Adler - 2003 - A Socio-Legal Approach to Administrative Justice

Adler ADMINISTRATIVE JUSTICE 323

© Blackwell Publishing Ltd. 2003

A Socio-Legal Approach toAdministrative Justice*

MICHAEL ADLER

The first part of this article outlines two complementary approaches to enhancingadministrative justice. Internal mechanisms, which can be put into place by govern-ment departments and public bodies themselves, are contrasted with externalmechanisms, which result in the imposition on government departments and publicbodies of principles enunciated by courts, tribunals, and ombudsmen. Lawyers areall too familiar with the external approach but tend to be much less familiar withthe internal approach. The article seeks to redress this imbalance. It emphasizesthe importance of the internal approach, not as an alternative but, rather, as acomplement to the external approach and develops a framework for analyzingadministrative justice in terms of “trade-offs” between different normative modelsof administrative decision-making. The second part of the article demonstrateshow this approach to the study of administrative justice has informed research onthe impact of computerization on social security in the United Kingdom; on deci-sion making in the Scottish prison system; on the assessment of special educa-tional needs in England and Scotland; and on the computerization of social securityin thirteen countries. The article concludes by attempting to show that this approachto the study of administrative justice satisfies all the defining characteristics of thesocio-legal paradigm.

I. INTRODUCTION

The aims of this article are to develop a theoretical framework for analyzing

administrative justice and to show how it has been applied in empirical

research on administrative decision-making. The term administrative justice

is used in the article to refer to the principles that can be used to evaluate the

* The author would like to record his considerable indebtedness to Roy Sainsbury, BrianLonghurst, Enid Mordaunt, Sheila Riddell, and Paul Henman, who not only collaborated withhim in the research described in this article but also helped to clarify his thinking on administrat-ive justice. He would also like to thank Denis Galligan, Neville Harris, and Simon Halliday fortheir very helpful criticisms of earlier versions of this article, and two anonymous referees whosecomments forced him to sharpen up his arguments.

Address correspondence to Professor Michael Adler, School of Social and Political Studies,University of Edinburgh, Adam Ferguson Building, George Square, Edinburgh EH8 9LL, UnitedKingdom. Telephone: 0044-131-650-3931; e-mail: [email protected].

LAW & POLICY, Vol. 25, No. 4, October 2003 ISSN 0265–8240© Blackwell Publishing Ltd. 2003, 9600 Garsington Road, Oxford OX4 2DQ, UK,and 350 Main Street, Malden, MA 02148, USA.

324 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

justice inherent in administrative decision-making. This comprises procedural

fairness, which is concerned with the process of decision making, that is, with

the ways in which individual citizens are treated, as well as substantive justice,

which refers to the outcomes of decision making, that is, to the benefits or

burdens that are conferred on individual citizens. The use of the term justice

in this context has been questioned but, in light of the fact that the principles

referred to above are normative principles that embody legitimate expecta-

tions, it is contended that this is an entirely appropriate use of the term.

The article is in two parts. The first part of the article outlines two com-

plementary approaches to the enhancement of administrative justice – an

approach that emphasizes external forms of accountability and focuses on the

small number of decisions that are the subject of an appeal or complaint and

an approach that emphasizes internal forms of accountability and focuses on

the much larger number of first-instance decisions. External mechanisms of

accountability are necessary for the achievement of justice in administrative

decision-making but, on their own, they are not sufficient to ensure adminis-

trative justice. An argument is therefore made for the importance of internal

mechanisms, not as an alternative but, rather, as a complement to external

mechanisms. The article develops a theoretical framework for analyzing

administrative justice in terms of the “trade-offs” between different normative

models of administrative decision-making and, in the second part, demon-

strates how it has informed empirical research on the impact of computer-

ization on social security in the United Kingdom; on decision making in

the Scottish prison system, on the assessment of special educational needs

in England and Scotland, and on the computerization of social security in

thirteen countries.

It should be noted that most of the examples cited in the first part of the

article are British and that three of the four studies described in second part

refer to Britain (the fourth study was explicitly comparative). However, the

general approach to administrative justice outlined in the article undoubtedly

has more general application. The article concludes by distinguishing four

key modes of legal scholarship: law in action; doctrinal or “black-letter” law;

legal, political, and social philosophy; and socio-legal studies and attempts to

demonstrate that the approach outlined in the paper exemplifies the socio-

legal paradigm.

II. DEVELOPING A THEORETICAL FRAMEWORK FOR ANALYSING

ADMINISTRATIVE JUSTICE

One orientation to administrative justice equates it with a set of principles

that are imposed on government departments and public bodies by institu-

tions that are external to them, in particular by the courts but also by

tribunals and agencies like ombudsmen, as a result of individual challenges to

Adler ADMINISTRATIVE JUSTICE 325

© Blackwell Publishing Ltd. 2003

administrative decisions. Another orientation sees it as a set of principles that

are put into place internally, that is, by government departments and public

bodies themselves, as a result of some internal monitoring of their adminis-

trative decisions. The distinction is an analytic one, and, in practice, the two

orientations are frequently combined. This makes a great deal of sense because

the aims of the two approaches are, in many respects, complementary –

the former is backward-looking and is primarily concerned with providing

individuals with a mechanisms for redressing any injustice they may have

experienced while the latter is forward looking and aims to promote justice

in administration for the benefit of all – and, although it is undoubtedly

important that mechanisms for redressing injustice should exist, it is equally

important that efforts should be made to prevent injustice from arising in the

first place.1 Each of the two orientations is considered in turn.

A. EXTERNAL ORIENTATIONS

1. Procedural Protection for “Adjudicative” Decisions

As Harlow and Rawlings (1997:495) note, there has been a general tendency

for the courts to model the administrative process in their own adjudicative

image. Fuller (1978:353) has argued that the distinguishing characteristic of

adjudication is to confer “on the affected person a peculiar kind of participa-

tion in the decision, that of presenting proofs and reasoned arguments for a

decision in his favor.” This implies that, for a decision to be adjudicatory,

certain procedural constraints must be placed on the decision maker. The

model of procedure that facilitates the presentation of “proofs and reasoned

arguments” is exemplified by the criminal and civil courts but is also found in

other settings, such as tribunal proceedings, which share some of the same

features. Actual decision making may have all or some or none of the char-

acteristics of adjudication, and Harlow and Rawlings (1997:496) conceive of

a sliding scale – the closer to the ideal type outlined above, the more

judicialized the process and the more the courts will insist on applying the

rules of natural justice; the further away, the less judicialized the process and

the less the courts will insist on applying such principles.2

Seventy years ago, the Donoughmore Committee on Ministers Powers

(Great Britain 1932:74) sought to distinguish three categories of decision

making, namely “judicial,” “quasi-judicial” and “administrative” decision-

making, each of which was based on a different type of dispute; but its

thinking was crude and its arguments circular. The Committee’s approach

had some attractions for “green light theorists,” who assume that the state

is the only effective guarantee for individual freedom, because it serves to

insulate administrative decisions from legal scrutiny, and emphasize the role

of legislation and regulation rather than the use of the courts. However, this

approach has been criticized on the grounds that it is very difficult (if not

326 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

impossible) in practice to separate these three types of functions, and because

one effect of applying it would be to deprive large numbers of decisions

of any procedural protection on the grounds that they are “administrative.”

The Committee’s approach had little attraction for “red light theorists,”

who assume that the state is a threat to the freedom of the individual and

favor a strong role for the courts in scrutinizing the legality of administrative

decisions.3

Exponents of “flexible protection” claim that it provides a way around the

problems of classifying decisions into those that are “adjudicatory” and those

that are “administrative” and a way of applying the rules of natural justice

only to the former while the latter are not given any protection at all. Mullan

(1975:25) recognizes that different types of decision making have different

procedural requirements and argues that the more closely they resemble

“straight law/fact determinations resulting in serious consequences for those

concerned, the more legitimate is the demand for procedural protection;

while the more closely they resemble broad, policy-oriented decisions, the

less they are in need of such protection.” Although this approach avoids the

problems associated with a rigid division of decisions into two types and

claims to recognize that different types of decisions require different forms of

procedural protection, it operates with a model of administrative justice that

provides no protection for decisions that involve the application of policy. A

further problem is that the optimum degree of protection is not instantly

recognizable and that a fair measure of discretion is required.

This position has led Harlow and Rawlings (1997:510) to question how far

it is the court’s job to pursue the optimum form of procedure for different

kinds of decision. They question whether judges have the information or the

expertise needed to determine the particular form of procedural protection

that is appropriate for different types of decisions and, even, whether such an

activity is consistent with the rule of law. Leaving these normative issues

aside, they claim (ibid.:516) that, in recent years, there has been a measure of

increased judicial activism and of greater flexibility of response. However,

such developments still leave broad swathes of “administrative” decisions

unprotected by the courts.

