Information Systems

for Public Sector

Management
Working Paper Series
Paper No. 1

Information Systems
and Public Sector
Accountability

RICHARD HEEKS

July 1998

Published with the support of:

http://www.comnet.mt

ISBN: 1 902518 02 0

Published Institute for Development Policy and Management

by: University of Manchester, Precinct Centre, Manchester, M13 9GH, UK
Tel: +44-161-275-2800/2804 Fax: +44-161-273-8829
Email: [email protected] Web: http://www.man.ac.uk/idpm

View/Download from: http://www.man.ac.uk/idpm/idpm_dp.htm#isps_wp

 Table of Contents

Abstract.................................................................................... 1

1. THE CONTEXT OF PUBLIC SECTOR REFORM ......................... 2

1a. Components of Public Sector Reform.................................. 5

2. DEFINING ACCOUNTABILITY.................................................... 6

3. ACCOUNTABILITY OUTCOMES ASSOCIATED WITH

NEW COMPUTERISED INFORMATION SYSTEMS ....................... 8
3a. Essential Support for Accountability.................................... 8
3b. Support for Less than Accountability................................... 9
3c. Alteration of the Balance of Accountabilities ....................... 10
3d. Undermining of Accountability............................................ 14

4. FACTORS DETERMINING ACCOUNTABILITY

OUTCOMES ....................................................................................... 17
4a. Information Technology ...................................................... 17
4b. Wider Systems Design......................................................... 18
4c. The Process of Information System Design.......................... 19
4d. Organisational and Environmental Factors........................... 20

5. CONCLUSIONS.............................................................................. 21
5a. IT, IS and Accountability .................................................... 21
5b. Achieving Accountability .................................................... 22
5c. Implications for Information Systems Design....................... 23

REFERENCES .................................................................................... 25
Information Systems and Public Sector
Accountability

Richard Heeks
IDPM, University of Manchester, UK
[email protected]

Abstract

Increased accountability is a key component of public sector reform. This paper

investigates the relationship between accountability and both information technology
and information systems. It concludes that IT has had a very mixed impact,
supporting accountability in some cases, but also skewing or undermining it in others.
Information systems - computerised or not - are an essential part of public sector
accountability. However, new IS have a flexible impact, with accountability outcomes
being determined mainly by management decisions and by wider organisational and
environmental factors. As such, the broader context must be conducive to
accountability before accountability IS initiatives will work.
 1. THE CONTEXT OF PUBLIC SECTOR REFORM

Public sector reform is, if generally defined, change within public sector organisations
(PSOs) that seeks to improve their performance. As such, public sector reform has
been an ongoing process since the inception of institutions that we might now label
'public sector'.

However, public sector reform is typically defined more narrowly. It is seen as a

process that grew up particularly from the 1970s for which three overarching (and
interlinked) causes may be identified.

Crisis in the public sector

If all had been well in the public sector, no consistent trend, let alone ideology, of
reform would have emerged. However, a perception of problems with the public
sector, even of crisis in some countries, emerged during the 1970s. The perceived
problems were focused on:

• Inputs. In a number of countries, the public sector was seen to require

unsustainably large and/or unsustainably increasing public expenditure.

• Processes. There was concern about examples of waste, delay, mismanagement

and corruption within the public sector, all of which contributed to inefficiency in
the conversion of public expenditure into public services. In particular, public
servants were seen as sometimes making decisions in their own interests rather
than in the interests of the public. That they were able to do so was seen as a twin
failure. First, of centralisation, which made decision makers too remote from the
locus of decision information and action. Second, of unaccountability, which made
decision makers too remote from those outside the organisation who were affected
by their decisions.

• Outputs. Finally, there was a perceived problem with outputs. Concerns were
widespread in a number of countries that the public sector was not delivering what
it should, from adequate defence and control of crime through support for
agriculture and industry to health, education, housing, social welfare and a hundred
other responsibilities.

2
The sense of difficulties came to cover both what the public sector was doing (the
public sector's role) and also how it was doing it (public sector organisation and
management).

A renewed ideology

If there had been no ideological peg on which to hang many of the trends of change
within the public sector, reform measures would have been less clearly recognised and
probably less strongly promoted. Such a peg emerged slowly after the Second World
War and with gathering pace from the 1970s in the form of 'neo-liberalism'. Neo-
liberalism is a resurgence in the ideas of liberalism that can be traced back to John
Locke and Adam Smith in, respectively, the seventeenth and eighteenth centuries. It
provides a substantial theoretical framework that can be used to justify a set of public
sector reforms.

