Undercutting Justice – Why legal representation should not be allocated by the market

Abstract

The adversarial legal system is traditionally praised for its normative appeal: it protects individual rights; ensures an equal, impartial, and consistent application of the law; and, most importantly, its competitive structure facilitates the discovery of truth – both in terms of the facts, and in terms of the correct interpretation of the law. At the same time, legal representation is allocated as a commodity, bought and sold in the market: the more one pays, the better legal representation one gets. In this article, I argue that the integration of a market in legal representation with the adversarial system undercuts the very normative justifications on which the system is based. Furthermore, I argue that there are two implicit conditions, which are currently unmet, but are required for the standard justifications to hold: that there is (equal opportunity for) equality of legal representation between parties, and that each party has (equal opportunity for) a sufficient level of legal representation. I, therefore, outline an ideal proposal for reform that would satisfy these conditions.

Full text

Article
Undercutting Justice –
Why legal representation
should not be allocated by
the market
Shai Agmon
University of Oxford, UK
Abstract
The adversarial legal system is traditionally praised for its normative appeal: it protects
individual rights; ensures an equal, impartial, and consistent application of the law; and,
most importantly, its competitive structure facilitates the discovery of truth – both in
terms of the facts, and in terms of the correct interpretation of the law. At the same
time, legal representation is allocated as a commodity, bought and sold in the market: the
more one pays, the better legal representation one gets. In this article, I argue that the
integration of a market in legal representation with the adversarial system undercuts
the very normative justifications on which the system is based. Furthermore, I argue that
there are two implicit conditions, which are currently unmet, but are required for the
standard justifications to hold: that there is (equal opportunity for) equality of legal
representation between parties, and that each party has (equal opportunity for) a suf-
ficient level of legal representation. I, therefore, outline an ideal proposal for reform that
would satisfy these conditions.
Keywords
adversarial legal system, legal justice, lawyers, legal representation, marketisation,
commodification, levelling down objection, moral limits of markets, market in legal
representation
Corresponding author:
Shai Agmon, Department of Politics and International Relations, University of Oxford, Oxford OX1 3UQ, UK.
Email: shai.agmon@politics.ox.ac.uk
Politics, Philosophy & Economics
1–25
ªThe Author(s) 2020
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1470594X20951886
journals.sagepub.com/home/ppe

There can be no equal justice where the kind of trial a man gets depends on the amount of
money he has.
Justice Hugo Black, 1956 Griffin v. Illinois, 1956.
Introduction
Legal Justice should not be for sale. Although most people find this uncontroversial,
almost no one challenges the fact that legal representation (namely, lawyers and legal
resources required for representation) – one of the most significant components of
contemporary adversarial legal systems – is traded on the free market. This article offers
a normative analysis of this discrepancy.
The Anglo-American adversarial method of adjudication is often defended as norma-
tively appealing. First and foremost, this is for the supposed fact that its competitive
structure facilitates the discovery of truth, both in terms of the facts, and in terms of the
correct interpretation of the law. This, in turn, supposedly secures a high probability of
delivering just results (Wertheimer, 1988: 309–310).
1
Second, it is seen as a method that
protects individual rights and ensures an equal, impartial and consistent application of
the law. The method is characterised by legal competition that consists, roughly, of the
following three components: (a) a passive impartial tribunal (normally a judge and/or
jury); (b) formal rules of procedure; and (c) two or more competing parties that are
responsible for presenting their own case as best as they can and for opposing their
adversaries (Assy, 2015: 11–12; Gerber, 1987: 4–5; Luban, 1988: 57). As modern
adversarial legal systems have become increasingly complex and professionalised,
nowadays one cannot normally make one’s case in court effectively without professional
legal representation (Assy, 2015: 10; Fuller and Winston, 1978: 383; Luban, 1985: 46).
Thus, ideally, component (c) of the adversarial method – the preparation and presenta-
tion of the legal argument –is performed by lawyers who advocate for their client with
one-sided partisan zeal, with the sole aim of their client winning.
2
The influence of legal representation on the adjudicative process intensifies in adver-
sarial systems. This is due to the institutionalised passivity of the tribunal, which leaves
control over significant parts of the process in the hands of the lawyer. In addition to
being in charge of litigation during the trial, the lawyer is also responsible for the entire
legal preparation process: from deciding on strategies for questioning witnesses, to
constructing legal arguments (to which the client is bound), to summoning experts (thus,
legal representation cannot be reduced only to the argumentative skills of the lawyers,
but also includes the access she has to quality legal resources) (Luban, 1988: 57–58;
Schwartz, 1983: 546).
3
The impartial tribunal only receives pre-processed legal material;
its role is limited to assessing both sides’ arguments, and in so doing trying to extract the
truth of the matter.
Thus, the quality of legal representation influences, in theory and practice, courts’
decisions: the better the lawyer is, the greater the chance that the court favours her
client’s case. Therefore, allocating legal representation by the market – i.e. the more
2Politics, Philosophy & Economics XX(X)