Disputes between the citizen and the state are more likely to be heard by a

tribunal than by a court and, in the 1950s, the Franks Committee (Great

Britain 1957) enunciated three principles that apply to tribunal decision-

making. According to Franks (ibid.: para. 42), these principles are openness,

fairness, and impartiality: openness requires publicity for the proceedings and

knowledge of the essential reasoning underlying the decision; fairness requires

the adoption of clear procedures that enable the parties to understand their

rights, present their case fully, and know the case that they have to meet;

while impartiality requires freedom from the influence, real or apparent, of

departments concerned with the subject matter of their decisions. However,

it should be noted that these principles apply to tribunals that hear appeals

Adler ADMINISTRATIVE JUSTICE 327

© Blackwell Publishing Ltd. 2003

from first-instance, administrative decisions and not to the decisions them-

selves, and that, in any case, they represent “good practice” and are not

enforceable. Although the role of the Council on Tribunals, set up under the

Tribunals and Inquiries Act 1959, is to keep the constitution and working of

tribunals under review, its powers are very limited.

2. Procedural Protection for “Administrative” Decisions

The first Parliamentary Commissioner for Administration (PCA) was

appointed in 1967 to deal with grievances from individuals who felt they had

suffered an injustice arising from maladministration by a central government

department for which there was no available remedy. “Maladministration”

was not defined in the Parliamentary Commissioner Act 1967 that estab-

lished the PCA, although the Leader of the House of Commons, Rt. Hon.

Richard Crossman, MP, described it as including: bias, neglect, inattention,

delay, incompetence, ineptitude, perversity, terpitude, arbitrariness and so

on” (Great Britain 1966: col. 51). Subsequently, a Parliamentary Commis-

sioner, Sir William Reid, elaborated on Crossman’s list by giving more

examples of what the term covers (Great Britain 1993: para. 7).4 “Injustice”

was likewise not defined in the Act but, for a complaint to be upheld, the

PCA must conclude that the individual suffered some kind of loss that would

otherwise not have occurred.

A number of other ombudsmen have subsequently been established to

deal with complaints about services in the public sector that fall outside the

PCA’s remit. They include the Health Services Ombudsmen for England and

Wales and for Scotland; the Commissioners for Local Administration (CLA)

for England and Wales and for Scotland; the Housing Association Ombuds-

men for England and Wales and for Scotland; and the Scottish Parliament-

ary Ombudsman.5 In addition, some public services, for example, the Inland

Revenue; the Customs and Excise Department; the Police Service (in England

and Wales, and in Scotland); the Prison Service (in England and Wales, and

in Scotland); and the Driver and Vehicle Licensing Agency (DVLA) have set

up procedures for investigating complaints of maladministration against them.

There are also ombudsmen covering various private-sector services such as

banking, broadcasting, building societies, estate agents, funerals, insurance,

investments, legal services, and pensions but, with the exception of the Legal

Services Ombudsmen for England and Wales and for Scotland and the

Pensions Ombudsmen, they do not have statutory powers.6

Ombudsmen use inquisitorial methods to investigate allegations that

maladministration has given rise to injustice. Although, in a given case, they

may find evidence of maladministration that has not given rise to any identi-

fiable injustice, their approach to administrative justice is, in effect, an instru-

mental one. Where a complaint that maladministration has given rise to

injustice is upheld, they may impose remedial action on the organization that

328 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

was the subject of the complaint; award compensation to the complainant;

or instruct the organization to issue an apology.7 In addition to providing

remedies for those who complain, ombudsmen may order the organization

complained of to modify its administrative procedures. However, although

an organization may wish to do so to avoid more complaints in the future,

there is, in general, no systematic check on whether or not it has done so.

Ombudsmen are independent of the organization against which the com-

plaint is made; no legal (or other) representation is required, and complain-

ants are not subject to any charges. Independence should, in theory, guarantee

impartially, but this may be compromised in practice by the fact that the

ombudsman’s staff is often drawn from, and sometimes return to, the

organizations against which complaints are made.8 In addition, there are

often barriers that have to be surmounted before a complaint can be made to

an ombudsman, and, except for the CLA in Northern Ireland, ombudsmen’s

remedies are not legally enforceable (Himsworth 1985). In spite of these

shortcomings, and in the absence of any systematic empirical evidence on the

subject, official reports indicate that the opportunity of complaining to an

ombudsman has contributed, if only to a small extent, to enhancing the

justice inherent in administrative decision-making.

While most people have looked to courts, tribunals, and other forms of

accountability, such as ombudsmen, which are external to the locus of adminis-

trative decision-making, as the means of achieving administrative justice,9

the available evidence does not suggest that this approach is particularly

effective on its own. This is, in part, because few of those who experience

injustice actually appeal to courts, tribunals, or ombudsmen; in part because

court, tribunal, and ombudsman decisions have a limited impact on the

corpus of administrative decision-making.10 As a result, as Ison (1999:23)

points out, “the total volume of injustice is likely to be much greater among

those who accept initial decisions than among those who complain or appeal.”

There is, of course, considerable variation in the impact of different forms

of external accountability on administrative decision-making. The more

authoritative the judgments are, the more publicity is given to them, and the

stronger the enforcement procedures are, the greater the impact they are

likely to have and the more effective they are likely to be in achieving admin-

istrative justice.11 In considering the impact of judicial review, Halliday (2001)

identifies three sets of conditions for successful impact: the model of adminis-

trative justice enunciated in the judgment must be coherent and consistent

with the existing body of law; decision makers (administrators) must be

“legally conscientious,” that is, they must be committed to legality; and the

decision-making environment must ensure that the model of administrative

justice enunciated in the judgment is given priority over other models. More

often than not, one or more of these conditions is not met, and it is for this

reason that Mashaw (1974:776) has argued that additional safeguards, such

as internal quality controls or quality assurance systems, are needed to

enhance administrative justice.

Adler ADMINISTRATIVE JUSTICE 329

© Blackwell Publishing Ltd. 2003

B. INTERNAL ORIENTATIONS

1. Mashaw’s Approach

In his pioneering study of the American Disability Insurance (DI) scheme,

Mashaw (1983:21–22) detected three broad strands of criticism leveled against

it: the first indicted it for lacking adequate management controls and produ-

cing inconsistent decisions; the second for not providing a good service and

failing to rehabilitate those who were dependent on it; and the third for not

paying enough attention to “due process” and failing to respect and uphold

the rights of those dependent on it. He claimed that each strand of criticism

reflects a different normative conception of the DI scheme, that is, a different

model of what the scheme could and should be like. The three models are

respectively identified with bureaucratic rationality, professional treatment,

and moral judgment.

Mashaw (ibid.:24) defines “administrative justice” (the justice inherent in

routine day-to-day administration) in terms of “those qualities of a decision

process that provide arguments for the acceptability of its decisions.” Two

points of clarification are called for here: by “acceptability,” it should be

assumed that Mashaw means something like “legitimacy,” and by “decisions

that he is referring to “outcomes.” Although one could infer from this that

he takes an instrumental approach to administrative justice, this is not the

case. On the contrary, his definition of administrative justice acknowledges

that it can contribute to the acceptability of decisions or outcomes even when

these decisions or outcomes are incorrect. The significance of administrative

justice is that it can legitimate such decisions.

It follows from Mashaw’s definition that each of the three models he

described is associated with a different conception of administrative justice.

Thus, there is one conception of administrative justice based on bureaucratic

rationality, another based on professional treatment, and a third based on

moral judgement. According to Mashaw (ibid.:31), each of these models is

associated with a different set of legitimating values, different primary goals,

a different organizational structure, and different cognitive techniques. These

are set out in the Table 1 below.

Although this is very helpful, the association of fairness with one of the

models (the moral judgment model), and the implication that the two other

models are “unfair,” is unfortunate. In addition, Mashaw’s characteriza-

tion of the three models reflects an exclusively internal orientation to admin-

istrative justice in that it makes no reference to external mechanisms for

redressing grievances. With this in mind, Table 1 has been revised. Table 2

renames the three models (it refers to them as a bureaucratic model, a profes-

sional model, and a legal model), alters the ways in which they are charac-

terized,12 and highlights redress mechanisms that include external as well as

internal procedures for achieving administrative justice. This is important

because internal and external procedures should not be seen as alternatives,

330 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

Table 1. Three Normative Models of Administrative Justice – Original Formulation

Model

Bureaucratic

Rationality

Professional

Treatment

Moral Judgment

Source: As set out by Mashaw (1983:31).

Structure or

Organization

Hierarchical

Interpersonal

Independent

Cognitive

Technique

Information

Processing

Clinical

Application

of Knowledge

Contextual

Interpretation

Primary Goal

Program

Implementation

Client

Satisfaction

Conflict

Resolution

Legitimating

Values

Accuracy and

Efficiency

Service

Fairness

and it is important to recognize that, for good reasons, they are frequently

combined.

Mashaw (1983:23) contends that each of the models is coherent, plausible,

and attractive and that the three models are competitive rather than mutually

exclusive. Thus, they can and do coexist with each other. However, other

things being equal, the more there is of one, the less there will be of the other

two. His insight enables us to see both what trade-offs are made between the

three models in particular cases and what different sets of trade-offs might be

more desirable. His approach is a pluralistic one, which recognizes a plurality

of normative positions and acknowledges that situations that are attractive

for some people may be unattractive for others.

Mashaw’s pluralism can be contrasted with the communitarian version of

pluralism adopted by other writers on justice, most notably by Walzer. Walzer

(1983:6) accepts that “the principles of justice are themselves pluralistic in

form [and that] different social goods ought to be distributed for different

reasons, in accordance with different procedures, by different agents.” How-

ever, he also claims (ibid.:8–9) that “the meaning of the goods in question

determines their distribution” and argues that “if we understand what it is,

what it means for those for whom it is a good, we understand how, by

Table 2. Three Normative Models of Administrative Justice – Revised Formulation

Model

Bureaucratic

Professional

Legal

Source: Based on Adler and Longhurst (1994:44).