In its crudest form - 'market good, government bad' - neo-liberal thinking emphasises
what it sees as the economic efficiency of markets and the forces of competition and
individual decisions. It also emphasises what it sees as the inefficiency of governments
and the forces of collective, planned intervention. Three particulars flow from this
viewpoint:
• that, wherever possible, there should be a 'rolling back of the state'; in other words,
the replacement of the state with privately-owned institutions
• that the main justification for the continued existence of the state is its role in
helping markets to function more efficiently
• that, where state institutions remain, they should, wherever possible, be opened up
to true or quasi-market forces of competition.
Neo-liberalism therefore had something to say not just about the role of the public
sector, but also about the way in which it might be organised and managed.

Political will and power

A sense of crisis and an ideology of reform are necessary, but not sufficient conditions
for reform. There must also be a third element: that of the political will and power to
enact reform. Three main components of the political economy in most countries can
be identified that influence this.
• The populus at large has often borne the brunt of public sector crisis, has typically
longed for reform, but has had only limited political capacity to have those reforms
enacted.

• Politicians and public servants have often been divided, with conflict between
those supporting and those resisting reform.

• Local and global capital has sometimes been divided but has more generally
sought reform in the belief that this will reduce business costs and increase
transaction speed.

In some countries, particularly the developing and transitional economy countries,

there has been a fourth component of political economy:

• International organisations have often been a powerful - sometimes the most

powerful - driving force behind public sector reform. These agencies have the
political power to pursue reform because many countries, struggling with both
international trade and domestic spending deficits, have had to request external
sources of financial assistance. Since the 1970s, such requests have often been met
with 'conditionalities'. These are requirements by international finance
organisations for changes in the recipient country's economic and institutional
structures. In return for agreement to begin implementation of such changes,
financial assistance is provided.

These agencies have the political will to pursue reform because they are largely
driven by a neo-liberal agenda. Reform measures concerned with government
policy include: removing trade barriers, encouraging foreign investment, floating
the local currency, privatising government-owned enterprises, and so on. Other
measures - often falling under the wider heading of 'good governance' - are
basically those of public sector reform outlined below in section 1a, which aim to
improve the internal workings of the public sector.

The political economy of every country is different. However, one stereotypical

outcome of these various forces has been a process of reform that is driven largely
from outside the public sector, whilst being resisted by at least some portion of public
servants.

4
 1a. Components of Public Sector Reform

Where public sector crisis prompted the call "Something should be done", neo-liberal
ideology provided the response "Something can be done" and, in some situations,
political driving forces demanded that "Something will be done". This "something"
has come in the form of various measures that fall collectively under the heading of
'public sector reform'.

There is no consistent menu of elements that make up a programme of public sector

reform. Typical components, however, include:
• increased efficiency: improving the input:output ratio within the public sector; the
rationale of such reforms is to address the large size of public sector expenditure
and/or the inefficiency of many of its processes
• decentralisation: the transfer of decision making to lower, more localised levels of
the public sector; the rationale of such reforms is to reduce the costs of centralised
decision making, and to create more flexible and responsive decision making
• increased accountability: making public sector staff more accountable for their
decisions and actions; the rationale of such reforms is to increase the pressure on
staff to perform well, to make them more responsive to recipient groups, and to
reduce corrupt practices
• improved resource management: increasing the effective use of human, financial
and other resources; the rationale of such reforms is clear from their definition
• marketisation: increasing the use of market forces within the workings of the
public sector; the rationale of such reforms is that market relations will drive costs
down and increase efficiency and/or effectiveness.

In this paper, we will look at the third element: accountability.

 2. DEFINING ACCOUNTABILITY

The interest in accountability within public sector reform was defined above as a desire
to make public sector staff more accountable for their decisions and actions. In more
detail, this means:
• that some set of recipients receives information about the outcomes of decisions
made by identified individuals who are source decision makers
• that those sources can be made to explain their decisions
• and that some sanctions can be imposed if the explanations are unsatisfactory.

But to whom should public servants be accountable? There is a broad set of

accountabilities in the public sector, including:

• Managerial accountability: to senior managers within the organisation. For

example, public servants may be held accountable by their immediate boss for their
attendance record.

• Political accountability: to those institutions that provide the political legitimacy

of the organisation. For example, PSO senior managers may be held accountable
by politicians for the overall achievements of their organisation, or for particular
projects that have been implemented.

• Financial accountability: to those institutions that provide the financing for the
organisation. For example, project managers may be held accountable by a
funding organisation for the expenditure on their project.

