms.
Senior managers viewed this move to greater
external accountability as essential to the protec-
tion and di?erentiation of the credibility and legit-
imacy of Amnesty’s ‘‘highly trusted brand” (S).
This, in turn, should directly facilitate (and/or pro-
tect) the achievement of Amnesty’s core mission of
undertaking research and action focused on pro-
moting human rights. In contrast they also per-
ceived threats to the brand, and the consequent
achievement of the mission, arising from any per-
ceived absence of external accountability:
We stand for something, we say you can help
change the world and they [donors and mem-
bers] commit to that so they are placing their
trust in us. If you lose that, if you damage
6
Shortly after the ?nal interviews were undertaken for this
study, Amnesty Ireland also came under public attack from
some media quarters for the nature of their ‘stop violence
against women’ campaign. One in?uential commentator
claimed that ‘‘[t]he idea that there are groups or individuals
out there seeking to deny adequate protections to women
genuinely at risk from violence is a ?gment of Amnesty’s
imagination” (Waters, 2005). This commentator further con-
tended that ‘‘through lack of rigour and even-handedness and
the adoption of a pernicious ideological position, Amnesty has
undermined its own venerable human rights tradition” (Waters,
2005). Amnesty Ireland’s Director publicly defended the
organisation from these attacks (Love, 2005) and this also led
to some heated comment in the letters pages of high circulation
broadsheet newspapers such as The Irish Times and The Sunday
Business Post.
810 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
that relationship, it is very hard to get it back
. . . [Demonstrating] a commitment to
accountability helps us maintain that credi-
bility and trust. (S)
Amnesty is completely dependent on its rep-
utation and accountability is central to that.
We know from independent studies that
Amnesty is the most trusted brand in Europe
and the second most trusted brand in the
world . . . We are so, so careful about what
we say and do, if we say it you’ve got to trust
it if Amnesty says it’s true. (S)
Amnesty Ireland’s increasing competition with
other NGOs for public funds heightened this per-
ceived need to distinguish itself through its com-
mitment to external accountability, especially as
there was widespread evidence of ‘‘donor fatigue”
(M) threatening ongoing funding. This di?erentia-
tion was considered crucial to Amnesty’s survival
in a tough ‘‘NGO marketplace” (S) where it was
‘‘competing for ?nances and hearts and minds”
(M). For example, development NGOs were
viewed as posing a speci?c threat to the Amnesty
‘brand’ given their emerging use of ‘‘the language
of human rights” (M) in their campaigning. This
endangered ‘‘the [marketed] vision of Amnesty in
terms of protecting human rights” (S). Develop-
ment NGOs could also display e?ectiveness more
easily in their campaigns which many managers
believed conformed with various public donors’
desire to fund only ‘‘[NGOs] which [we]re e?ective
and c[ould] demonstrate that they [we]re not wast-
ing money” (M).
The distinction between the Tro´ caires, Con-
cerns, Goals etc. [large Irish development
NGOs] and the human rights organisations
like Amnesty, for example, was getting fuzzy
was blurring to an extent. If you were to
open The Irish Times [Irish broadsheet news-
paper] for example and you were to see the
full page spread from Amnesty Interna-
tional, from Tro´ caire, or Concern, it [was]
sometimes not that easy to actually distin-
guish who it was who put in the full page
spread . . . I suppose, the market place for
goodwill or for charitable organisations [is]
getting more cluttered . . . and a lot of the
development agencies have started to use
the language of human rights as well and
have been speaking in terms of rights based
development . . . There is a recognition that,
?rst of all, Amnesty needs to keep its share
of the market . . . [Hence,] Amnesty felt that
there was a need to be accountable to the
general public in terms of how our money
was being spent but also to do it in terms
of somehow distinguishing . . . between us
and the other organisations. (M)
While much of this concern to embrace external
accountability was motivated by a preoccupation
with upward accountability to potentially power-
ful mission-contingent stakeholders, there was,
however, also widespread recognition of the need
and desirability for Amnesty to additionally be
accountable to a much broader range of stake-
holders – in particular the bene?ciaries of
Amnesty’s human rights campaigns (or their rep-
resentatives). Given that the nature of managers’
identi?cation of the stakeholders to whom they
needed to be accountable was a crucial factor in
the quest to realise the opportunities of greater
external accountability, while de?ecting the threats
of inadequate accountability, the following sub-
section proceeds to explore perceptions on stake-
holder identi?cation and accountability in depth.
Upward accountability crowding-out downward
accountability
As the core of Amnesty’s mission revolves
around promoting human rights, managers were
especially keen to indicate a need for downward
accountability in that bene?ciaries ‘‘should ide-
ally” (M) constitute a key stakeholder group for
Amnesty. However, the perceived practical di?-
culties of providing any form of accountability to
these bene?ciaries as well as the fact that ‘‘you
don’t see them demanding accountability [meant]
they [we]re not explicitly considered” (M). Hence,
they were deftly, albeit reluctantly, downplayed by
many interviewees. There was a sense that while
mission achievement did result from assisting these
groups or individuals, accounting to them as
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 811
stakeholders (by, for example, asking them or their
representatives to assess Amnesty’s e?ectiveness)
was not where the priority for accountability in
Amnesty should or could lie. This was also in?u-
enced by bene?ciaries’ perceived inability to
enforce accountability on Amnesty:
Your primary bene?ciaries, who are the peo-
ple, the poor people who are a?ected by
human rights violations are key stakehold-
ers. You can keep on patting yourself on
the back and say you are doing good work
but are the people you are working on behalf
of, do they think you are doing good work?
Are you asking them what they think of your
work? . . . That’s part of the accountability
system . . . but it’s so much harder [in prac-
tice] to be accountable to these people. (S)
In terms of accountability, I suppose we
should leave aside discussion of the fact that
the victims of human rights abuses that
Amnesty would be working for would ideally
be the primary stakeholders. (M)
The di?culties of identifying this ‘‘very large
constituency” (M) of bene?ciaries (or their repre-
sentatives) were deemed enormous ‘‘despite a clear
obligation to defend people whose rights [we]re
being abused” (S). Prioritising easily de?ned and
more powerful stakeholders such as donors was
considered more relevant for pragmatic survival
and growth related reasons, even if this caused
some concern among Amnesty researchers work-
ing in the ?eld.
7
Within Amnesty it is very di?cult, as the
researcher working on country X will feel
?rst and foremost that they are accountable
to the victims of human rights violations or
on behalf of whom they work, but there is
no formal accountability to them . . . but then
they have to be accountable to the Amnesty
membership and then they might also have
to be accountable to external funders . . . they
often feel that the prioritization is not in the
right place. (M)
Managers made no reference to any explicit
consideration of means of enhancing dialogue with
(or accountability to) other bene?ciaries or their
representatives. While Amnesty’s research draws
heavily on key informants or representatives of
bene?ciaries, there was little evidence that these
individuals were also used to in?uence the nature
of Amnesty’s work and its overall e?ectiveness.