Characteristic

Remedy

Administrative review

Second opinion or

complaint to a

professional body

Appeal to a court or

tribunal

Mode of

Accountability

Hierarchical

Interpersonal

Independent

Legitimating

Goal

Accuracy

Expertise

Legality

Mode of

Decision Making

Applying rules

Applying

knowledge

Asserting rights

Adler ADMINISTRATIVE JUSTICE 331

© Blackwell Publishing Ltd. 2003

whom, and for what reasons it ought to be distributed.”13 Although Walzer

accepts that social meanings are historical in character and that what is

regarded as just and unjust changes over time, he nevertheless assumes a

degree of normative consensus in a given community, which stands in stark

contrast with Mashaw’s assumption of normative conflict.

The trade-offs that are made, and likewise those that could be made,

reflect the concerns and the bargaining strengths of the institutional actors

who have an interest in promoting each of the models, typically civil servants

and officials in the case of the bureaucratic model; professionals and “street

level bureaucrats” (Lipsky 1980) in the case of the professional model; and

lawyers, court and tribunal personnel, and groups representing clients’ inter-

ests in the case of the legal model.

These trade-offs vary between organizations and, within a given organ-

ization, between the different policies delivered by that organization and

between the different stages of policy implementation. They also vary over

time and between countries. In the case of the (American) DI scheme, Mashaw

concluded that, in the early 1980s when he carried out his study, bureau-

cratic rationality was the dominant model and, notwithstanding variations

within and between countries, it is at least arguable that bureaucracy is, and

always has been, the dominant model as far as the administration of social

security is concerned. However, the professional model may be dominant

in other policy fields, for example, in the delivery of health care. Although

the legal model is clearly of greater importance in countries with a strong

rights culture, it has rarely supplanted the dominant bureaucratic and pro-

fessional models.

Mashaw’s approach is a very attractive one. Although, Bayles (1990:163–

89) and Galligan (1996:25) both recognize that a uniform set of procedural

principles does not apply across the board and that the appropriateness

of any set of procedural principles depends on the characteristics of the

decision-making process in question, their approaches to procedural fairness

are less sophisticated than Mashaw’s in that they assume that the different

forms of decision making are sui generis. The great strength of Mashaw’s

approach is his recognition that different models of decision making coexist

with each other and that each of them is associated with a different conception

of administrative justice. The administrative justice of any given instance of

decision making is not represented by the procedural principles associated with

the single model that best describes that form of decision making but, rather,

by trade-offs between each of the models for which there is some evidence.

Mashaw’s approach has been subjected to a number of criticisms. Although

he contends that the three models described above, and only these three

models, are always present in welfare administration, this claim can be

disputed. The bureaucratic, professional, and legal models have, in many

countries, been challenged by a managerial model associated with the rise

of new public management (see Hood 1991, 1998), a consumerist model that

focuses on the increased participation of consumers in decision making (see

332 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

Clarke & Newman 1997) and a market model that emphasizes consumer

choice (see Johnson 1995).

A second criticism is that, in assessing the relative influence of the three

models, Mashaw ignores their absolute strengths. Consider two situations in

which the strengths of three models are given weights of 30, 20, and 10 units

and 3, 2, and 1 units – although they are identical in a relative sense, they are

quite different in absolute terms and clearly refer to what are, in reality, very

different situations. “Strong” balances are very different from “weak” bal-

ances in ways that Mashaw’s analysis does not bring out very well.

A third criticism is that Mashaw takes the policy context for granted.14

However, just as different orientations to administration, that is, to how

programs should be run, can be understood in terms of a number of normat-

ive models that are in competition with each other, so different orientations

to policy, that is, to what programs aim to achieve, can also be understood in

this way. As shown below, Adler and Longhurst (1994) have demonstrated

that Mashaw’s approach can be applied to competing models of policy as

well as to competing models of administration. Each of several competing

models of policy may, in theory, be combined with each of several competing

models of administration. The resulting “two-dimensional” model is neces-

sarily more complex than a “one-dimensional” model but its characteristics

are similar in that it not only makes it possible to understand the trade-offs

that are made between different combinations of policy and administration

in particular cases, but also makes it possible to see what different sets of

trade-offs might be more desirable.15 Since, applying the terminology used

in this article, the models of policy refer to outcomes while the models of

administration refer to process, the two-dimensional model provides a way

of combining procedural fairness with substantive justice.

2. Developing Mashaw’s Approach

In light of the criticisms above, some modifications of Mashaw’s analytic

framework are clearly called for. First, it is important to recognize that, in

addition to the bureaucratic, professional, and legal models identified by

Mashaw, some additional models of administrative justice need to be con-

sidered. Three such models are a managerial model, a consumerist model,

and a market model, although they are not necessarily all present in every

administrative system. Second, account needs to be taken of the absolute as

well as the relative strengths of these models. Third, the approach can be

applied to competing normative models of outcomes, that is, to substantive

justice, as well as to competing normative models of process, that is, to

procedural fairness, and used to analyze the interactions between them.

An extended analytic framework, involving six normative models of

administrative justice, is set out in Table 3 below. A brief explanation of this

extended analytic framework is called for. During the postwar period, most

public welfare services in the United Kingdom were shaped by the bureaucratic

Adler ADMINISTRATIVE JUSTICE 333

© Blackwell Publishing Ltd. 2003

Table 3. Six Normative Models of Administrative Justice

Model

Bureaucratic

Professional

Legal

Managerial

Consumerist

Market

Characteristic

Remedy

Administrative

review

Second opinion or

complaint to a

professional body

Appeal to a court or

tribunal (public law)

Publicity

“Voice” and/or

compensation

through Consumer

Charters

“Exit” and/or court

action (private law)

Mode of

Decision-Making

Applying rules

Applying

knowledge

Asserting rights

Managerial

autonomy

Consumer

participation

Matching supply

and demand

Legitimating

Goal

Accuracy

Expertise

Legality

Efficiency

gains

Consumer

satisfaction

Profit making

Mode of

Accountability

Hierarchical

Interpersonal

Independent

Performance

indicators

Consumer

Charters

To owners or

shareholders

and professional models outlined above, although the trade-off between them

varied from one policy domain to another. However, by the mid-1980s, the

pattern of administration had come under attack. It was variously criticized

for lacking neutrality and being biased against certain groups; for its failure

to contain the growing demand for cost savings; for having a vested interest

in the maintenance and expansion of existing structures and not promoting

the “public interest”; and, as a “monopoly provider,” for being insulated

from competitive pressures to become more efficient and more responsive

to the demands and preferences of consumers. New and better forms of

management were championed as the most appropriate response to these

criticisms. “Managerialism,” as this approach came to be known, challenged

the powers and prerogatives of bureaucrats and professionals in the name of

managers who demanded the “freedom to manage” the attainment of pre-

scribed standards of service. It gave priority to achieving efficiency gains,

introduced different forms of financial and management audit to assess how

well the prescribed standards of service had been met, rewarded staff who

performed well, and, in theory at least, sanctioned those who did not (see

Clarke & Newman 1997). Perhaps inevitably, the introduction of these new

managers frequently led to struggles for power and control within welfare

organizations. Managerialism can thus be characterized in terms of manager-

ial autonomy, the pursuit of efficiency gains, the use of performance indicators

to assess accountability, and the possibility of drawing attention to the fact

that prescribed standards have not been met as a means of putting pressure

on management to improve their standards.

334 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

Consumerism has, likewise been a central reference point in the drive for

public sector reform from the mid-1980s onwards (ibid.: chap. 6). Like man-

agerialism, it has been taken up as a response to criticisms of the bureaucratic

and professional, and the reshaping of welfare services around consumer

choice has been visible in a number of reforms, in particular in the introduc-

tion in the UK of the “Citizen’s Charter” (Great Britain 1991; Page 1999).

Consumerism embodies a more active view of the service user, who is seen as

an active participant in the process rather than a passive recipient of bureau-

cratic, professional, or managerial decisions. It can thus be characterized in

terms of the active participation of consumers in decision making, consumer

satisfaction, the introduction of consumer “charters,” and the use of “voice”

(Hirschman 1970), together with the possibility of obtaining compensation

where the standards specified in the charter are not met as available remedies.

Markets constitute the final model in the extended analytic framework and

have many of the characteristics of the managerial and consumerist models

(although the reverse is not necessarily the case). Decision making in the

market involves the matching of demand and supply and is made with refer-

ence to the price mechanism. Individuals are viewed as rational economic

actors who choose the producer who best satisfies their wants or preferences.

The legitimating goal of the producer organization is profit-making, while

the prevailing mode of accountability is to the owners or shareholders. In

contrast to consumerism, where the consumer can use “voice” as a remedy,

and can obtain compensation through consumer charters if the specified

standards have not been met, markets provide the possibility of “exit” (ibid.).

In addition, an aggrieved individual may be able to raise a court action for

compensation where he or she suffers some measurable loss from an admin-

istrative decision. Internal or quasi-markets (Le Grand 1991; Le Grand &

Bartlett 1993) have some but not all of the characteristics of the market

model just outlined.