• Public accountability: to citizens outside the organisation. For example, a

Minister or Permanent Secretary may be held accountable by the public for corrupt
activities within their Ministry. This is a combination of:
- ultimate political accountability, since public institutions are supposed to derive
ultimate legitimacy from the citizenry, and
- client accountability, since citizens are normally the intended recipients of the
services a PSO provides.

Other accountabilities (Lawton & Rose 1991) include the professional accountability
of staff to their professional peer group (e.g. social workers to other social workers);
and legal accountability to the judiciary. All of these accountabilities are illustrated in
figure 1.

6
 Figure 1: Accountabilities in the Public Sector

External
Accountabilities

Internal
Accountabilities
Legislature/
Senior Judiciary
Politicians
Managers

Managerial Legal
Political

Public
Servant
Professional
Financial
Professional
Peers
Public

Finance Providers
(Government,
Donors, Etc.)

Citizens/
Clients

Within public sector organisations, accountability is often that of the individual for
decisions made. Where accountability runs outside the organisation to other
recipients, accountability is often aggregated to the level of the whole organisation. In
other words, the whole organisation is somehow seen as accountable for the actions of
its staff. Even in this case, individual senior staff may be held accountable by the
external recipients. Those senior staff, in turn, may hold more junior individuals
responsible within the organisation for externally-reported decisions and outcomes.
 3. ACCOUNTABILITY OUTCOMES ASSOCIATED WITH NEW
COMPUTERISED INFORMATION SYSTEMS

To what extent can the use of new computerised information systems support these
various accountabilities? Evidence is drawn here from a number of cases, from which
emerge four principal accountability impacts associated with information systems,
particularly computerised information systems:
• essential support for accountability
• support for less than accountability
• alteration of the balance of accountabilities, and
• undermining of accountability.

These will now be investigated in turn. In each case, an attempt will be made to
disentangle the broader role of information systems and the narrower role of
information technology.

3a. Essential Support for Accountability

The creation of new information systems is an essential component in the creation of

accountability. When a decision is taken, information about that decision and its
outcomes must flow to all those to whom the decision maker is accountable. Without
such an information flow and without the information system to carry that flow, there
can be no accountability because there can be no knowledge of the decision.

"Essential to effective accountability for results (as

well as good management) is credible, timely and relevant
information on the performance of services and programs."
(Treasury Board Secretariat 1994:53)

Accountability information systems are basically management information systems.

They provide for the monitoring of decision performance, and are intended to assist in
the control of that performance. As well as enabling new accountabilities, they can
also improve existing ones through provision of more timely, consistent and formal
information about performance.

In some cases, use of information technology forms an almost essential element of

these new information systems. In one Asian railways information system, for

8
instance, computerisation provided automated monitoring and comparison functions
that would have been very hard to achieve using a non-computerised system (Heeks
1995).

However, computers are by no means always needed. One system of bringing local
government officials in India to account used paper copies of financial statements
pasted up in a public place, with no computers involved (Bajaj & Sharma 1995).

3b. Support for Less than Accountability

Three possible types of accountability information system can exist, as illustrated in

figure 2.

Figure 2: Types of Accountability Information System

Inputs Process Outputs

Control Comparison Monitoring

Openness Reporting
Accountability
IS IS
IS

Those information systems with just a monitoring mechanism merely support

reporting: recipients have information on decisions and outcomes but cannot judge
them. Those with monitoring and comparison mechanisms merely support openness:
recipients can judge whether decisions and outcomes meet acceptable performance
standards but can do nothing with this information. Only those information systems
that provide monitoring, comparison and control mechanisms can be said to truly
support accountability: allowing recipients to take actions that affect the source
decision maker.
For example, in India, a new (paper-based) public information system was created with
the intention of improving accountability on rural development projects (Roy 1996).
However, it only supported openness: citizens were able to monitor reported
expenditure on local projects and compare it with actual project outcomes, thus
identifying discrepancies in which public servants had embezzled funds. What the
system did not do is provide a mechanism for actually holding decision makers
accountable.

Information technology has tended to be used mainly to help create reporting or, to a
lesser extent, openness information systems. It is rarely possible for IT to be applied
in a full accountability system that automates control.

3c. Alteration of the Balance of Accountabilities

New information systems can expand the scope of accountability by delivering

accountability information to new recipients. However, the effect in practice has been
uneven. Those groups or institutions with the financial and political power to impose
new information systems on public sector organisations are those to whom greater
accountability has accrued.