The problems, as opposed to the possibilities, of
accountability to such stakeholder groups were
emphasised. Managers also displayed little knowl-
edge of the plethora of holistic accountability
forms emerging within development NGOs aimed
at enabling greater engagement with bene?ciaries,
as outlined in the exploring NGO accountability
section above (see O’Dwyer, 2005). Thus, although
there was recognition of the desirability of being
downwardly accountable, mechanisms aimed at
enacting downward accountability were not evi-
dent in Amnesty Ireland.
This general lack of engagement with more
holistic accountability forms contrasted with the
level of active attention to upward accountability
mechanisms which became central to the emerging
external accountability focus. This was partly
re?ected in the business-focused terminology used
by many managers. There was widespread use of
terms like ‘‘NGO competition”, ‘‘donor markets”,
‘‘branding”, ‘‘added value” and ‘‘the business of
Amnesty”. This language was consistent with the
business-oriented approach to the formulation of
the current Amnesty International strategic plan
(Williamson, 2005a). While some managers out-
lined their unease with this terminology, they indi-
cated that it was now prevalent in their ‘‘corporate
NGO” (S) world. One senior manager insisted that
while he ‘‘hate[d] the term [corporate NGO]” (S),
an aversion to ‘‘corporate speak” (S) was too often
used to imply that NGOs were in some way di?er-
ent when it came to applying standards of
accountability. This, he suggested, was not sus-
tainable in the future ‘‘given the amount of money
7
Some of the campaigns resulting from the recent Amnesty
International mandate change abolishing the ‘home country’
rule (which previously precluded a national section campaign-
ing on issues in their own country) such as the campaign to stop
violence against women and girls have, nevertheless, explicitly
recognised and engaged with bene?ciaries as key stakeholders
where this was practicable (S). However, this was primarily due
to these stakeholders being more easily accessible and more
likely to be vocal and in?uential.
812 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
that’s now going through the accounts of global
NGOs” (S) and the consequent scrutiny that this
was attracting:
Traditionally, NGOs have tried to evade cor-
porate speak and corporate accountability,
in a sense that NGOs were di?erent. Di?er-
ent is okay, but what is not okay is di?erent
standards and we’ve got to hold ourselves to
the highest standards . . . To do that di?er-
ently is okay, but we must make sure we’re
hitting the right standards. (S)
Consistent with this business-focus, and in com-
mon with the ?ndings prevalent in much corporate
accountability research (Unerman & Bennett,
2004), senior managers especially focused external
accountability practices upwards towards those
stakeholders who appeared to hold the most eco-
nomic and political in?uence over Amnesty Ire-
land. A separation in Amnesty Ireland’s
accountability relationship between the wider
membership base (‘ordinary members’) and more
in?uential ‘high net worth donor’ members
(‘extraordinary members’) was evident. ‘Ordinary’
members of Amnesty Ireland were considered key
stakeholders in that participatory decision making
processes were established to make sure they felt
part of the organisation, and to try to enable a
sense of shared responsibility for campaigns
between ‘ordinary’ members and head o?ce sta?
in Dublin. An inferred need to demonstrate e?ec-
tiveness to these members was considered central
to maintaining members’ faith and motivation:
We must manage this relationship [with
members] in a constructive way otherwise
you are losing members, bringing new people
in, setting them up, investing time, money
and energy [and] to what end? . . . If you go
to talk to anyone in Amnesty and say ‘well
what motivates you or what would motivate
you to be more active or to stay with the
organisation?’ [They will say] . . . ‘when you
know what is going on, you are informed
about what is going on, when you are part
of the process, you know what you are
expected to do, and crucially you know what
the outcome [of your activity] is, feedback all
the time’. So, that sense of feeling . . . that we
have delivered, that we have achieved, it’s a
central part of the whole principle of
accountability, it’s central to the whole psy-
chological principle of maintaining motiva-
tion and activism. (S)
However, this desire to exhibit accountability to
‘ordinary’ members met with considerable apathy.
Most ‘ordinary’ members did not actively demand
accountability. Once they joined Amnesty, they
exhibited ‘‘a great sense of loyalty and trust” (S)
and remained passive as long as they ‘‘hear[d] a
story . . . about how their involvement [had] made
a di?erence” (S). This ‘‘lack of [member] engage-
ment” (S) rendered many internal participatory
decision making processes redundant and disap-
pointed some middle managers:
I have to say that [from] my own experience
at annual conferences over the last few years,
you could put up complete acts of ?ction as
the ?nancial accounts for the year and the
membership wouldn’t even notice. (M)
This passiveness appears to have led to a devel-
oping and, for some, worrying divide between the
board and central management sta? and the wider
‘ordinary’ membership. ‘‘Unquestioning trust”
(M) was now placed by members in the ‘‘profes-
sional trained group of sta?” (M) located in Dub-
lin whose ‘‘specialisation ha[d] . . . outstripped the
average member in terms of what [Amnesty]
d[id]” (M), further indicating a general lack of
accountability demands from ‘ordinary’ members.
In the absence of actual information and
accountability demands from most of these ‘ordin-
ary’ members, the needs of ‘extraordinary mem-
bers’ – ‘‘high net worth in?uential donors” (S)
and foundations who represented a ‘‘signi?cant
cohort of supporters beyond the membership
[and] who [were] investing in . . . [Amnesty] because
they believe[d] in what [it was] doing” (S) – were
prioritised. Senior managers avoided referring to
these stakeholders as ‘members’ per se. ‘Extraordi-
nary’ members were seen as somewhat external to
Amnesty and were a?orded the role of overseers of
Amnesty’s performance given their emerging in?u-
ence. An increased consciousness about how
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 813
Amnesty was seen and judged by these stakehold-
ers led senior managers in particular to actively
engage in attempts to ensure they presented an
image of Amnesty in line with these stakeholders’
often presumed expectations, which revolved
around increased demands for reporting on how
speci?c donations were achieving an impact:
These are people who give their time, like for
example artists, or signi?cant in?uentials who
give us their name, or people who give us large
sums of money ... [Their] belief has to be
restored or matched and recognised in terms
of a sense of accountability owed to them. (S)
Amnesty Ireland had only recently instigated
plans to focus speci?cally on these high net worth
donors, who were growing in number in Ireland’s
rapidly advancing economy. Prioritising account-
ability to them was crucial given they were per-
ceived as an especially critical source of funding
and support in Ireland in the coming decade:
High net worth donors are quite discerning
and they are going to be very careful where
they put their money and ?nancially they
are going to be critical [for Amnesty Ireland]
otherwise there is not going to be any money.