In a recent book, Halliday (2003) argues that the managerial, consumerist,

and market models of administrative justice elaborated above are better

understood as components of a single model, which he associates with the

“new” public management (see Hood 1991, 1998). However, I think this

formulation obscures more than it reveals. This is because each of the three

models has its own internal logic and because these three models are “com-

petitive” in the sense that Mashaw uses this term to describe the relationship

between the bureaucratic, professional, and legal models in his account of

administrative justice. It follows that the managerial, consumerist, and mar-

ket models of administrative justice can be, and frequently are, combined.

When they are combined, the combination does provide a good account of

the “new” public management. However, they need not be combined in this

way, and each of the models is sometimes encountered on its own, that is,

not in combination with the others.

It may seem a little strange to refer to some of these models, in particular,

the market model, as models of justice. After all, markets are often regarded

Adler ADMINISTRATIVE JUSTICE 335

© Blackwell Publishing Ltd. 2003

as threats to justice, which undermine its achievement in practice. This strange-

ness is more apparent than real and results from equating the concept of

“justice” with substantive justice rather than with procedural fairness. In

seeking to demonstrate that procedural fairness is an important component

of overall justice and that the idea that “everyone should receive what is due

to them” (Miller 1976:20) applies just as much to procedures as to outcomes,

this article adopts a very different starting point. However, how people should

be treated is very dependent on context, and, as the discussion above tries to

make clear, different conceptions of procedural fairness are associated with

different types of decision making.

Consider the case of a lottery. Everyone who takes part has a legitimate

expectation that they will be treated in exactly the same way (without any

reference to their personal circumstances or characteristics) and that they

will have exactly the same chance of being selected as everyone else.16 Some

lotteries stipulate that an individual can only be selected once, while others

permit multiple selection, thereby enabling an individual who has already

been selected to be selected again. However, in both cases, the outcome is (or

ought to be) completely random and determined only by chance. If lottery

outcomes are judged in terms of any of the familiar principles of distributive

justice (need, desert, rights, or strict equality), or any combination of these

principles, they will no doubt be considered unjust. However, this is not the

case for lottery procedures, which should incorporate the principles of chance

and equality that form the basis of the legitimate expectations referred to

above. A similar argument applies to markets.

In a market, consumers can legitimately expect to have the opportunity to

express their wants in terms of their preferences and the freedom to choose.17

Likewise, producers can legitimately expect to be able to respond to con-

sumer choice and produce goods and services up to the point where the

value to the consumer of his (her) marginal purchase just exceeds the cost of

supply – beyond that point the consumer will not be interested in purchasing

the service, and it will not pay the producer to go on producing it. The price

mechanism stabilizes as well as controls these transactions (where there are

shortages, higher prices act as an incentive to economize, but, where the

market is flooded, prices will fall, and this will encourage consumers to buy

more), while the profit motive should make supply responsive to demand,

and encourage technological innovation and progress. Competition should

lead to greater efficiency – successful producers will thrive (and make big

profits) while unsuccessful ones will go to the wall. However, since market

theorists tend to treat the initial distribution of resources as an exogenous

variable, assessing the final distribution of resources in terms of any of the

familiar principles of distributive justice, or any combination of principles,

will usually conclude that, like the lottery, they are not just.18 But, although

market outcomes are very rarely just, and it follows that markets do under-

mine attempts to achieve substantive justice, market procedures may incor-

porate the principles underlying the legitimate expectations referred to above

336 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

and may thus be fair.19 It is in this sense, and only in this sense, that it makes

sense to talk about the market as a model of justice.

3. Normative Theorizing

One of the attractions of Mashaw’s approach is that it makes it possible to

understand the trade-offs that are made between different justice models in

particular instances and to see whether another set of trade-offs might be

more desirable, not merely for those institutional actors who have an interest

in promoting each of the models in play but for all concerned, that is, in

some overall sense. However, the attempt to arrive at a “better” balance

between the different models in play and identify a different set of trade-offs

that will enhance “the public interest” raises a number of problems. Since

there is no “magic formula,” this exercise necessarily involves the exercise of

judgment. However, this does not make it an arbitrary exercise. It is one that

can, and should, be informed by empirical data. In this connection, perform-

ance measures of various kinds and audit data are important, and, to the

extent that the primary justification for public services is that they should

serve the public, it can be argued that the results of user surveys and public

opinion data are especially important.

III. APPLYING THIS THEORETICAL FRAMEWORK IN EMPIRICAL

RESEARCH ON ADMINISTRATIVE JUSTICE

The approach to administrative justice outlined above has informed four

pieces of research undertaken by the author over the last fifteen years. These

comprise research on the impact of computerization on social security in the

UK, on decision making in the Scottish prison system, on the assessment of

special educational needs in England and Scotland, and on the computeriza-

tion of social security in thirteen countries. In each case, an attempt was

made to specify a number of different models of administrative justice and to

collect data that would enable the strength of these models to be assessed.

A brief account of each of these pieces of research is set out below.

A. THE IMPACT OF COMPUTERIZATION ON SOCIAL SECURITY IN THE UK

The three main aims of the “Operational Strategy” (Great Britain. Department

of Health and Social Security 1982:1), the massive program to computerize

the entire social security system that the UK government attempted to put

into place in the 1980s, were:

• to improve operational efficiency, reduce administrative costs, and

increase the flexibility of the operational system to respond to changing

requirements;

Adler ADMINISTRATIVE JUSTICE 337

© Blackwell Publishing Ltd. 2003

• to improve the quality of service to the public, for example, by treating

customers in a less compartmentalized benefit-by-benefit manner and

more as “whole persons” with a range of social security business, and to

improve the provision of information to the public;

• to modernize and improve the work of social security staff.

Of these aims, the first reflected the interests of the government, the second

reflected the interests of the claimant, and the third reflected the interests of

the staff.

By 1989, the Operational Strategy had run into deep trouble – the costs

had escalated so steeply that, unless improvements in quality of service were

taken into account, it was clear that the future of the program was in doubt.

Against this background, the Department of Social Security (DSS) commis-

sioned some research on “quality of service” and, in particular, on the “whole

person” concept (Adler & Sainsbury 1990). Using a consultative procedure

known as the Delphi Method (Adler & Sainsbury 1996), four panels of

experts (drawn from DSS staff; welfare rights officers; representatives of

pressure groups, academics, and researchers; and persons with backgrounds

in other organizations or from overseas social security systems) were invited

to comment on the desirability and feasibility of a number of different con-

ceptions of quality of service and of the whole person concept, and on a

number of different models of organization. In addition to these questions,

the research attempted to investigate the effects of computerization on admin-

istrative justice, that is, on the justice inherent in routine day-to-day admin-

istration (Adler & Sainsbury 1991a, 1991b).

The research assumed that the three normative models of administrative

justice set out in Table 2 above were all present in the administration of social

security but claimed that the Operational Strategy would further strengthen

the dominance of the bureaucratic conception of administrative justice at the

expense of the two competing conceptions. Thus, it was likely to lead to an

even more bureaucratized system rather than one that was more sensitive to

the needs and circumstances of claimants or one that made it easier for them

to assert their rights. The main reasons for this were that the DSS adopted a

“top-down” orientation to computerization that gave priority to the interests

of the government rather than a “bottom up” orientation that would have

given priority to the interests of claimants or staff; and that the aim of the

program was to make administrative savings rather than to improve quality

of service (whatever that might mean). The study concluded that the overall

effect of the program was certainly to alter but not necessarily to enhance the

procedural fairness of administrative decision-making in social security.

B. DISCRETIONARY DECISION-MAKING IN THE SCOTTISH PRISON SYSTEM

During the late 1980s, a program of research on day-to-day administrative

decision-making in the Scottish prison system, focusing on adult, male,

338 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

long-term prisoners, who constituted the largest and, arguably, most prob-

lematic of the various groups that made up the prison population, was under-

taken (Adler & Longhurst 1994). Using a mixture of documentary analysis,

observation, and interviews with a wide range of individuals inside and out-

side the Scottish Prison Service (SPS), the research investigated a number of

important areas of decision making, associated with classification, transfers,

regimes, and accountability, in detail.20 In each case, it sought to establish

what decisions were accomplished; why the existing system operated in the

way it did; what problems were created by existing practices; for whom they

were problematic; to what extent they gave rise to pressures for change;

what alternatives to the present system were being canvassed; and what their

implications for day-to-day decision making would be. The same methods

were used to study the policymaking process by carrying out a detailed exam-

ination of a series of policy documents that were published by the SPS

during the period of our research.

The period was one of great turbulence for Scottish prisons – a spate of

violent disturbances had given rise to a vigorous debate about what prisons

were for and how they should be run. Although some people argued that this

was a bad time to study Scottish prisons because so much was changing, and

it was far from clear what the eventual outcome would be, it actually turned

out to be a very good time, because the arguments advanced by powerful inter-

ests were forcefully expressed and easily accessible, and because their struggles

for control were overt and visible. This enabled us to see the processes at

work particularly clearly, and made it easier to construct a theoretical frame-

work than it would have been in more settled and less turbulent circum-

stances. Using an iterative procedure based on “wide reflexive equilibrium”

(Rawls 1971:46–53), a mutual adjustment between this theoretical frame-

work and the empirical reality that was the focus of our research was achieved.