one pays, the better-quality legal representation one gets – starts to look controversial, as
the rich enjoy a structural advantage (see Hadfield, 2000: 956).
4
To be sure, the potential influence of economic inequalities on the legal system has
not been ignored. Legal scholars have long been concerned with the problem of ‘access
to justice’ – people’s inability to afford access to the court system (for example see
Rhode, 2004). Other worries have arisen in relation to the ethical dilemmas deriving
from the commercialisation of the legal profession (see Luban, 1988: 177–237; Marko-
vits, 2008; Wendel, 2010). However, this work has mainly focused on remedies designed
to minimise the problematic results which are caused by the influence of economic
inequalities within the existing structure of adversarial systems (e.g. drafting an ethical
code for lawyers, supporting free legal clinics, etc.). Almost all critics have refrained
from challenging the fundamental structure of the whole system; that the legal repre-
sentation is allocated by a market mechanism.
5
This attribute has been taken as a given,
and has therefore rarely faced normative scrutiny.
6
This absence is particularly striking in light of the wide agreement that other legal
functions, like prosecution and the judiciary, should not be marketised. Even staunch
defenders of the market (e.g. Friedman, 2002: 25; Hayek, 1960: 229; Nozick, 1974: 14–
15; Rand, 1964: 131 and others) do not suggest extending it to the judicial system. In
fact, libertarians usually regard adjudication, including legal representation (although
not explicitly oftentimes), as the quintessential role of the state.
In this paper, I attempt this scrutiny, and argue that the integration of a market in legal
representation into the adversarial system undercuts the very normative justifications on
which the system is based.
7
The adversarial system has been justified in a way that
assumes the absence of a market in legal representation. But the undeniable existence of
such a market undermines the only merits that made the system normatively appealing in
the first place. Furthermore, I argue that this undercutting reveals two implicit necessary
conditions for the standard justifications to hold (which are currently unmet): that there
should be (equal opportunity for) equality of legal representation between parties, and
that each party should have (equal opportunity for) a sufficient level of legal represen-
tation. I therefore outline an ideal reform to the system that would meet these
conditions.
8
The paper is structured as follows. First, I characterise the legal representation mar-
ket. Second, I present the three main normative justifications for the adversarial system,
which together make up the ‘normative appeal’ of the system. Third, I argue that the
legal representation market undercuts this normative appeal, in both the adversarial civil
and criminal systems. Fourth, I show how this undercutting reveals the unstated require-
ments of equality and sufficiency. In this part, I also present my proposal for reform and
reply to some objections to it. Finally, I conclude.
Part I: The legal representation market in adversarial legal
systems
Before presenting my argument, a few preliminary comments are required to clarify
what I mean by the ‘legal representation market’ or ‘market mechanism’. In this paper,
these terms denote only three of the market’s typical characteristics: allocation according
Agmon 3