So they are a critical stakeholder to be fac-
tored in . . . A plan to focus on them [high
net worth donors] is beginning and is only
starting at the moment. X [name of board
member] is very aware that to be serious in
working with some of these people they are
going to expect to see fuller reporting of
the e?ectiveness of their donations. (M)
The next sub-section illustrates how this pre-
dominant focus on upward accountability to these
‘extraordinary members’ has led to the promotion
of narrow, often underspeci?ed, hierarchical
accountability forms.
Hierarchical accountability and the emergence of
performance measurement
Concerns to develop the ‘extraordinary mem-
bers’ donor base and eliminate the likelihood of
their exit had initially led to informal, private,
face-to-face accountability mechanisms evolving
on an ad hoc basis in order to reassure them as
to Amnesty’s e?ectiveness in using their donations.
Senior managers evidently disciplined themselves
in advance of these meetings to ensure they pre-
sented a proper ‘face’ to these donors, focused
on illustrating Amnesty’s e?ectiveness:
High net worth donors are very developed
relationships, we have an open conversation
with them whether they are foundations or
individuals where we’re telling them where
we’re having an impact and showing them
the evidence, but that doesn’t necessarily
have to be public. (S)
Among Amnesty Ireland’s senior managers
there was a widely held perception that these
stakeholders had a preference for indicators of e?-
cacy, and this fuelled a desire to develop perfor-
mance accountability indicators. This drive to
further and more formally establish ‘‘indicators
of achievement” (S) to show how this ‘‘investing”
(S) public got ‘‘bang for their buck” (S) was con-
sidered crucial in order to ‘‘retain relevance” (S)
and compete ‘‘in [the] crowded [NGO] market”
(M). The aforementioned focus on brand di?eren-
tiation through demonstrating accountability for
performance was additionally motivated by an
emerging ‘‘success driven [Irish] society” (S) from
which these future donors would materialise
demanding veri?able proof of Amnesty’s impacts:
Our reading of trends is that there is going to
be a much more signi?cant public awareness
and understanding of the e?cacy and impact
of NGOs going forward and the public is
going to be much more judicious around
how it’s investing its charity money . . . The
public is becoming much more discerning
and they’re much more anxious to get bang
for their buck, and they’re beginning to look
for indicators of achievement, they’re begin-
ning to look for indicators of impact, they’re
beginning to look for indicators of e?cacy
and that means that the reporting mecha-
nisms and the communication mechanisms,
the conversation between the organisation
and the donor . . . has to be much more
814 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
informed and based on much better informa-
tion. (S)
A further impetus towards this strategy of
developing performance measurement metrics
internally (the de?ning of which would therefore
be under the control of Amnesty) was the possibil-
ity of inappropriate measures being externally
‘‘imposed” (M). For example, certain managers
were incensed at recent print media usage of a
funding/administration ratio as a measure of Irish
NGO performance. Not only was this ratio
deemed severely simplistic and inappropriate, but
other NGOs were accused of abusing it by de?ning
‘administration’ narrowly thereby improving their
reported ratio. This externally imposed focus on
‘‘bogus measures of e?ciency” (M) neglected any
substantive consideration of NGO e?cacy and
made ‘‘administration heavy NGOs like Amnesty”
(M) appear ine?cient:
Di?erent organisations use di?erent indica-
tors. There is a notion that you can quantify
administrative costs and you see organisa-
tions, particularly the development organisa-
tions, saying that they spend a certain
percentage of their money on administration
which, of course, is an entirely notional per-
centage. It all depends on your internal de?ni-
tion of administration and some
organisations like Amnesty would argue that
the costs of hiring a car to drive somebody
from Khartoum to Darfur is administration
whereas other organisations would argue that
no, that’s ?eld based activity, that’s the real
work. I mean it’s hiring a car, what’s the di?er-
ence whether you’re hiring the car here?
Amnesty takes the view that everything is
about administration . . . we regard ourselves
as an administration . . . that people invest in
because they know that we have the know-
how to get the right people into a car from
Khartoumto Darfur, to get the right informa-
tion back out. That is about administration
and we are much more upfront about it. (S)
Within the above contexts, many managers
elaborated on a quest within the Irish section to
‘‘prove” Amnesty’s success at ‘‘delivering change”
(S) by developing internally and reporting exter-
nally on ‘‘indicators of e?cacy” (S) in order to
‘‘measure progress” (S). This had involved a
‘‘rethink of the operational planning process” (S)
where Amnesty Ireland now ‘‘focus on where
[they] can be e?ective” (S) and is consistent with
the current Amnesty International strategic plan,
which stresses the ‘‘development and implementa-
tion of speci?c projects that target measurable out-
comes” as part of a process of promoting ‘‘an
outcome-oriented approach to [Amnesty’s] work”
(Amnesty International, 2004, p. 31). Thus far,
managers have concentrated on establishing
mainly quanti?able measures of e?ectiveness on
an ad hoc basis. These include ‘‘shifts in public
opinion [measured] in terms of media cover-
age,. . .number of changes in speci?c pieces of leg-
islation [and] changes in the way the judicial
system works” (S). Amnesty also now engages in
market research aimed at enabling such forms of
campaign assessment:
We’ve tried to become more rigorous about
setting . . . qualitative and quantitative indi-
cators of achievement, and trying to measure
campaign performance against those indica-
tors so all of our campaigns are set against
a range of qualitative and quantitative
benchmarks which we measure against, they
are internal to our work and planning . . .
For example, we can measure resource allo-
cation . . .we know if we put a certain amount
of people on an issue for a certain amount of
time what happens . . . that’s one way of mea-
suring what we’re doing. (S)
Amnesty’s campaign against racism is a speci?c
example of this emerging focus on performance
measurement. One of the campaign aims was to
increase public awareness of racismin Irish society.
To determine its e?ectiveness, Amnesty managers
needed to establish some measure of public aware-
ness which could be assessed at the beginning of the
campaign and after a period of campaigning.
Quantitative and qualitative measures of media
coverage over the campaign period (number and
nature of articles etc.) were used as proxies for
changes in public opinion. The extent of the media
coverage at the beginning of the campaign was then
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 815
used as a ‘‘benchmark” (S) against which coverage
at the end of the campaign (or at points throughout
the campaign) was assessed to discover whether
opinion had changed over the campaign period:
We are much more speci?c, we are much
more focused in terms of how we structure
our campaigns and agree the outcomes and
how you measure that and how you set the
goals. We have never really had money for
research but we have started doing [market]
research. Maybe at the beginning of the cam-
paign getting questions into the research.
[Asking] what are people’s views on racism?
Who do they think is taking the lead? Do
they think the government is doing enough?