While the research on the Operational Strategy had utilized the three

normative models of administrative justice set out in Tables 1 and 2 above,

the turbulence of the prison system resulted in the calling into question of the

various normative models of substantive justice as well, and this suggested

that they also needed to be considered. Each of the justice models was asso-

ciated with a distinctive discourse and “ends discourses,” which are con-

cerned with what prisons are for, were distinguished from “means discourses,”

which are concerned with how prisons should be run. The discourses in play

are set out in Tables 4 and 5 below.

The ends and the means discourses were combined to produce a discourse

matrix that summarizes the discursive structure of the Scottish prison system

at the time, and the individuals, groups, and institutions whose discourses

were associated with each of the cells in the matrix were identified. This is set

out in Table 6 below.

The research was based on the twin assumptions that groups in particular

settings produce discourses that reflect their interests, and that discursive

struggles lie at the heart of the power struggles that are endemic in every

Adler ADMINISTRATIVE JUSTICE 339

© Blackwell Publishing Ltd. 2003

Table 4. Characteristics of Three Competing “Ends Discourses”

Source of

legitimacy

Focus

Dominant

concerns

Rehabilitation

Discourse

Improving the

individual

“Deviant” individual

Socializing the

prisoner back into

society through the

provision of training

and treatment

Control

Discourse

Control of disruption;

smooth running of

establishments

“Disruptive”

individual

Good order and

discipline; protection

of prison staff

Normalisation

Discourse

Prevention of negative

effects of prison; treating

prisoners like individuals

in the community

“Normal” individual

Minimum security;

contact between the

prisoner and his or her

family; improved living

conditions

setting. However, although they always exist, they are particularly evident in

periods of flux. The importance of external and contextual factors is that

they structure the power relations between internal interest groups and shape

the outcome of the power struggles between them. By applying the theoretical

Table 5. Characteristics of three competing “Means Discourses”

Source of legitimacy

Focus

Dominant concerns

Accountability

for decisions

Legal

Discourse

Rule of law

On individual

prisoners

Respect for

prisoners rights

External

Professional

Discourse

Knowledge based

on experience

On establishments

Leadership,

experience,

judgment,

enhancing the

institutional ethos

Negotiated

Bureaucratic

Discourse

Rules and regulations

On the system

Uniformity, consistency,

fidelity to the rules

Internal

Table 6. Discourse Matrix for the Scottish Prison System

Bureaucracy

Professionalism

Legality

Rehabilitation

Parole Board

Barlinnie Special Unit

“Left realist”

academics, SCCL

Normalization

HQ – Administration

Division

Most governors

Rights-orientated

governors, SACRO

Control

HQ – Operations

Division

SPOA

Judges and

courts

340 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

framework outlined above to various areas of decision making and to the

policy process, the research was able to demonstrate that the struggle for

control between the interest groups represented in the discourse matrix had

profound implications for procedural fairness and substantive justice. In addi-

tion, a new form of managerial discourse was identified (Adler & Longhurst

1994:236–38), which, at the end of research, had assumed a position of domin-

ance in Scottish penal policy.

C. THE ASSESSMENT OF SPECIAL EDUCATIONAL NEEDS IN ENGLAND

AND SCOTLAND

The assessment of special educational needs is a very complex process that

can be divided into a number of discrete and overlapping stages and involves

input from a large number of individuals, including educational psycholo-

gists, education officers, head teachers, class teachers, medical practitioners,

social workers, “named persons,” parents, and children. This process was

structured by a statutory “Code of Guidance” in England (Great Britain.

Department for Education 1994)21 and by a non-statutory “Manual of Good

Practice” in Scotland (Scottish Office Education and Industry Department

1998). The outcomes of the assessment process were extremely significant for

those involved, in particular for children and their parents, since they had a

direct interest in the type of education the child received and the resources

that were made available for this purpose. However, as with many decentral-

ized decision-making processes, there are wide variations in outcomes be-

tween and among (local) educational authorities in England and Scotland.22

Although the Code of Guidance and, to a lesser extent, the Manual of Good

Practice have undoubtedly led to a degree of standardization, there are still

wide variations between authorities in the procedures used to assess special

educational needs in the two countries.

The research on the assessment of special educational needs involved an

empirical study of the fairness of the different procedures used in England

and Scotland (Riddell, Adler, Mordaunt & Farmakopoulou 2000; Riddell,

Wilson, Adler & Mordaunt 2002). It aimed to describe the range of practices

that constitute statutory assessment in England and in Scotland and to ana-

lyze the nature of the justice inherent in them. Documentary analysis was

supplemented by interviews with key informants, including politicians, civil

servants, and representatives of professional organizations, volunteer organ-

izations, and pressure groups. A postal survey was administered to (local)

education authorities in England and Scotland to elucidate variations in the

ways in which children are assessed, “statemented” (in England) or “recorded”

(in Scotland), and outcomes were investigated through the secondary

analysis of official statistics. However, the main thrust of the research

consisted of a more detailed exploration of assessment at a local level. We

examined the roles of the key players, the extent to which parental prefer-

ences were congruent with professional identifications of need and official

Adler ADMINISTRATIVE JUSTICE 341

© Blackwell Publishing Ltd. 2003

determinations of policy, and the ways in which outcomes were negotiated.

The postal survey of (local) education authorities in England and Scotland

enabled us to identify variations in process north and south of the border

and provided a basis for selecting four contrasting (local) education author-

ities (two in England and two in Scotland) for in-depth fieldwork. In each

of these authorities, sixteen case study pupils with a range of special needs

were selected. The sample of sixty-four children included contested as well as

noncontested cases.23 Case papers for these children were analyzed, meetings

were observed, and interviews were conducted with most of the key actors

(i.e., educational psychologists, education officers, head teachers, SENCOs,24

class teachers, medical practitioners, social workers, “named persons,” parents,

and, where appropriate, children).

The research adopted an analytic framework that can be represented as

a “half way house” between those set out in Tables 2 and 3 above in that

it comprised four models of administrative justice – it incorporated the

consumerist but not the managerialist or market models of administrative

justice.25 The interviews with expert informants and the postal survey revealed

a wide variety of practices between and within the two countries. In Scotland,

it was clear that professionalism, strongly supported by bureaucracy, was the

dominant configuration. The legal model was very weak in that the assess-

ment and recording process was not seriously constrained by legal norms,

while the consumerist model, as manifested by parental participation, was

only marginally stronger.26 In England, the dominance of the professional

model was more effectively challenged by the bureaucratic, legal, and con-

sumerist models, all of which were stronger than in Scotland. The Code of

Guidance not only embodies bureaucratic norms but is legally binding; its

provisions confer strong procedural as well as substantive rights on parents.

The legal model was strengthened by the establishment of the Special Educa-

tional Needs Tribunal (SENT), which can hear appeals against a wide range

of local education authority decisions,27 while the consumerist model was

given a boost by the establishment of Parent Partnership Officers whose

existence enhances parental participation. There were no equivalents of the

SENT or Parent Partnership Officers in Scotland.

The in-depth fieldwork in the four contrasting education authorities yielded

some very rich accounts of the ways in which different trade-offs between the

competing normative models of decision making structured the relationships

between parents, professionals, and officials in different ways (some examples

can be found in Riddell, et al. 2002). To the extent that administrative justice

results from a trade-off between the features of competing normative models

of decision making, the position in England is clearly more complex than in

Scotland because more models are currently in play. In considering the pos-

sibility of alternative trade-offs, it is evident that Scotland might learn from

England by strengthening those models that would shift the balance of power

towards parents and children, who are currently in a rather weak position.

For example, a more robust appeals system might be introduced; Records of

342 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

Needs might specify the resources to be provided; assessment procedures,

including timescales, might be tightened; the provision of advocacy and mech-

anisms for enhancing parental participation might be strengthened; and

public access to information improved.28 Conversely, England might learn

from Scotland by seeking to find fairer ways of allocating resources and

better ways of balancing individual preferences and collective policy goals.

D. THE COMPUTERIZATION OF SOCIAL SECURITY IN THIRTEEN COUNTRIES

The most recent of the four pieces of research described in this article is a

comparative study of the use of computer technology in social security in

thirteen OECD countries, comprising ten Western European countries, Aus-

tralia, Canada, and the United States. Data were generated by two expert

informants in each country by means of electronic mail.29 They were asked to

complete a structured questionnaire, and, if their responses were unclear or

incomplete, they were asked to provide supplementary information. Preliminary

findings were distributed electronically, and informants were invited to correct

and comment on them. In this way, we were able to obtain valuable feedback

and to check the accuracy of our findings and the validity of our conclusions.

One of the aims of the study was to assess the impact of computerization

on administrative justice in a systematic manner. Two indicators have were

selected for each of the six models included in Table 3, and respondents

were asked to rate them on a 1–5 scale (where 1 = generally very important;

2 = generally important; 3 = important in some areas; 4 = not very important;

5 = unimportant). They were then asked whether computerization had made

each of them more or less important. A 1–5 scale was used here, too (where

1 = greatly increased importance; 2 = increased importance; 3 = much

the same; 4 = decreased importance; 5 = greatly decreased importance). The

scores for the two expert informants from each country were averaged. The

twelve indicators are listed in Table 7 below.