This result has been seen particularly in the case of new information systems that are
information technology-based and of powerful groups that are external to the PSO,
such as:

• Central government. As central governments seek to enact reform measures, they

seek greater control over, and accountability of, other parts of government. They
have often used new computerised information systems in this effort. In Brazil, for
example, the Ministry of Economics and Finance introduced a centralised,
computerised financial administration information system (Reinhard & Zwicker
1996). This reported, and sanctioned payment for, the financial transactions of
most government agencies. The effect was to increase the accountability of those
agencies to central government.

10
• Donors. Funding agencies have been particularly active in formalising
accountability relationships through the introduction of new computerised
information systems:

"Many donors now increasingly look for mechanisms for

effective project management, particularly financial
management to be available. This is seen as providing
the necessary feedback and control information to ensure
managerial accountability over project implementation. …
That donors require this type of information from
projects has led to their undertaking a programme of
micro[computer] introduction to facilitate it." (Charlton
1988:8)

Evidence indicates that such computerised systems will tend to drive out other
information systems and channels (Burns 1984, Avgerou & Mulira 1996, Sahay 1996).
This occurs for a number of reasons:

• Attention: human attention capacity is limited. Where a recipient attends to formal

computing information, s/he must necessarily remove attention from other
information systems and channels. (This applies to any type of information system,
IT-based or not, and means that some of the shifts described below can occur with
or without computerisation.)

• Information channel speed: recipients select information channels mainly on the

basis of speed and convenience rather than quality of information (Hardy 1982).
Once installed, IT-based accountability systems will have a tendency to be relied
on more and more because of this, again driving out other information systems and
channels.

• Personal preference: some recipients prefer to receive information from inanimate

objects. They will tend to focus increasingly on computerised information systems
to the exclusion of other sources.

• The mask of objectivity and data quality. Insofar as computer-mediated

communication channels are given greater (and often undue) credibility than other
channels, the former will tend to drive out the latter. This issue is discussed again
below.
The result is often a shift in the balance of information flows and accountability from
citizens to central government and/or donors. This, in turn, leads PSOs to become
more responsive to the demands of central government and/or donors, and less
responsive to the needs of client groups.

There may also be a shift in focus from internal management to external groups. This
has long been a problem with the external reporting of information in the public sector:
that it consumes a large amount of internal resource, yet tends to undermine internal
information systems and internal management processes (Rathore 1977, Lucas 1994).
Internal information needs may be forgotten or, at best, be satisfied only where they
overlap with the provisions of these external information systems.

This overlap between external information provision and internal information needs
may be limited. Public sector systems for external reporting of accounts, for example,
generally have to adhere to specified strict accounting principles, many of which are
not relevant to internal needs. More than this, as noted below, there are frequent
pressures to falsify externally-reported information.

Some public sector organisations have been driven to maintain two separate
information systems: one for external consumption, and the real figures for internal
use. Examples have been presented to the author from the public sector in one East
African and one South Asian nation. In the former case, senior civil servants used one
set of information for their own management purposes. A different set of information
- mainly personnel and financial - was used to provide donors with a 'rosier-than-
reality' picture of the progress of public sector reform.

In other cases, the situation might be reversed. The imposition of information systems
may derive internally where, for example, a powerful senior manager pushes through
their introduction. Again, there is likely be a growing imbalance of accountability in
these situations, in this case away from external groups and institutions towards
internal senior management.

Three scenarios have therefore been described that result from the introduction of new
information systems, each of which leads to a shift in accountabilities:
• between external groups
• from internal management to particular external groups
• from external groups to internal management.

12
In all three cases, there is a process of polarisation in which the public at large - as
both citizens and clients - are the 'accountability losers'.

As noted, these shifts can take place without the introduction of information
technology. However, many of the new information systems are computer-based, and
computerisation seems to give a boost to the imbalances of accountability that
develop. This is ironic since much has been made of the potential for IT to create a
greater direct accountability of PSOs to the public. There are, for instance, systems
that provide public sector information via the Internet in many countries. Yet the
information provided in such systems often consists of details about how to apply for
public services; regulations governing business; and rose-tinted press releases
('cyberganda'). The presence of all this information helps divert attention from the fact
that only rarely do such systems provide the type of information that allows public
servants to be held to account.

Where accountability information is provided online, it can still reinforce inequities of

accountability because of the accessibility barriers that computerised information
systems raise (Cottrill 1995, Stoll 1995). Recipients must have access to the
technology, and must have the skills and confidence to access and interpret the
information provided. Central government and donor organisations have the resources
to ensure they can overcome these barriers. This is the case for only a minority of
public citizens in all countries.

When accountability information is available on paper, accessibility barriers are lower.

Even here, some barriers remain, for recipients must generally be word-literate and
statistically-literate to read and interpret the output. Citizens are still likely to be at a
greater disadvantage than central government or donor organisations, though the
'disadvantage gap' should be less than with computerised systems.