And then revisiting that in a two year time
frame, so at the end of the ?rst two years
of this major campaign we have some bench-
mark as to whether there is a shift in public
opinion The other would be in each phase
you might say that public awareness is a goal
of the campaign. (S)
While this performance accountability focus
pervaded the senior manager interviews, we did
not uncover many detailed examples, other than
those mentioned above, relaying exactly how this
emerging outcome-oriented accountability focus
was being implemented. Rather, senior managers
were deliberating at a strategic level on the need
for implementation of detailed performance met-
rics, with only a few such metrics (as the time of
the interviews) having been developed and imple-
mented at the operational level.
8
Middle managers
had di?culties with several aspects of the perfor-
mance metrics that had already been implemented
and were both wary and sceptical of the strategic
push for ever more such metrics. This was consis-
tent with many middle managers’ admittance that
they were struggling to develop accountability
mechanisms to ‘‘rationalise” (M) their campaign
activities. Indeed, this focus on the often inferred
accountability desires of powerful stakeholders
has operated to downplay some managerial con-
cerns about the potentially dysfunctional conse-
quences for Amnesty’s mission (and internal
cultural norms) of adopting primarily hierarchical
accountability forms. These concerns are discussed
in the following sub-section.
Performance accountability and mission
achievement – problematic bed fellows?
Although several managers emphasised the
importance of the strategic pressure towards per-
formance accountability, there was dissent and
much soul searching among middle managers in
particular with respect to the ultimate contribution
this could make to the achievement of Amnesty’s
mission. Firstly, it was claimed that measuring
‘success’ was often impractical and the strategic
pressure could lead to a focus on convenient but
potentially misleading quanti?able performance
measurement devices. Secondly, by acting counter
to an internal culture which resisted taking direct
external credit for campaign successes, the trend
could act to diminish the potential of future cam-
paigns if such proclaiming of direct in?uence dis-
pleased powerful campaign targets such as
governments. Thirdly, by providing external and
internal visibility to so-called ‘failed’ campaigns,
de?ned using frequently ?awed measures,
Amnesty’s mission-led focus might be distorted.
For example, this might threaten the future of
key long term campaigning (such as lobbying)
for which no direct cause and e?ect could often
be shown. The remainder of this sub-section
expands on these perceived accountability-related
threats to mission achievement.
Several of our middle management interviewees
appeared somewhat confused and anxious about
the strategic accountability trend towards greater
use of performance measurement. In particular,
there was widespread concern as to the feasibility
of developing performance measures:
I come across a lot of desire to do it [measure
and report on impacts] but little or nothing
in terms of the ability to do it. (M)
8
This is consistent with the latest Amnesty International
strategic plan which is also a somewhat short on speci?cs
regarding the exact accountability mechanisms that should be
used.
816 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
I really don’t think we have very e?ective
tools to measure our impact . . . however, if
we could be convinced by some sort of for-
mula or mechanism that was true I think
we would be more open to change. (M)
How do you measure raising awareness? It
takes a long time and maybe when something
does happen in the end it is quite di?cult to
measure, I’m not sure how to do it . . . how
do we know that our campaign on mental
health improved the lot of people with men-
tal health problems? It is a very di?cult thing
for us to measure. (M)
It may be the case that you work on some-
body who is in jail for 3 years and then they
are released and you think great, but then
you discover if you hadn’t taken up their
case and had all the publicity they might
have only been in for a year. How do you
measure that, how do you know whether
it’s right or wrong? (M)
There appeared to be some tension between cer-
tain senior and middle managers as to the ease
with which performance accountability measures
might be introduced, although senior managers
also sometimes seemed to di?er as to the possibil-
ities of accountability in this vein. Middle manag-
ers were closer to the campaign coalface and
closely involved in developing or supporting spe-
ci?c campaigns. Many were dealing with frustra-
tions on a daily basis and the idea that their
campaigns might now be assessed in what they
were concerned could be a ?awed manner added
to their sense of annoyance, as did the prospect
of unwelcome critical scrutiny. These managers
were also more aware of the potentially time con-
suming nature of the development and enaction of
these new mechanisms of accountability, and the
impact this could have on the amount of ‘real’
work they were able to do:
There are some complicated formulae for
measuring how much press work you get,
to compare press coverage this year as
opposed to last year . . . I give up to be honest
because the amount of time that I would
have to invest in collecting all that data
would not be worth it in the end and I could
be o? doing some real work. (M)
Some people do not welcome that sort of
scrutiny as it might make their or my work
look ine?ectual or pointless . . . most people
would be shocked [at the scrutiny]. (M)
Much of this apprehension was re?ected in the
view that aiming for clear, simple measures might
prove misleading. One middle manager articulated
his unease with this trend by using his recollections
of an attempt to formally measure the impact of
an international campaign in Peru aimed at reduc-
ing killings and disappearances. In this situation,
numerical impact measurement was seen as wholly
misleading as it ignored the underlying context
thereby simplifying the inherent complexity in
the circumstances being assessed:
When I was working in the campaigns pro-
gramme in Amnesty in London, Amnesty
had run a campaign on human rights in
Peru, it was in the late 1980s when the war
in Peru was in full swing, there were massive
numbers of disappearances and killings and I
had not been involved in the campaign . . .
[but] I was asked to go in and do the evalua-
tions. So I sat there with the researcher in
Peru and said ‘so, what was the main focus
of the campaign?’ It was to try and push
the local authorities to exert more account-
ability over the military to reduce the num-
ber of killings and disappearances. Okay,
so, during the course of the campaign did
the number of killings and disappearances
decrease? He said ‘oh no, they increased
hugely’. I said ‘oh, so the campaign was a
disaster then?’ He said ‘no, I think they
would have increased more if the campaign
hadn’t happened’. The main dynamic was
the war and the number of innocent civilians
was a factor of the intensity of the war, he
felt that the politicians had felt the interna-
tional pressure from the letters and had
started to try to control a lot of it more, that
there were feelings that some senior military
commanders had felt some pressure and it
had gone down the line and in some places
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 817
there had been a reduction in the killings,
[but] it is impossible to measure that. (M)
A further key anxiety about the accountability
trend related to a concern that Amnesty should
never take direct external credit for the outcomes
of successful campaigns. This ‘‘culture of reti-
cence” (S) was deemed sacrosanct by senior man-
agers in particular and there was widespread
insistence that there was little possibility of this
commitment being reneged upon:
Externally, Amnesty is culturally opposed to
taking credit for work. It’s a cultural tradi-
tion in Amnesty where we just don’t and
won’t and probably will never take credit. I
can tell you now the names of individual
people, I can tell you about major cases
where we made huge positive gains, where
if it were not for Amnesty situations
wouldn’t have changed. Whole governance
systems in countries have changed and we
know that it is because of Amnesty . . .