The use of two expert informants for each country was intended to pro-

vide a check on the accuracy of the data generated by the study. Neverthe-

less, doubts concerning its validity and reliability will still be raised. However,

early results (Adler & Henman 2001), based on an analysis of data from ten

countries, indicated that, with one exception, bureaucracy was the dominant

model of administrative justice in social security in the countries included in

our study, and that computerization had reinforced its dominance. Com-

puterization appears to have had a very significant effect in promoting the

managerial model of administrative justice, which in many countries is now

the second most important model. In contrast to this, there is little evidence

of the professional model, and there appears to have been a tendency for

computerization to reduce its importance. There is likewise little evidence of

the market model and computerization has had a minimal effect so far on

the ability of users to choose their preferred service provider – the two excep-

tions here are Belgium and Finland where employers and employees can

Adler ADMINISTRATIVE JUSTICE 343

© Blackwell Publishing Ltd. 2003

Table 7. A Framework for Assessing the Impact of Computerization on ProceduralFairness in Social Security

Indicator Model Importance Impact

Rating (on Rating (on

a 1–5 scale) a 1–5 scale)

In making decisions about entitlement to

benefit, social security institutions apply

well-established rules

Dissatisfied customers can complain to a

professional body

The government purchases social security

services from nongovernmental

service providers

Social security institutions are expected

to meet performance targets

Social security institutions are expected

to abide by customer charters

Claimants can check and correct

personal records

Dissatisfied customers have their

cases reviewed internally

Indicators are used to assess staff

performance

In making decisions to benefit entitlement,

staff exercise administrative discretion

Claimants can choose between more than

one social security institutions

Claimants participate actively in decision

making

Dissatisfied customers appeal to an

independent court or tribunal

choose which social security fund should provide statutorily defined social

security benefits.30 While the importance of the legal model of administrative

justice appears to be stronger in some respects than in others, computeriza-

tion appears to have had a minimal effect on it. Finally, the importance of

consumerism differs from country to country, and computerization has had

a mixed response on it.31

Among these broad shifts, computerization appears to have had little effect

on the availability of appeal procedures but has had a greater effect on the

procedures for determining entitlement to benefit. In particular, the assessment

of entitlement has become increasingly automated, involving an increased

reliance on rules and a correspondingly decreased use of discretion. This has

been supported by an increased reliance on the managerial model, in particu-

lar on performance measures, to bring about improvements in the delivery of

Bureaucracy

Professionalism

Markets

Managerialism

Consumerism

Legality

Bureaucracy

Managerialism

Professionalism

Markets

Consumerism

Legality

344 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

social security benefits. However, there is little evidence that this “top-down”

type of accountability is being matched by an increased emphasis on the legal

and consumerist models of administrative justice that embody “bottom up”

orientations.

The validity of these conclusions assumes both that the indicators of each

of the six models of administrative justice are appropriate ones and that the

assessments made by our expert informants are accurate. However, it is already

clear that computerization has had an impact on the trade-offs that are made

between the six different models of administrative justice outlined above in

that it alters the ways in which decisions are made, the ways in which they can

be challenged, and the ways in which individuals are treated by social security

institutions. Although many similarities were observed among the countries

in the study, some differences between these countries were also apparent.

IV. CONCLUSION

By adopting a relativistic orientation to administrative justice, Mashaw’s

approach, and the approach adopted in all four of the studies outlined above,

both challenge the view that there are any invariant principles of administrat-

ive justice that apply in all contexts.32 This may, at first, seem surprising but,

on reflection, should not be since it is generally agreed that this is true of

social or distributive justice. Administrative justice is no less a contested

concept than social justice in that, although it can be defined in a relatively

uncontroversial or uncontentious way (as “a proper balance between com-

peting claims to procedural protection”), the terms in which it is defined (i.e.,

what constitutes “a proper balance” and even what are to count as “claims”)

are the subject of considerable disagreement. Like social justice, it is an

“essentially contested concept” (Gallie 1964). Compared to the external focus

on mechanisms of redress, two of the great merits of an internal orientation are

that it focuses on the myriad of first-instance decisions rather than the much

smaller number of decisions that are the subject of an appeal or complaint

and that it analyzes them directly rather than at one remove and through a

“legal prism.” This is not to deny the important contribution that external

redress mechanisms undoubtedly make to the promotion of administrative

justice. The point is that external modes of redress promote particular concep-

tions of administrative justice and may have a limited impact on the justice

inherent in administrative decision-making. They need to be combined with

the internal orientation that has been outlined and illustrated in this article.

In terms of the framework for analyzing administrative justice, the attempt

to assess the impact of computerization on social security in the UK in terms

of the three normative models identified by Mashaw was undoubtedly the

simplest. By contrast, the study of decision making in the Scottish prison

system was perhaps the most ambitious. With its focus on the discourses of

procedural fairness and substantive justice and on the discursive struggles

Adler ADMINISTRATIVE JUSTICE 345

© Blackwell Publishing Ltd. 2003

between the individuals and groups who were the carriers of these discourses,

it was able to give a dynamic account of the dramatic changes that were

taking place in the Scottish prison system at the time. Although this study

anticipated the rise of managerialism as another normative model of adminis-

trative justice, it was only later that this was fully elaborated. The study of

the assessment of special educational needs in England and Scotland utilized

a framework based on four normative models of administrative justice and

yielded a rich account of the different forms that procedural fairness can

take and the different responses to them of each of the parties concerned. As

noted above, it has been used to indicate how administrative justice could be

enhanced in both countries by some different trade-offs between the competing

models. An extended framework comprising six models was used in the com-

parative study of computerization in the social security systems of thirteen dif-

ferent countries and has enabled us to understand the impact of computers

on the trade-offs between the different normative models of administrative

justice. However, in this case, the advantages of a systematic approach will

inevitably be offset by concerns about the reliability and validity of the data.

Although all of the studies have their limitations, it is contended that,

considered together, they demonstrate the power of a particular approach to

studying administrative justice and the validity of the assumptions that underlie

it. It is also contended that they satisfy each of the three defining character-

istics of the socio-legal paradigm – they all adopt an external perspective to

legal process that seeks to analyze administrative justice in terms of concepts

and categories that are derived from the social sciences; they all focus on

routine, rather than leading, cases; and they are all informed by philosophical

analysis.33 Whether or not they are successful is for others to judge.

michael adler is Professor of Socio-Legal Studies in the School of Social and PoliticalStudies at Edinburgh University. He has recently completed a review of research on theexperiences, perceptions, and expectations of tribunal users for the Lord Chancellor’sDepartment and, with financial support from the Nuffield Foundation, is currently carry-ing out a developmental study of administrative grievances, exploring the problemspeople experience in dealing with government departments and other public bodies, theadvice they seek, whether or not they challenge the decisions in question and the impactof these problems on their lives. He hopes that this will lead to a national survey of theadministrative grievances people experience and the effectiveness of the available mech-anisms for challenging decisions that are experienced as unjust or unfair.

NOTES

1. A similar argument applies in respect to health care — preventative medicineaims to improve levels of health in the community while health-care systems areset up to treat episodes of illness that individual members of the communityexperience. One is not more important than the other, and there is a strong casefor combining the two.

2. The rule against bias asserts that decisions must be made by an impartial judgeand may be set aside where there are grounds for reasonable suspicion of bias;

346 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

the right to a hearing requires that each party should have the opportunity ofknowing the case against him (or her) and stating his (or her) own case.

3. For a full account of “red light” and “green light” theories, see Harlow andRawlings (1997:29–66, 67–90).

4. The further examples were rudeness (though that is a matter of degree); unwill-ingness to treat the claimant as a person with rights; refusal to answer reasonablequestions; neglecting to inform a complainant on request of his or her rights orentitlement; knowingly giving advice that is misleading or inadequate; ignoringvalid advice or overruling considerations that would produce an uncomfortableresult for the over-ruler; offering no redress or manifestly disproportionateredress; showing bias, whether because of color, sex, or any other grounds;omission to notify those who thereby lose a right of appeal; faulty procedures;failure by management to monitor compliance with adequate procedures; cavalierdisregard of guidance that is intended to be followed in the interest of equitabletreatment of those who use a service; partiality; and failure to mitigate the effectsof rigid adherence to the letter of the law where that produces manifestly unequaltreatment.

5. Under the Scottish Public Services Ombudsman Act 2002, a single ScottishPublic Services Ombudsman has now taken over the work formerly carried out byfour separate ombudsmen: the Scottish Parliamentary Ombudsman, the ScottishHealth Services Ombudsman, the Scottish Local Government Ombudsman, andthe Housing Association Ombudsman for Scotland.

6. For a very comprehensive account of public sector ombudsmen in the UK, seeSeneviratne (2002). For a more general account, see Harlow and Rawlings(1997:391–422, 423–55).

7. Remedial action involves correcting the injustice caused by the maladministrationand reversing the decision made because of it; compensation can cover direct andindirect losses resulting from maladministration; and an apology would normallycome from the head of the organization complained of. Note that remedialaction is not always feasible while compensation may be appropriate when remed-ial action is not. However, where the losses caused by maladministration arepsychological rather than material, it may be difficult to put a monetary value onthem. Sometimes an apology may be all that the complainant wants.