This imbalance against the public is unlikely to be redressed unless the mass media
plays a role (Roy 1996). This role must be that of receiving, interpreting and
publicising accountability information, and then of pressurising public officials on
behalf of citizens.
 3d. Undermining of Accountability

From the previous sub-section, we may conclude that information systems in general
can skew the balance of public sector accountabilities. Computerised information
systems seem especially likely to contribute to this skew. Continuing the specific
focus on information technology, potential impacts are discussed below which may
sometimes undermine the whole process of accountability.

Loss of paper record

Computerisation may be associated with a loss of paper records, and this can have a
significant impact on accountability. As Cain (1996) points out, computerised records
currently provide a far poorer basis for accountability than paper-based records
because of their intangibility and malleability. Unless there is a concerted and
continuing effort to maintain paper records, accountability can be undermined with the
advent of computerisation.

The mask of data quality

Where they know they are being judged by others on the basis of the data provided,
sources will tend to provide inaccurate data (Whetton 1986, Chambers 1996,
Chepaitis 1996). The expectation is therefore that large amounts of the information
provided by accountability information systems cannot be relied upon.

The problem with computerisation is that the introduction of IT per se has little effect
on data quality. Worse, the - often inaccurate - data within computerised
accountability systems is given credibility because of the perceived objectivity of
computers. Computerisation can lead recipients to believe in inaccurate accountability
information, thus undermining the process of accountability.

Even if the data is accurate, there may be problems because of the way in which
computerised systems tend to drive out other information channels. Where
accountability IS are computer-based, recipients may increasingly take the
accountability IS data at face value and may base accountability decisions on it alone.
Where recipients ignore other information channels and fail to cross-check via informal
consultations, the validity of the accountability process is likely to suffer.

14
Increased opportunities for corruption

In an African public works department, computerisation of the payroll system allowed

a computer operator to enter the details of 'ghost workers', whose wages he was able
to collect (Heeks 1995). Computerisation undermined accountability in this case by
opening access to corruption opportunities for computing staff, whilst reducing the
ability of non IT-literate managers to monitor and control the payroll system.

Loss of human input

Separate from the issue of the accountability information systems themselves is a

problem raised by the introduction of other computerised information systems into
public sector organisations. Such systems automate human labour and thus reduce
human inputs to the decision-making process, sometimes to the extent of removing
them altogether.

This leads to a confusion and undermining of accountability, because it produces the

question: "Who is to be held accountable when computerised information systems are
involved in decisions and subsequent actions that are judged to be sub-standard?".
This question keeps thousands of lawyers busy worldwide in liability lawsuits.

Although one might look forward to the day when PCs are locked up in jail for their
misdemeanours, information technology cannot be held accountable for its actions.
Mistakes are frequently and conveniently blamed on 'computer errors' or 'software
bugs' but, in law and in current reality, this is not reasonable. Such mistakes can
almost always be traced back to a human root.

The problem for accountability, to the delight of the lawyers, is the difficulty of which
human root to trace it back to. A faulty decision outcome arising from use of a typical
management support system (MSS) could derive from many sources, including:
• MSS development problems, such as poor analysis or poor specification or poor
design of the system
• MSS implementation problems, such as the introduction of a software bug or
inadequate training
• sub-standard manufacture of the MSS hardware, such as the installation of a faulty
chip
• MSS operational problems, such as mistyping or poor maintenance or virus
infection
• MSS use problems, such as human error by the manager making use of the system
output.

Each source in itself can be many-layered. For example, if the software system has a
bug that leads to damaging decisions, who is to be held accountable: the programmer
who put in the bug; the system tester who failed to spot it; the system designer who
chose that particular programming language and complex software design; the project
manager overseeing the whole process; the PSO manager who commissioned the
system; etc.?

The use of computerised information systems in decision making can therefore cloud
more than clarify accountability.

16
4. FACTORS DETERMINING ACCOUNTABILITY OUTCOMES

Information systems can support or alter or undermine accountability. How can we

explain this variety of different impacts? The first step is to understand both
information technology and information systems as part of a broader context.
Drawing from innumerable other works on information systems (e.g. Kling 1987,
Walsham 1993, Bell & Wood-Harper 1998), a summary framework of IT and IS
context is presented in figure 3.

Figure 3: Information Systems as Social Systems

Environment

Political Economic
Organisation

Strategies Structure
Information System
People Processes
Technology

Information
Hardware Software
Data Knowledge

Telecommunications

Management
Politics Culture
Technical
Institutional

Socio-cultural

Each one of this diagram's factors, alone or in combination, can be seen to have an
effect on accountability outcomes. Here, the factors will be discussed under four
headings.