Another example is major pieces of interna-
tional law [that] have been put in place pre-
cisely because of Amnesty’s work, we can
track our involvement, we can know that
we were responsible but we are never going
to say that. (S)
Taking explicit credit also risked proving coun-
terproductive in that intransigent governments
might react by reversing decisions Amnesty
claimed to have directly in?uenced. This was a
particularly sensitive issue given the aforemen-
tioned nature of relations with the Irish govern-
ment at the time of the interviews:
. . . if we are campaigning on a prisoner case
like X [name of prisoner] for example, has
just been released in Turkey, a prisoner of
conscience, human rights activist and lawyer.
We would not go out trumpeting, we would
celebrate the fact that she had been released,
but we won’t go out and claim necessarily
that it was a direct result [of our actions]
. . . The Turkish government is never going
to say ‘we’ve released this person because
of Amnesty’s campaigning’ . . . and they
might even reverse some decisions if we
trumpet too much . . . so it’s kind of policy
almost not to [claim credit]. (S)
The apparent threat to this ‘‘culture of reti-
cence” from the emerging performance account-
ability focus led us to probe as to how the two
could comfortably co-exist. Some senior managers
found this questioning provocative and insisted
that ‘‘it was not a black and white issue” (S) and
that there were occasions where de?nitive credit
could be taken; for example, in relation to Christ-
mas card campaigns to get prisoners released.
However, other middle and senior managers con-
ceded that this apparent contradiction was some-
thing which had been regularly ‘‘discussed in the
corridors” (M) given the clear tensions exposed
by the emerging commitment to establishing and
reporting on measures of achievement. One middle
manager was highly sceptical and claimed that the
cultural reticence was ‘‘a fairly quaint thing that
was rooted in Amnesty’s infancy” (M) and was
subtly overridden by Amnesty’s use of ‘‘highly
suggestive” (M) testimonies in their advertising:
We have the rule that we never claim credit
for X and Y but we cheat on that all the time
by pumping our literature with highly sug-
gestive testimonies from individuals we may
have helped. (M)
However, he conceded that the emerging focus
on accountability for achievement would prove
more problematic for this culture if senior manage-
ment continued to prioritise its importance within
Amnesty.
Attempts to measure e?ectiveness, fuelled by an
eagerness to demonstrate cause and e?ect, also
meant that so-called ‘failed’ campaigns could take
prominence with crucial work like lobbying being
ignored. This could lead to Amnesty Ireland being
judged in a manner that might actually ‘‘undermine
people’s faith in the work of the organisation” (S)
and potentially challenge Amnesty’s ability to
‘‘embrace . . . failure as an opportunity for learn-
ing” (Roberts, 1991, p. 366; Slim, 2002) contrary
to its widely expressed desire to become a learning
organisation (Amnesty International, 2004).
You’re open to challenge . . . and you’re open
to highlighting a lot of the cases that are not
818 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
successful. Our aim is to set certain issues
[on] the agenda, to achieve change in the leg-
islation on human rights practices of a par-
ticular government. Sometimes it works,
sometimes it doesn’t work but that’s not to
say that that the work isn’t important, that
you have to keep pushing in all of that. If
you were to do a substantive test and say,
‘well, X percentage of it was e?ective and
delivered a tangible result and X percentage
didn’t’, well, a lot of it is trial and error
and where is the balancing point between
those two and the need to be realistic and
to be accountable to the members and to
the public and at the same time not under-
mine people’s faith in the work of the organi-
sation? (S)
The ?nal section summarises the key ?ndings
evident in the case narrative above and analyses
their implications for the development of more
holistic accountability forms in Amnesty Ireland.
It concludes with some consideration of the poten-
tial lessons for other advocacy NGOs in the con-
text of broader trends in NGO accountability.
Discussion and conclusions
The complex nature of accountability necessi-
tates its examination in context in order to illumi-
nate the multi-faceted ways it is experienced and
enacted in speci?c organisational settings (Rob-
erts, 1991, 2001; Roberts & Scapens, 1985). Within
NGO practice generally, and in the arena of
human rights advocacy NGOs in particular,
increasing attention is being a?orded to issues of
accountability (see, for example, ActionAid Inter-
national, 2006; Ebrahim, 2005; ICHRP, 2003).
However, this has not been accompanied by exten-
sive academic examinations of the emergence of
accountability in speci?c NGO contexts. This
paper contributes towards ?lling this research
gap by analysing the development of accountabil-
ity mechanisms in the speci?c context of the Irish
section of the human rights advocacy NGO,
Amnesty International. It advances an emerging
research agenda in the social accountability litera-
ture investigating the accountability of NGOs
more generally (see, for example, Dixon et al.,
2006; O’Dwyer and Unerman, 2007). More specif-
ically, it adds in-depth empirical insights to dem-
onstrate that in a quest to capitalise on the
potential bene?ts a?orded by greater accountabil-
ity, in particular forms of performance account-
ability, (and de?ect the threats from inadequate
accountability) certain directions may be taken in
the adoption of accountability mechanisms which
may, perversely, damage an NGO’s ability to
achieve its mission. In the case of Amnesty Ire-
land, we have revealed and evaluated the factors
which may render certain accountability mecha-
nisms, with a hierarchical accountability focus,
counterproductive and damaging in this manner.
We have also revealed and analysed why Amnesty
Ireland might nevertheless adopt such potentially
inappropriate mechanisms.
The emerging focus on hierarchical account-
ability in Amnesty Ireland is driven by concerns
about Amnesty’s future viability and relevance.
It is accompanied by an apparent absence, espe-
cially among senior managers, of knowledge
about, or apparent desire to experiment with,
more holistic forms of accountability. This absence
is surprising for two reasons. Firstly, implicit pri-
ority is given to an accountability conception that
is at variance with Amnesty’s public calls for
greater holistic accountability among other organ-
isations, particularly corporations. For example,
Amnesty often argues from a moral perspective
that corporate responsibilities, and duties of
accountability allied to these responsibilities,
should be recognised in relation to all those upon
whom a corporation’s actions may impact (now
and in the future). This perspective is evident in
Amnesty’s recent campaigning for greater corpo-
rate accountability, especially through its work in
the United Kingdom based CORE Coalition push-
ing for regulations on corporate social and envi-
ronmental accountability (CORE, 2005).
Secondly, experiments by other advocacy NGOs
with holistic forms of accountability (ActionAid
International, 2006; Blagescu et al., 2005; Cronin
& O’Regan, 2002; O’Dwyer, 2005) are seemingly
overlooked. These experiments have placed a key
focus on assessing wider impacts on a broad range
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 819
of constituencies, using the perspectives of these
constituencies (or their representatives), and allow-
ing these perspectives to feed into organisational
decision making.
Several possibilities for Amnesty Ireland can be
gleaned from these experiments that could guide it
towards embracing more holistic accountability
forms. For example, of particular relevance to
Amnesty, especially given its recently broadened
mandate, is ActionAid’s Accountability, Learning
and Planning (ALPS) system.