8. A recent British study of regulatory bodies in government (Hood et al. 1999:60–65) indicates that, compared to other types of regulator, a relatively small pro-portion of the staff working for the PCA, the Health Service Commissioner, andthe CLA have experience of employment in “regulatee” organizations (ibid.:Table 3.2). Using terminology developed by Black (1975), their “relational dis-tance” (RD) was placed in the lowest of three categories. The RD of HousingAssociation Ombudsmen was placed in the middle category while that of thePrisons Ombudsman was placed in the highest category. However, it shouldbe noted that the measures of RD used in this study are relative to the otherregulators in the study and are not absolute measures.

9. As Halliday (2001) reminds us, in addition to its impact on the individual whopetitions the court, judicial review can have an impact on government policy orlegislation as well as on routine decision making within government. This appliesequally to other external forms of accountability like tribunals and ombudsmen.For a more general discussion of the impact of judicial decisions on publicadministration, which highlights the lack of empirical evidence, see Richardsonand Sunkin (1996). For an up-to-date discussion, based on an empirical study ofthe impact of judicial review on three local government homeless persons units,see Halliday (2003).

10. The few empirical studies that have been carried out indicate that judicial reviewhas a limited impact on administrative decision-making in the United Kingdom.

Adler ADMINISTRATIVE JUSTICE 347

© Blackwell Publishing Ltd. 2003

See, e.g., Mullen, Pick and Prosser (1996:113–34) and Halliday (2000). Likewise,research indicates that tribunals also have a limited effect on first-instance deci-sion makers. In a study of decision making in social security (Baldwin, Wikeleyand Young 1992), just over half (52.6 percent) of adjudication officers claimedthat, in making first-instance decisions, they were not at all influenced by atribunal’s likely response to an appeal. This compares to a quarter (25.0 percent)of officers who claimed that tribunals had a procedural effect in that the pro-spect of an appeal led them to be more thorough and document their decisionsmore fully.

11. There is likewise considerable variation between countries with different legalsystems. Generally speaking, the impact of higher courts is greater than that oflower courts, and the impact of courts is greater than that of tribunals. It is moredifficult to generalize about ombudsmen since there is so much variation betweencountries in their scope and standing.

12. Note that the third model is characterized in terms of legality rather than fair-ness. For a discussion of legality, which is a synonym for “the rule of law,” seeSelznick (1980:11–18).

13. Agreement on the principle of justice that should determine how the goods inquestion are distributed may not be the end of the matter since there may beconsiderable disagreement about how to operationalize the principle in question.For a full discussion, see Elster (1992).

14. See, e.g., the reviews of Mashaw (1983) by Boyer (1984) and Maranville (1984).15. There are, of course, many views about what different sets of trade-offs between

competing models of policy and competing models of administration might bemore desirable. These are associated with different political ideologies and reflectdifferent conceptions of what would be in the public interest. In ascertainingwhich trade-offs might be more desirable, Rothstein (1999) points out that it isimportant to ensure that those who are affected should regard the policy as justand its implementation as fair. However, what is desirable is not necessarilyfeasible. The trade-offs that are made in practice reflect the power and theinterests of different groups of institutional actors who are attached to, sponsor,and promote the different models of administrative justice.

16. Like other methods of distribution, making decisions by lot can be used todistribute burdens as well as benefits. In the first case, we frequently refer tosomeone “drawing the short straw”; in the second case, to someone “having thewinning ticket” or “having the winning number.” For a discussion of the advant-ages and disadvantages of lotteries, see Duxbury (1999).

17. For an illuminating discussion of the ethics of markets, see Sen (1985).18. Unless, however, it is argued that the initial distribution of resources accurately

reflects preexisting property rights or that the final distribution reflects the pro-ducers’ right to enjoy what they produce. The best account of the argumentbased on prior entitlements can be found in Nozick (1974) and of the argumentbased on producers’ rights in Bauer (1981). Both arguments are very effectivelycriticized in Sen (1985).

19. Unlike perfect procedural justice, where there is an independent criterion fordetermining how benefits and burdens should be distributed and a procedure thatis certain to have this result can be specified, and imperfect procedural justice,which has only the first of these features (there is an independent criterion fordetermining what the outcome should be but no way of specifying a procedurethat will always produce this outcome), pure procedural justice has neither ofthese features. However, although there is no independent criterion for deter-mining what the outcome should be, there is a correct or fair procedure andthe consequences of applying it, whatever they may be, should be regarded asjust, provided that the procedure has been properly followed. In this kind of

348 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

procedural justice, the justice of any outcome is founded on the fairness of theinstitutional arrangements from which it arises. For a fuller discussion of “perfect,”“imperfect,” and “pure” forms of procedural justice, see Rawls (1971:85–86).Rawls makes it clear (ibid.:270–74) that markets exemplify pure procedural justice,and it is clear that lotteries do as well.

20. These included the initial allocation of prisoners to establishments, transfersbetween establishments, security categorizations, the allocation of work and educa-tional placements, the distribution of privileges, the handling of requests andgrievances, appeals to the domestic courts, the PCA and the ECHR, and theactivities of the Prisons Inspectorate.

21. In 1999, the government circulated a number of proposals in a consultationdocument that led to a revised code that came into effect in the school year2000–2001.

22. The percentage of pupils who are statemented in England (2.9%) is higher thanthe percentage of pupils who are recorded in Scotland (1.9%), but there are widevariations between authorities in both countries. Among pupils who arestatemented or recorded, a higher proportion of pupils attend mainstream schoolsin England (58%) than in Scotland (37%). In Scotland, some special schoolpupils do not have records of needs, although all special school pupils in Eng-land are statemented. In spite of this, the percentage of the age group who attendspecial schools is higher in England (1.2%) than in Scotland (1.0%). The figuresrelate to the years 1996, 1997, or 1998, and the comparisons are therefore notstrictly correct.

23. We were unable to achieve the 50:50 split between contested and noncontestedcases that we had hoped to. This was because there were very few contested casesin two of the authorities. Most of the “contested” cases were resolved beforethey reached the point of legal challenge.

24. In England, Special Educational Needs Co-ordinators implement the school-based stages of assessment and coordinate reviews for all children with specialeducational needs.

25. Unfortunately, the fieldwork and the development of our analytic frameworkdid not go hand in hand. In the early stages of our research, we thought ofmanagerialism as an advanced form of bureaucracy rather than as a sui generismode of decision making. Moreover, after considerable debate, we concludedthat the market could be best understood as an exogenous variable, i.e. as anaspect of the external environment in which administrative decision-making takesplace, rather than as a form of decision making in its own right. Although wechanged our position on these two issues in the course of the project, this changeoccurred too late for it to have any major effect on our research design, on theconduct of our research or on our research findings. However, we were ratherquicker to appreciate the distinction between consumerism, manifested in termsof user participation in decision making, and legalism, expressed in terms ofchallenges to official decisions and/or the manner in which they are reached.Thus, the analytic framework that informed the study incorporated the con-sumerist but not the managerialist or market models of administrative justice.

26. Despite the fact that parents were actively deterred from adopting the role of aconsumer, there were fears among some local authorities that the process wasbecoming too adversarial, prompting an attempt by the Association of PrincipalEducational Psychologists to persuade the Scottish Executive to abolish the pro-cess of recording.

27. The SENT was set up in 1994 and heard its first cases in 1995. Parents have theright to appeal to the SENT in relation to: an LEA’s decision to make and main-tain a Statement; an LEA’s decision to make a statutory assessment; the schoolnamed on the Statement; the assessment of the child’s SEN; the measures proposed

Adler ADMINISTRATIVE JUSTICE 349

© Blackwell Publishing Ltd. 2003

by the LEA to meet the child’s SEN. For a very thorough evaluation of theSENT, see Harris 1997. There is, as yet, no equivalent of the SENT in Scotland.

28. Many of these proposals would, no doubt, be strongly opposed by the Associ-ation of Principal Educational Psychologists. See note 25 above.

29. At the outset, we assumed that there would be a number of advantages to thisprocedure. We thought that it would constitute an efficient method of datacollection and an effective means of understanding the detailed operation ofpolicies and procedures in different countries, i.e., that it would reduce thedanger of misunderstanding the situation in a particular country, overcome lan-guage barriers, and provide a useful source of informed advice on research designand the interpretation and analysis of the information provided. In practice, itsometime proved to be extremely difficult to identify informants with the necessaryexpertise and persuade them to take part, and we received fewer comments onour preliminary findings than we had expected.

30. The main effect of computerization on the market model appears to have beento increase the extent to which governments purchase the delivery of social secur-ity from nongovernmental organizations.

31. An analysis of the data from all thirteen countries in the study confirms thesefindings (see Adler & Henman forthcoming).

32. Some followers of Mashaw, e.g., Sainsbury (1992), have attempted to develop aless relativistic conception of “administrative justice.” Adopting the perspectiveof the individual citizen, Sainsbury argues that administrative justice, definedas those “qualities an administrative decision ought to exhibit, which providearguments for the acceptability of its decisions,” has two invariant components.These are, first, accuracy and, second, fairness, the latter comprising promptness,impartiality, participation, and accountability. According to Habermas (1992),participation is a means of advancing rational discourse and, as such, plays animportant role in legitimating the rule of law and the role of the state.

33. See MacCormick (1994). A similar characterization can be found in the Economicand Social Research Council’s (ESRC) review of socio-legal studies (ERSC 1994).