4a. Information Technology

Information technology generally provides a reporting of information that transmits

information from the source decision maker to the recipient. At best, it can provide an
openness that compares this information to pre-set norms and alerts recipients when
there is some deviation. It cannot create complete accountability because it has no
innate accountability property. Thus, IT is not always essential in accountability
information systems. Indeed, as described above, IT seems to skew or undermine
accountability more easily than increasing it.

4b. Wider Systems Design

If information technology alone cannot create accountability, then simultaneously-

introduced, wider systems components are required if accountability is to be achieved.
These components - people, management and processes - are part of the overall
information system, but they are also related to wider organisational systems
components such as strategies and structures. They must exist around both the
accountability source and the recipient:

Source factors

• People. Over and above most other factors, there must be a motivation of public
servants to act accountably. This, in turn, depends on many other factors such as
current rewards within the job for performing well; the likelihood of detection and
punishment for wrong-doing; and the organisational culture's viewpoint on
accountability.

• Management. Within public sector organisations, there must be a system of

managerial responsibility that informs decision makers of their responsibilities and
identifies those who are responsible. One common fudge in accountability
information systems is the identification of a shortcoming but a failure to identify
who specifically is responsible for that shortcoming. This is corporate
accountability without individual accountability.

If the accountability information system is to move beyond just reporting, some

type of performance standard must be set. Reported information can then be
compared against this standard to determine what is and is not satisfactory. Both
source and recipient should be aware of these measures.

• Processes. The processes must exist to gather accountability information and

deliver it to relevant recipients.

18
Recipient factors

• People. Those wishing to hold public servants accountable must have the skills to
access information systems and interpret results. They must also be motivated to
hold public servants accountable. This last point cannot always be assumed: some
public information systems in the West are marked less by the enthusiasm of a
minority of citizens than by the apathy and disinterest of the majority.

• Management. Those wishing to hold public servants accountable must have the
authority to demand explanations from public servants, and to affect those public
servants with their subsequent actions. Appropriate structures must therefore be in
place to permit this. Taking a UK example, one of the aims of the Citizen's
Charter - besides publicising information about performance - has been to create an
easy-to-use complaints and feedback system.

• Processes. Comparison and control processes must exist that allow recipients to
interpret accountability information in the light of performance standards; to
communicate with public servants in order to supplement the formal accountability
information with informal contextual information; and then to decide upon and
implement appropriate control actions.

Returning to an earlier example, the rural poor in the state of Rajasthan in India
undertook an accountability action (Roy 1996). Here, the people components were
definitely in place as regards the recipients, who had a deep desire to hold public
servants accountable for money that was being corruptly diverted from development
projects. However, the management authority, structures and processes to then call
the public servants to account and affect them with actions were not present. This is
therefore a case of openness without accountability: new information flows made
citizens aware of wrong-doing by public servants, but they had little ability to interact
with the public servants and little power to correct those wrong-doings.

4c. The Process of Information System Design

The method of design strongly shapes the design and operation of the information
system it produces (Bhatnagar 1986, Braa 1996). The designers of an accountability
information systems must therefore themselves be accountable to the intended
recipients of accountability information.
In some cases, this is so. Where central government or donor organisations introduce
an accountability information system, for example, they often fund it and 'call the shots'
as far as system design is concerned.

The same is less likely to be true of managerial or public accountability systems. Here,
the systems designers may have little meaningful interaction with the intended
recipients. Instead, they design the information system to the requirements they
believe the recipients have, or to the requirements of the source decision makers. In
either case, the resulting information system may not support accountability.

4d. Organisational and Environmental Factors

Given the flexibility of information technology and information systems, accountability

impacts are strongly determined by the management decisions of those creating the
accountability information systems. In turn, these decisions will themselves be
affected by a range of wider factors.

Taking the Indian case (Roy 1996) once again, accountability information systems
were largely shaped by an organisational politics of self-interest and a culture of
corruption prior to the actions of a non-government organisation (the MKSS:
Mazdoor Kisan Shakti Sangathan) in pressurising government for more access to
information. In theory, before the MKSS' actions, there were accountability systems
passing information up to higher levels within the public sector organisations involved
and to external, central government institutions. In practice, the politics and culture
ensured there was no accountability because the information transmitted was largely
false and because higher levels were not motivated to try to hold development staff
accountable.