9
Like Amnesty,
ActionAid embraces a rights-based approach to
its work focused on eradicating injustice (and pov-
erty), and its four rights-based goals and six inter-
national themes for 2005–2010 resonate strongly
with many of Amnesty’s global goals,
10
hence, cer-
tain campaigning goals of both NGOs overlap.
The ALPS approach to accountability eschews
annual reports of activity and replaces them with
annual participatory reviews of various forms
which prioritise consultation with marginalised
groups, often in innovative ways that suit bene?-
ciaries through using techniques such as storytell-
ing, theatre, people’s art and song (ActionAid
International, 2006, p. 8). These accountability
mechanisms are speci?cally aimed at improving
transparency, critical learning, re?ection, and
embracing failure, issues which Amnesty prioritis-
es in its current strategic plan. Furthermore, de-
emphasising activity reports – which are central to
Amnesty International’s current annual reports –
frees up campaigning time and allows a greater
focus on broader aspects of accountability. There-
fore, ALPS reduces unnecessary bureaucracy while
prioritising core accountability to bene?ciaries. It
is also under continual revision in response to ben-
e?ciaries’ concerns. Given the centrality of learn-
ing to Amnesty’s strategic plan, some study of
this system could reap bene?ts in pursuit of this
objective. Moreover, accounts of the di?culties
ActionAid have experienced in developing this sys-
tem are available to guide organisations like
Amnesty if they wish to experiment with this
approach (see, Scott-Villiers, 2002).
While approaches such as ALPS suggest some
possibilities for holistic accountability in Amnesty
Ireland, our analysis also indicates that key obsta-
cles need to be addressed. At a minimum, senior
managers need to become more receptive to, and
knowledgeable about, broader holistic account-
ability forms.
11
This means that the overt focus
on potentially powerful external others needs to
be tempered to allow space for more sophisticated
accountability forms. Our perception from the
overall interview material is that senior managers’
perspectives seem strongly in?uenced by trends at
the international level of Amnesty as re?ected in
the latest strategic plan. Therefore, unless some
shift occurs at the international level towards
embracing greater holistic accountability then it
appears unlikely that senior managers will pro-
mote and develop this focus within the Irish sec-
tion. This international level impediment to
greater holistic accountability could equally be a
factor in other national branches of Amnesty.
However, Amnesty International’s more busi-
ness-like focus on growing market share and dem-
onstrating narrow forms of e?ectiveness (see,
Williamson, 2005a, b) make this change in focus
unlikely.
Within this context, it was middle managers
who seemed most concerned at the potentially
counterproductive nature of a trend towards forms
of hierarchical accountability through perfor-
mance metrics. Unless middle managers are
enabled to make their concerns heard in an e?ec-
9
We did not address this issue in our interviews because
before we analysed the ?nal interviews we had not appreciated
how applicable these holistic accountability mechanisms being
developed by some other NGOs could be to Amnesty. This is an
area that could form the basis for future study.
10
ActionAid International’s goals are: poor and excluded
people and communities will exercise power to secure their
rights; women and girls will gain power to secure their rights;
citizens and civil society across the world will ?ght for rights
and justice; states and their institutions will be accountable and
democratic and will promote, protect and ful?l human rights
for all. ActionAid’s six international themes encompass:
women’s rights; the right to education; the right to food; the
right to human security during con?icts and emergencies; the
right to life and dignity in the face of HIV/AIDS; and the right
to just and democratic governance (ActionAid International,
2006).
11
We should make clear that are not suggesting that holistic
accountability forms are entirely absent in Amnesty Ireland or
that senior managers reject greater holistic accountability
outright, merely that the possibilities are not being explored.
820 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
tive manner then the trend may continue
unabated.
Some hope may emerge from Amnesty Interna-
tional’s involvement in the INGO accountability
charter. This charter indicates that in balancing
the di?erent views of stakeholders, Amnesty Inter-
national will be ‘‘guided by [its] mission” (INGO,
2006, p. 2) and as part of its accountability it will
‘‘encourage input by those people who may be
directly a?ected [by its work]” (INGO, 2006, p.
4) – not just donors, members and bene?ciaries.
We argue that this should be embraced in practice.
If the prevailing motives and strategic direction for
accountability evident from our case analysis are
not broadened then Amnesty may be exposed to
risks that its concern to ‘prove’ its e?ectiveness
to ‘extraordinary members’ through external
accountability mechanisms will, in line with the
expectations of many middle managers, lead to
the adoption of speci?c accountability practices
that mute elements of its mission and re-direct its
focus away from certain potential bene?ciaries.
12
While the key ?ndings in this case are context
speci?c, it is possible to draw some lessons that
may apply to other advocacy NGOs given the
similarity in structure and focus on rights-based
advocacy of many of these organisations. First,
NGO managers need to remain attentive to their
core mission whatever pressure they may feel, real
or imagined, to narrowly account for perfor-
mance to powerful stakeholders. Satisfying per-
ceived accountability requirements of prominent
supporters/donors may not necessarily yield
improved mission achievement and could actually
lead to mission drift as has happened in some
development NGOs now competing with
Amnesty in ‘the human rights space’ (Ebrahim,
2005; Lloyd, 2005; Najam, 1996; Williamson,
2005a). While recognising that, in a context of
increased competition for funds among NGOs
whose activities are in e?ect coalescing, NGOs
might (like Amnesty) use a commitment to exter-
nal accountability as a tool to distinguish them-
selves from other NGOs, managers need to
ensure that accountability to the organisational
mission is not demoted in this competitive quest
for enhanced funding.
Second, NGO managers need to develop strat-
egies to manage the potential tensions between
rigid forms of hierarchical accountability and
mission achievement. For example, sta? need to
be able to formally discuss the perceived impacts
of accountability mechanisms with senior manag-
ers and with each other. They also need to be
given ?exibility to operate outside the bounds of
strict hierarchical accountability requirements.
Our case suggests that some managers had not
‘bought in’ to the trend and strategic-level push
towards greater hierarchical accountability.
Organisations therefore need to ensure that any
shifts in accountability such as those revealed in
this study are discussed and assessed in-depth
and not introduced on an ad hoc basis without
clear justi?cation in terms of their potential
impact on a broad range of stakeholders. This
means that internal communication on both the
reasons for and the development of new account-
ability mechanisms is crucial.
Third, if accountability mechanisms are merely
used as control and justi?cation instruments,
rather than as tools for learning or for disseminat-
ing ?ndings, then mission drift will become more
likely as NGOs become more distant from their
bene?ciaries. Advocacy NGOs therefore need to
strive to ?nd a balance between mechanisms that
ascertain and respond to the needs of powerful
patrons, and mechanisms which can more directly
support mission accomplishment by ascertaining
and responding to the needs of sta? and bene?cia-
ries – in addition to other stakeholders upon
whom the NGOs’ activities might impact. As
Ebrahim (2005) suggests, ?nding this balance will
require a reorientation toward more holistic
accountability forms which prioritise learning pro-
cesses and accountability to mission in order to
place hierarchical accountability in proper
perspective.