REFERENCES

adler, michael, and paul henman (2001) “e-Justice: A Comparative Study ofComputerisation and Procedural Justice in Social Security,” International Reviewof Law, Computers & Technology 15(2):195–212.

adler, michael, and paul henman (forthcoming) The Computerisation of SocialSecurity in Thirteen OECD Countries. London: IBM.

adler, michael, and brian longhurst (1994) Discourse, Power and Justice: Towardsa New Sociology of Imprisonment. London: Routledge.

adler, michael, and roy sainsbury (1990) Putting the Whole Person Concept intoPractice: Final Report (Parts I and II). Edinburgh: Department of Social Policy,Univ. of Edinburgh.

adler, michael, and roy sainsbury (1991a) “Administrative Justice, Quality of Serviceand the Operational Strategy.” In The Social Implications of the Operational Strat-egy, edited by M. Adler & R. Williams. Edinburgh: Dept. of Social Policy andSocial Work, Univ. of Edinburgh.

adler, michael, and roy sainsbury (1991b) “The Social Shaping of InformationTechnology: Computerisation and the Administration of Social Security.” In TheSociology of Social Security, edited by M. Adler, C. Bell, J. Clasen & A. Sinfield.Edinburgh: Edinburgh Univ. Press.

adler, michael, and roy sainsbury (1996) “Alternative Approaches to the Com-puterisation of Social Security: Reflections on a Delphi Exercise.” In Gazing into the

350 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

Oracle: The Delphi Method and its Application to Social Policy and Public Health,edited by M. Adler & E. Ziglio. London: Jessica Kingsley.

baldwin, john, nick wikeley, and richard young (1992) Judging Social Security:The Adjudication of Claims for Benefit in Britain. Oxford: Clarendon Press.

bauer, p. t. (1981) Equality, The Third World, and Economic Delusion. Cambridge,Mass.: Harvard Univ. Press.

bayles, michael (1990) Procedural Justice: Allocating to Individuals. Dordrecht: KluwerAcademic Publications.

black, donald j. (1975) Behavior and the Law. New York: Academic Press.boyer, barry (1984) “From Discretionary to Bureaucratic Justice,” Michigan Law

Review 82:971–80.clarke, john, and janet newman (1997) The Managerial State: Power, Politics and

Ideology in the Remaking of Social Welfare. London: Sage.duxbury, neil (1999) Random Justice: On Lotteries and Legal Decision-Making.

Oxford: Oxford Univ. Press.economic and social research council (ersc) (1994) Review of Socio-Legal Studies:

Final Report. Swindon: ESRC.elster, jon (1992) Local Justice: How Institutions Allocate Scarce Goods and Necessary

Burdens. Cambridge: Cambridge Univ. Press.fuller, lon l. (1978) “The Forms and Limits of Adjudication,” Harvard Law Review

92:353–409.gallie, w. b. (1964) “Essentially Contested Concepts.” In Philosophy and the Historical

Understanding, by W. B. Gallie. London: Chatto & Windus.galligan, d. j. (1996) Due Process and Fair Procedures: A Study of Administrative

Procedures. Oxford: Clarendon Press.great britain (1991) The Citizen’s Charter: Raising the Standard. Cm. 1599.

London: HMSO.great britain. council on tribunals (1993) Annual Report of the Council on Tribunals

for the Period. London: HMSO.great britain. department for education (1994) Code of Practice on the Identifica-

tion and Assessment of Special Educational Needs. London: HMSO.great britain. department of health and social security (1982) Social Security

Operational Strategy: A Framework for the Future. London: DHSS.great britain. parliament (1932) Report of the Donoughmore Committee on Minis-

ters Powers. Cmnd. 4050. London: HMSO.great britain. parliament (1957) Report of the Franks Committee on Tribunals and

Inquiries. Cmnd. 218. London: HMSO.great britain. parliament. house of commons (1966) Parliamentary Debates, 5th

ser., vol. 734. London: HMSO.habermas, jürgen (1992) Between Facts and Norms: Contributions to a Dis-

course Theory of Law and Democracy. Trans. by W. Rehg. Cambridge: PolityPress.

halliday, simon (2000) “The Influence of Judicial Review on Bureaucratic Decision-Making,” Public Law (spring):110–22.

halliday, simon (2001) “Ideal Conditions of Impact: Some Reflections on the Influ-ence of Judicial Review on Administrative Decision Making.” Paper presented tothe Law Program, Research School of the Social Sciences, Australian NationalUniversity, 5 September, Canberra.

halliday, simon (2003) Judicial Review and Compliance with Administrative Law.Oxford: Hart.

harlow, carol, and richard rawlings (1997) Law and Administration. 2d ed.London: Butterworths.

harris, neville (1997) Special Educational Needs and Access to Justice: The Role ofthe Special Educational Needs Tribunal. Bristol: Jordans.

Adler ADMINISTRATIVE JUSTICE 351

© Blackwell Publishing Ltd. 2003

himsworth, chris (ed.) (1985) Judicial Teeth for Local Ombudsmen? Edinburgh:Department of Public Law, University of Edinburgh.

hirschman, albert o. (1970) Exit, Voice, and Loyalty: Responses to Decline in Firms,Organizations and States. Cambridge, Mass.: Harvard Univ. Press.

hood, christopher (1991) “A New Public Management for all Seasons,” PublicAdministration 69:3–19.

hood, christopher (1998) The Art of the State: Culture, Rhetoric and Public Manage-ment. Oxford: Clarendon Press.

hood, christopher, colin scott, oliver james, george jones, and tony travers

(1999) Regulation inside Government: Waste Watchers, Quality Police, and Sleaze-Busters. Oxford: Oxford Univ. Press.

ison, t. g. (1999) “ ‘Administrative Justice’: Is it Such a Good Idea?” In AdministrativeJustice in the 21st Century, edited by M. Harris & M. Partington. Oxford: Hart.

johnson, norman (ed.) (1995) Private Markets in Health and Welfare: An InternationalPerspective. Oxford: Berg.

le grand, julian (1991) “Quasi-Markets and Social Policy,” Economic Journal101:1256–67.

le grand, julian, and will bartlett (eds.) (1993) Quasi-Markets and Social Policy.Basingstoke, UK: Macmillan.

lipsky, michael (1980) Street-Level Bureaucracy: Dilemmas of the Individual in PublicServices. New York: Russell Sage Foundation.

maccormick, neil (1994) “Four Quadrants of Jurisprudence.” In Prescriptive Form-ality and Normative Rationality in Modern Systems: Festschrift for R. S. Summers,edited by W. Krawietz, G. H. von Wright & D. N. MacCormick. Berlin: Duncker& Humboldt.

maranville, deborah (1984) “Review of J. L. Mashaw: Bureaucratic Justice,”Minnesota Law Review 69:325–47.

mashaw, jerry l. (1974) “The Management Side of Due Process: Some Theoreticaland Litigation Notes on the Assurance of Accuracy, Fairness and Timeliness in theAdjudication of Social Welfare Claims,” Cornell Law Review 59:772–824.

mashaw, jerry l. (1983) Bureaucratic Justice: Managing Social Security DisabilityClaims. New Haven, Conn.: Yale Univ. Press.

miller, david (1976) Social Justice. Oxford: Clarendon Press.mullan, d. j. (1975) “Fairness: The New Natural Justice,” University of Toronto Law

Journal 25:281–316.mullen, tom, kathy pick, and tony prosser (1996) Judicial Review in Scotland.

Chichester: J. Wiley.nozick, robert (1974) Anarchy, State, and Utopia. Oxford: Basil Blackwell.page, alan (1999) “The Citizen’s Charter and Administrative Justice.” In Administrat-

ive Justice in the 21st Century, edited by M. Harris & M. Partington. Oxford: Hart.rawls, john (1971) A Theory of Justice. Oxford: Clarendon Press.richardson, genevra, and maurice sunkin (1996) “Judicial Review: Questions of

Impact,” Public Law (spring):79–103.riddell, sheila, michael adler, enid mordaunt, and nadia farmakopoulou (2000)

“Special Educational Needs and Competing Policy Frameworks in England andScotland,” Journal of Education Policy 15(6):621–35.

riddell, sheila, alastair wilson, michael adler, and enid mordaunt (2002)“Parents, Professionals and Special Educational Needs Policy Frameworks in Eng-land and Scotland,” Policy and Politics 30(3):411–25.

rothstein, bo (1999) Just Institutions Matter: The Moral and Political Logic of theUniversal Welfare State. Cambridge: Cambridge Univ. Press.

sainsbury, roy (1992) “Administrative Justice: Discretion and Procedure in SocialSecurity Decision-Making.” In The Uses of Discretion, edited by K. Hawkins.Oxford: Clarendon Press.

352 LAW & POLICY October 2003

© Blackwell Publishing Ltd. 2003

scottish office education and industry department (1998) A Manual of GoodPractice in Special Educational Needs. Edinburgh: HMSO.

selznick, philip (1980) Law, Society, and Industrial Justice. New Brunswick, N.J.:Transaction Books.

sen, amartya (1985) “The Moral Standing of the Market.” In Ethics and Economics,edited by E. F. Paul, J. Paul & F. D. Miller, Jr. Oxford: Basil Blackwell.

seneviratne, mary (2002) Ombudsmen: Public Services and Administrative Justice.London: Butterworths LexisNexis.

walzer, michael (1983) Spheres of Justice: A Defense of Pluralism and Equality.Oxford: Basil Blackwell.