The actions of institutions in this case - the MKSS on behalf of the rural poor of
Rajasthan and, probably much more importantly, the mass media that picked up the
story and publicised it - changed the external environment, and led to changes at the
level of the state government. However, this did little to effect the ongoing politics
and culture. Overall, then, negligible progress was actually made in holding public
servants accountable.

20
 5. CONCLUSIONS

5a. IT, IS and Accountability

Clearly computerisation is no panacea for achieving accountability. Information

technology has no innate impact on accountability and, whilst it may support some
accountability information systems, it is not typically essential. Where present,
computers may skew or undermine accountability as much as support it.

Information technology may not be essential to accountability, but the same is not true
of information and information systems. Information flows (hence, information
systems) enable accountability and are essential to it. Nevertheless, like IT,
information systems are flexible and have no pre-determined impact on accountability.

In some cases, new information systems have been associated with positive
accountability outcomes for certain stakeholders. In other cases, though, even when a
working information system is produced, there may be components of IS failure, such
as:
• goals unattained for particular stakeholders: this will occur, for example, if the
new IS only assists reporting rather than accountability, or if the new IS
undermines accountability by giving credibility to inaccurate data
• undesirable outcomes for particular stakeholders: this will occur, for example, if
the introduction of a donor-oriented IS leads to a reduction in accountability to the
public.
In all, we may conclude that information systems, particularly their technology
component, have had only a limited positive impact on accountability initiatives in the
public sector.

Information systems per se do not cause or even significantly drive accountability.

Instead, the key to accountability outcomes lies within the shell of surrounding factors
illustrated in figure 3. Within this shell, it is management decisions about the
information systems and their design processes that are fundamental. These decisions,
in turn, are affected by factors such as organisational politics and culture, and the
external environment.

Thus, not surprisingly, accountability outcomes associated with IT are not equal
across all types of accountability but have tended to follow existing distributions of
power. As a generalisation, computerised accountability IS initiatives in the public
sector appear to have supported:
• financial accountability to some extent
• managerial accountability to a limited extent
• political accountability to a very limited extent, and
• public accountability hardly at all.
This hierarchy typically reflects the hierarchy of political will and power identified in
section 1.

5b. Achieving Accountability

For any type of accountability to be achieved, there must a fit between all the factors
identified in section 4, including:
• the process of information systems design
• the technology; the skills and processes of the information system
• personal motivations of the stakeholders
• management strategies and values
• organisational structures, politics and culture.

In certain public sector situations - currently characterised by limited accountability

and even corruption - this fit will be hard to achieve because it requires such a
combination of changes. This includes changes not just in the information systems
architecture but also, quite probably, in the way that information systems are designed
and implemented. In addition, there will be the wider linked changes that are required
in organisational political and bureaucratic structures, strategies, culture and so on.

Within each of these changes, there will be inertia at work that will tend to reinforce
the status quo, such as:
• a technological inertia because of the high financial and skills costs of introducing
IT or of changing the current computing architecture
• a management inertia because managers are used to making work decisions,
including those on information systems, in a particular way
• a political inertia because those involved will not wish to give up their information
and power by becoming more accountable.

It is possible to overcome such inertia, but only where change is driven from an
environmental and organisational level. In other words, referring back to figure 3,

22
change will be driven from the outer shells inward far more easily than it can be driven
from the inner shells outward.

In practice, this means that successful accountability initiatives will require consensus
amongst the key stakeholders about the desirability of accountability. Such consensus
may derive from factors within the organisation (such as self-interest, and reward and
punishment systems) or from outside the organisation (such as pressure from
politicians, international organisations, or the public via the mass media). Where such
consensus is absent, the risk of failure is high for any accountability IS project.

5c. Implications for Information Systems Design

The role of the IS design method in determining accountability outcomes has been
described above. In this final section, though, the focus will be on stakeholder
perceptions of likely accountability outcomes, and the impact of those perceptions on
IS design.

Although the impact may be limited in practice, public servants often perceive that
new information systems - especially computerised information systems - are going to
have a have significant effect of increasing accountability (and, hence, reducing
corruption). These perceptions will feed into the process of planning any new
information system. The effect of these perceptions is likely to be particularly marked
where a new computer system is being introduced in the presence of practices that are
currently unaccountable.

For instance, in one Asian government's pensions office, studied by the author,
computerisation was roundly resisted. Many factors were at first seen to underlie the
resistance including:
• fears of loss of jobs
• fears that staff would not have the necessary skills
• health and safety concerns.