to agents’ ability and willingness to buy and sell, and the innate tendency to produce
economic inequalities. Most importantly, the market is indifferent to its distributional
results: its sole concern is procedural, i.e. that all transactions are voluntary and non-
fraudulent (Nozick, 1974: 163–164; Satz, 2010: 17–21).
9
Since legal representation in the adversarial system is bought on the market, con-
sumers (clients) can use their money to obtain the best legal representation they can
afford. Producers (lawyers), on their part, can set their fees without being limited by
anything other than market forces (though public defenders and prosecutors, whom I
discuss later, are an exception to this rule). Unless some other normative criterion is
taken into consideration, producers generally aim to receive as much money as possible
from consumers. Consequently, the rich usually secure high-quality representation,
while the middle class and the poor, if they are able to hire a lawyer at all, can only
afford representation of lower quality. By ‘quality of representation’ I mean the profes-
sional level of lawyers, as well as to the quality of legal resources available to them, such
as experts and private investigators. These cost a lot of money too, and are an integral
part of the lawyer’s work (see Galanter, 1974: 97–98, 114.).
10
These characteristics are not only present in the legal representation market qua
market; they are exacerbated by its unique attributes. First, the legal system is particu-
larly complex: lawyers undergo expensive training, the laws themselves are complex, the
results of adjudication processes are unpredictable, and clients have trouble in assessing
the scope of the work that has to be done by the lawyer. Additionally, the adversarial
method is competitive and the stakes are usually high for the parties at trial. Clients,
therefore, are willing to pay as much as they can to tilt the scales of justice in their favour.
The combination of professional complexity and exceptional willingness to pay results
in high fees, and in considerable disparities in the quality of representation received by
clients (Hadfield, 2000: 963–999; Moorhead, 2011: 349–350).
So the legal representation market produces an advantage for the wealthy in terms of
the quality of representation available to them. In a legal competition, like the one which
takes place in an adversarial system, such an advantage means that the privileged party
has a de facto better chance for making its case. This common-sense claim underpins the
main argument of this paper, and is supported by the empirical evidence on the results of
the adversarial system.
11
As further evidence of the disparities which arise from the adversarial system, various
institutions which have been developed in countries in which the system exist – attesting
to the fact that the legal system itself recognises, or at least reacts to, the difficulties that
arise. Examples of this phenomenon include class action, legal aid and contingent fees
mechanisms (regarding Class Actions see Conte et al., 2011: 9–21; Jones, 2003: 75). But
these institutions have a negligible corrective force. Even if they were perfectly effective
(a controversial assumption), they are either confined to a very specific field within the
law (e.g. class actions) or simply insufficient to compensate for the disparities in the
system as a whole (e.g. legal aid) (Emons, 2000: 21–33; Rhode, 2004: 60–62, 122–144;
Rosen-Zvi, 2010: 721–722). Consequently, such mechanisms are no more than testi-
mony to the existence of the structural problems of the system.
The arguments so far should not be taken as showing that people can buy justice
per se. If this were the case, there would be no difference (in effectiveness terms)
4Politics, Philosophy & Economics XX(X)