The Amnesty ‘setting’ examined in this paper is
somewhat unusual in the NGO context, given the
organisation does not rely on governments or
12
Senior Amnesty executives in other countries such as
Germany have reiterated these concerns when publicly stating
their fears that narrow measures of ‘e?ectiveness’ (not justice or
compassion) now look like becoming the criterion that deter-
mines what Amnesty does and where it operates (Williamson,
2005b).
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 821
funding agencies for funding and it has a clearly
de?ned and unyielding mission. In contrast, NGOs
in the development sector tend, in the main, to rely
on direct state funding from governments and
funding agencies and often have a greater face-
to-face relationship with the bene?ciaries they
aim to serve. We suggest that future research
should extend our current examination of the
emergence and e?ects of particular accountability
forms in Amnesty Ireland to these distinctive
development NGO contexts, in addition to other
advocacy NGO settings, given the often con?icting
and complex pressures they face. This would serve
to enhance and deepen our understanding of how
di?erent forms of accountability manifest them-
selves in other NGO settings. Of particular interest
would be examinations of whether, how and why
responsibilities to original visions and missions
are subsumed by dominant, externally focused,
hierarchical conceptions of accountability. Fur-
thermore, while this study has focused primarily
on managerial conceptions of accountability pro-
cesses, future research could focus on how stake-
holders of individual NGOs (both those
stakeholders impacting on and those impacted by
NGOs) perceive their accountability relationships
with these NGOs. For example, within develop-
ment NGOs, case studies examining manifesta-
tions of speci?c government, donor and
bene?ciary accountability relationships could add
signi?cantly to our understanding of how account-
ability processes operate in these increasingly in?u-
ential contexts.
This broad range of opportunities and needs for
further research is partially a function of the lim-
ited attention that issues of NGO accountability
have received in the accounting and accountability
academic literature. But, more importantly, it is
also a result of the rapidly growing in?uence of
NGOs on many aspects of the lives of growing
numbers of people throughout the world. In this
context, research providing evidence of the limita-
tions, drawbacks and bene?ts of a range of speci?c
accountability mechanisms employed by a variety
of NGOs should help to foster more e?ective
NGO accountability mechanisms aimed at
enhancing NGOs’ ability to deliver the maximum
social and environmental bene?ts.
Acknowledgements
The assistance of Caitriona Brocklebank in the
initial stages of this paper’s development is greatly
appreciated. We would also like to gratefully
acknowledge the assistance of all of the interview-
ees who contributed to this study as well as the
excellent assistance of Sara Bennett in Amnesty
Ireland. The comments of Rob Gray, Sue Gray,
Michael Grayson, Sheila Killian, John Lannon,
Carlos Larrinaga Gonzalez, Richard Laughlin,
Sue Llewellyn, Victor Maas, Hugh McBride,
Christopher Napier, David Owen, Thomas Riise
Johansen, Crawford Spence, and Sander van Tri-
est on earlier drafts were highly helpful in develop-
ing the paper. We would also like to acknowledge
the supportive comments of participants at the
2005 European Accounting Association annual
conference in Gothenburg, the 2005 Irish
Accounting and Finance Association annual con-
ference in Limerick, the 2005 Critical Perspectives
on Accounting conference in New York, and the
2005 International Congress on Social and Envi-
ronmental Accounting Research at the University
of St. Andrews as well as the comments of two
anonymous referees. Further comments were
gratefully received from participants at research
seminars in the School of Management, Royal
Holloway, University of London and the Amster-
dam Business School, University of Amsterdam.
References
ActionAid International (2006). Accountability, learning and
planning system. ActionAid International. . Accessed 31.07.07.
Amnesty International (2004). Globalizing justice, Amnesty
International strategic plan. .
Amnesty International (2007). Statute of Amnesty International.
.
Accessed 15.03.07.
Andrews, D. (2001). Amnesty soldiers on, but campaign ads
were o? target. The Irish Times.
Blagescu, M., De Las Casas, L., & Lloyd, R. (2005). Pathways
to accountability: A short guide to the GAP framework.
London: One World Trust.
822 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824
Blagescu, M., Lloyd, R., Dombrowski, K., Kadosh, R., &
Oatham, J. (2006). 2006 Global accountability report. Lon-
don: One World Trust.
Buchanan, T. (2002). The truth will set you free: The making of
Amnesty International. Journal of Contemporary History,
37(4), 575–597.
Burall, S., Neligan, C., & Kovach, H. (2003). Power without
accountability. London: One World Trust.
Candler, G. (2001). Transformations and legitimacy in non-
pro?t organisations – The case of Amnesty International
and the Brutalisation thesis. Public Organization Review, 1,
355–370.
Cooper, S. M., & Owen, D. L. (2007). Corporate social
reporting and stakeholder accountability: The missing link.
Accounting, Organizations and Society., 32(7–8), 649–667.
CORE (2005). CORE submission to UK Company Law Reform
White Paper – June 2005. . Accessed 15.03.07.
Cronin, D., & O’Regan, J. (2002). Accountability in development
aid: Meeting responsibilities, measuring performance. A
research report for Comhlamh. Dublin: Comhlamh Aid
Issues Group.
Cummins, A. (2002). Amnesty condemns government delay on
convention. The Irish Times.
Dawson, E. (1998). The relevance of social audit for Oxfam
GB. Journal of Business Ethics, 17(13), 1457–1469.
Dey, C. R. (2007). Social accounting at traidcraft plc: A
struggle for the meaning of fair trade. Accounting, Auditing
and Accountability Journal, 20(3), 423–445.
Dixon, R., Ritchie, J., & Siwale, J. (2006). Micro?nance:
Accountability from the grassroots. Accounting, Auditing
and Accountability Journal, 19(3), 405–427.
Donnellan, E. (2001). Amnesty assails ?nding on racism
adverts. The Irish Times.
Ebrahim, A. (2003a). Making sense of accountability: Concep-
tual perspectives for Northern and Southern nonpro?ts.
Nonpro?t Management and Leadership, 14(2), 191–212.
Ebrahim, A. (2003b). Accountability in practice: Mechanisms
for NGO’s. World Development, 31, 813–829.
Ebrahim, A. (2005). Accountability myopia: Losing sight of
organizational learning. Nonpro?t and Voluntary Sector
Quarterly, 34(1), 56–87.
Edwards, M., & Fowler, A. (2002a). Introduction: Changing
challenges for NGDO management. In M. Edwards & A.
Fowler (Eds.), NGO Management. London: Earthscan.
Edwards, M., & Fowler, A. (Eds.). (2002b). NGO Management.
London: Earthscan.
Edwards, M., & Hulme, D. (2002a). Beyond the magic bullet?