What emerged during investigation, however, was that the main fear lay around the
issue of corrupt incomes. Pensions staff had the power to deny claimants access to
pension payments or to provide claimants with access to certain types of higher-
income pension. The staff were using this power to extract bribes from pensionable
claimants. They feared that computerisation would create accountability in the
pension payment process and would therefore remove this power. Hence, their true
reason for resistance.

Where resistance and lack of accountability are linked, there are three possible
reactions:
• if the new information system will not have an effect, make this (subtly) clear
• if the new information system will increase accountability and the stakeholders are
not that powerful, then 'tough it out', i.e. push on in the likelihood that resistance
can be overcome
• if the new information system will increase accountability and the stakeholders are
powerful, change the IS design plans so that the key unaccountable processes are
not computerised or are not exposed to monitoring by the computerised
information system.
'Toughing it out' can also be tried in the last case, and this would seem to be the
morally-correct route to take. However, it will greatly increase resistance to the new
information system and the risks of IS failure.

Whatever the reaction, it is clear that the link between new information systems and
accountability will have to be recognised in the planning of some information systems.
This link must also be teased out as a component of resistance to computerisation.

24
 REFERENCES

Avgerou, C. & Mulira, N. "A study of a university admission system in Uganda," in

Information Technology, Development and Policy, E.M. Roche & M.J. Blaine (eds),
Avebury, Aldershot, 1996, pp. 229-44

Bajaj, J.L. & Sharma, R. "Improving government delivery systems," Economic and
Political Weekly, 27 May 1995, pp. M73-80

Bell, S. & Wood-Harper, T. Rapid Information Systems Development (2nd edn.).

McGraw-Hill, London, 1998

Bhatnagar, S.C. Decision Support for District Administration: An Experiment in

Surendranagar. Indian Institute of Administration-Ahmedabad, mimeo, 1986

Braa, J. "Decentralisation, primary health care and information technology in

developing countries," in Global Information Technology and Socio-Economic
Development, M. Odedra-Straub (ed.), Ivy League Publishing, Nashua, NH, 1996, pp.
130-142

Burns, A. "Information technology - for better or worse?," in New Information

Technology, A. Burns (ed.), Ellis Horwood, Chichester, UK, 1984, pp. 187-215

Cain, P. "Making the transition to the electronic age," Information Technology for
Development, 7(4), 1996, pp. 159-67

Chambers, R. "The self-deceiving state," IDS Bulletin, 23(4), 1992, pp. 31-42

Charlton, J. "Dynamics of microcomputer policy development in African public

administration," Paper presented at the 1988 UK Development Studies Association
annual conference, University of Birmingham, Sept 6-9 1988

Chepaitis, E. "The problem of data quality in a developing economy: Russia in the

1990s," in Global Information Technology and Systems Management, P.C. Palvia et
al. (eds.), Ivy League Publishing, Nashua, NH, 1996, pp. 104-122

Cottrill, K. "Information ghetto" Computing, 6 April 1995, p. 25

Hardy, A.P. "The selection of channels when seeking information: cost/benefit vs. least
effort," Information Processing and Management, 18(6), 1982, pp. 289-93

Heeks, R.B. "Computerising corruption in developing countries," Information

Technology in Developing Countries, 5(2), 1995, pp. 2-5

Kling, R. "Defining the boundaries of computing across complex organizations", in

Critical Issues in Information Systems Research, R.J. Boland & R.A. Hirschheim
(eds.), John Wiley, Chichester, UK, 1987, pp. 307-362

Lawton, A. & Rose, A. Organization and Management in the Public Sector. Pitman,
London, 1991

Lucas, H. "New information technology - old information habits?," IDS Bulletin,

25(2), 1994, pp. 94-99

Rathore, R.S. "MIS in government environment," in Management Information

Systems, N.K. Kulshreshta (ed.), SIPA, Jaipur, 1977, pp. 32-41

Reinhard, N. & Zwicker, R. "Managing public information systems development," in

Information Technology for Management, Turban, E. et al. (eds.), John Wiley, New
York, 1996, pp. 798-801

Roy, B. "The right to information," Devel-L list ([email protected]), 25 April

1996

Sahay, S. Decision Support Systems. Personal communication with author, 1996

Stoll, C. Silicon Snake Oil. Macmillan, London, 1995

Treasury Board Secretariat. Working Together for Better Public Services. Treasury
Board Secretariat, Government of Canada, Ottawa, 1994

Walsham, G. Interpreting Information Systems in Organizations. John Wiley,

Chichester, UK, 1993

26
Whetton, J. "The politics of information: the case of an urban renewal programme,"
Paper presented at the 1988 UK Development Studies Association annual conference,
University of Birmingham, Sept 6-9 1988

© Richard Heeks, April 1998