Lessons and conclusions. In M. Edwards & A. Fowler
(Eds.), NGO Management. London: Earthscan.
Edwards, M., & Hulme, D. (2002b). NGO performance and
accountability: Introduction and overview. In M. Edwards &
A. Fowler (Eds.), NGO Management. London: Earthscan.
Goddard, A., & Assad, M. J. (2006). Accounting and navigat-
ing legitimacy in Tanzanian NGOs. Accounting, Auditing
and Accountability Journal, 19(3), 377–404.
Gray, R. H., Bebbington, J., & Collison, D. (2006). NGOs, civil
society and accountability: Making the people accountable
to capital. Accounting, Auditing and Accountability Journal,
19(3), 319–348.
HAP International (2003). The Accountability Work Plan
Requirements and Guidelines. Geneva: HAP International.
Haughey, N. (2001). Amnesty chief defends anti-racism posters.
The Irish Times.
Haughey, N. (2002). Amnesty critical of refusal to access Irish
prisons. The Irish Times.
Holland, K. (2004). Government urged to ratify treaty on
migrants. The Irish Times.
Hopgood, S. (2006). Keepers of the ?ame: Understanding
Amnesty International. Ithaca: Cornell University Press.
Hopwood, A. G. (2005). Editorial: After 30 years. Accounting,
Organizations and Society, 30(7–8), 585–586.
Huberman, A. M., & Miles, M. B. (1994). Data management
and analysis methods. In N. K. Denzin & Y. S. Lincoln
(Eds.), Handbook of Qualitative Research. Thousand Oaks,
California: Sage.
Hudson, A. (1999). Organizing NGOs’ international advocacy:
organizational structures and organizational e?ectiveness.
Paper presented at the NGOs in a Global Future conference.
International Council on Human Rights Policy (ICHRP)
(2003). Deserving trust – Issues of accountability for human
rights NGO’s. Draft Report. www.ichrp.org/paper_?les/
119_w_01.doc. Accessed 15.03.07.
International Non-Governmental Organisations (INGO)
(2006). Accountability charter. . Accessed 15.03.07.
Jordan, L., & Van Tuijl, P. (Eds.). (2006). NGO accountability:
Politics, principles and innovations. London: Earthscan.
Keystone (2006). Learning with stakeholders: A keystone guide
to stakeholder dialogue. London: Keystone.
Labanyi, D. (2004). McDowell stands over Amnesty apology
claim. The Irish Times.
Lloyd, R. (2005). The role of NGO self regulation in increasing
stakeholder accountability. London: One World Trust.
Love, S. (2002). One of Ireland’s last acts on Security Council
must be to prevent war. The Irish Times.
Love, S. (2004a). Irish lag behind in e?ecting a human rights
culture. The Irish Times.
Love, S. (2004b). Immigration bill is outdated and defective.
The Irish Times.
Love, S. (2005). Save your anger for the perpetrators of abuse.
The Irish Times.
Najam, A. (1996). NGO accountability: a conceptual frame-
work. Development Policy Review, 14, 339–353.
Nelson, P., & Dorsey, E. (2003). At the nexus of human rights
and development: new methods and strategies of global
NGOs. World Development, 31(12), 213–226.
O’Dwyer, B. (2004). Qualitative data analysis: exposing a
process for transforming a ‘messy’ but ‘attractive’ ‘nui-
sance’. In C. Humphrey & B. Lee (Eds.), Real Life Guide to
Accounting Research: A behind the scenes view of using
qualitative research methods. Amsterdam: Elsevier.
B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824 823
O’Dwyer, B. (2005). The construction of a social account: a
case study in an overseas aid agency. Accounting, Organi-
zations and Society, 30(3), 279–296.
O’Dwyer, B. (2007). The nature of NGO accountability:
Motives, mechanisms and practice. In J. Unerman, J.
Bebbington, & B. O’Dwyer (Eds.), Sustainability accounting
and accountability. London and New York: Routledge.
O’Dwyer, B., & Unerman, J. (2007). From functional to social
accountability: Transforming the accountability relation-
ship between funders and non-governmental development
organisations. Accounting, Auditing and Accountability
Journal, 20(3), 446–471.
Patton, M. Q. (2002). Qualitative evaluation and research
methods. Beverly Hills, California: Sage.
Roberts, J. (1991). The possibilities of accountability. Account-
ing, Organizations and Society, 16(4), 355–370.
Roberts, J. (2001). Trust and control in Anglo-American
systems of corporate governance: The individualizing and
socializing e?ects of processes of accountability. Human
Relations, 54(12), 1547–1572.
Roberts, J., & Scapens, R. (1985). Accounting systems and
systems of accountability. Accounting, Organizations and
Society, 10(4), 443–456.
Ron, J., Ramos, H., & Rodgers, K. (2004). Transnational
information politics: Amnesty International’s international
reporting, 1986–2000. Working paper, Montreal: McGill
University.
Scott-Villiers, P. (2002). The struggle for organisational change:
How the action aid accountability, learning and planning
system emerged. Development in Practice, 12(3–4), 424–435.
Slim, H. (2002). By What Authority? The legitimacy and
accountability of non-governmental organisations. The Inter-
national Council on Human Rights Policy. International
Meetings on Global Trends and Human Rights – Before
and After September 11th, Geneva. www.ichrp.org./paper_-
?les/119_w_02.doc. Accessed 15.03.07.
The Economist (2007a). Many rights, some wrong. . Accessed 28.09.07.
The Economist (2007b). Stand up for your rights. . Accessed 28.09.07.
Unerman, J., & Bennett, M. (2004). Increased stakeholder
dialogue and the internet: towards greater corporate
accountability or reinforcing capitalist hegemony? Account-
ing, Organizations and Society, 29(7), 685–707.
Unerman, J., & O’Dwyer, B. (2006a). On James Bond and the
importance of NGO accountability. Accounting, Auditing
and Accountability Journal, 19(3), 305–318.
Unerman, J., & O’Dwyer, B. (2006b). Theorising accountability
for NGO advocacy. Accounting, Auditing and Accountability
Journal, 19(3), 349–376.
United Nations (2006). Debating NGO accountability. New
York and Geneva: United Nations Non-Governmental
Liaison Service.
Vakil, A. C. (1997). Confronting the classi?cation problem: A
taxonomy of NGOs. World Development, 25(12),
2057–2070.
Waters, J. (2001). Amnesty not free of racism. The Irish Times.
Waters, J. (2005). Amnesty surrenders credibility. The Irish
Times.
Williamson, H. (2005a). Under attack. Development and Coop-
eration, April. . Accessed
15.03.07.
Williamson, H. (2005b). Free thinkers. The Financial Times.
824 B. O’Dwyer, J. Unerman / Accounting, Organizations and Society 33 (2008) 801–824

doc_966817001.pdf