between hiring a lawyer and bribing a judge. In cases where there is sufficient evi-
dence and the legal interpretation is conclusive, money is unlikely to influence the
court’s decision. Thus, the scope of my argument is limited to cases in which legal
argument is required. Moreover, even when a lawyer can influence a court’s decision,
spending tremendous amounts of resources on legal representation is limited and does
not necessarily lead to winning a case – all my argument highlights, and depends on, is
the fact that money increases one’s probability of winning. The influence of money,
nonetheless, can manifest in ways other than winning, such as the reduction of the
compensation one has to pay when found responsible, or the softening of a sentence.
Hence, the phenomenon I identify in the adversarial system should not be considered as
bribery, but rather a significant structural advantage enjoyed by the wealthy.
Part II: The normative appeal of the adversarial system
In this section I present the three most convincing normative justifications for the
adversarial legal system.
12
Although I present the arguments as justifications for the
system as a whole, some differ in application between civil cases and criminal cases. I
elaborate on these differences in the following sections.
13
Providing just results
Using Rawls’s terminology (1999: 73–78), the adversarial method of adjudication is an
instance of imperfect procedural justice, which is comprised of two (related) compo-
nents: imperfect results and an independent criterion of justice. The former means that
even in cases where the procedure (e.g. the law) has been carefully followed and the
proceedings properly conducted, the decision-maker (e.g. the court) may still reach a
wrong conclusion. The latter is best understood in contradistinction to a dependent
criterion for a just result. With such a criterion, any outcome of the procedure, correctly
followed, is just. The procedure itself is the normative watershed. For example, a fair
coin toss is a system with a dependent criterion of justice, because whoever turns out to
be the winner, the result will be just. With an independent criterion of justice, the way to
evaluate whether a result is just or not being to employ a criterion of justice independent
from the procedure: in the legal system, roughly, a just result of the adjudicative process
should be judged against the criterion of the accurate interpretation of the law, in
accordance with the facts of the matter (Rawls, 1999: 73–78). This implies, for instance,
that those who are innocent should not be convicted.
The crucial question that arises is whether the adversarial method, as an imperfect
procedural system, provides mostly just results (Wertheimer, 1988: 309).
14
There are
two arguments in favour of the conclusion that it does: the ‘invisible hand’ argument;
and the ‘division of labour’ argument.
The ‘invisible hand’ argument. The basic narrative underlying ‘invisible hand’ arguments is
of a process in which the result of numerous individuals, acting in their narrow self-
interest, results in the emergence of an optimal overall design, unplanned as well as
unforeseen by them (Ullmann-Margalit, 1978: 270–272).
Agmon 5

Regarding the adversarial method, the invisible hand argument is as follows. The
individuals are the parties; each is incentivised to zealously fight for their own case,
according to their self-interest, without any intention of achieving justice (yet within
ethical boundaries preventing intentional obstruction of justice), by providing the best
plausible legal argument. The aggregation of the products of such self-interest-based
actions provides the impartial tribunal with the best form of legal argument from each
party. Then, so the explanation goes, although each legal argument is biased in favour of
one party, the arguments and evidence in their best form ‘cancel out’ each other’s biases.
As Mill (1989 [1859]: 38) puts it: ‘three-fourths of the arguments for every disputed
opinion consist in dispelling the appearances which favour some opinion different from
it’. The unintended optimal design is thus of a judge provided with the best conditions to
extract the truth of the matter, as a result of biased, zealous representation for each party
(Luban, 1988: 69; Rhode, 2000: 53).
The ‘division of labour’ argument. The ‘division of labour’ argument is a comparative
argument that refers to the limits of the inquisitorial system. The main feature of such
a system is its ‘judge-driven trial’. The judge conducts the trial, investigates, summons
witnesses, and questions the parties. The lawyers are considered an independent part of
the justice system, not the extension of the client’s will (Luban, 1988: 94). Fuller (Fuller
and Winston, 1978: 383) claims that in such a system, justice can be achieved only if the
judge undertakes the role of the representative of all parties. In order to fully understand
the case, the judge must ‘permit himself to be moved by a sympathetic identification
sufficiently intense to draw from his mind all that it is capable of giving – in analysis,
patience and creative power’. The problem is that after identifying with each side, the
judge needs to put back the impartial cap on – a requirement that Fuller describes as
unrealistic. ‘When [the judge] resumes his neutral position’, he argues, ‘he must be able
to view with distrust the fruits of this identification and to be ready to reject the products
of his own best mental efforts. The difficulties of this undertaking are obvious. If it is true
that a man in his time must play many parts, it is scarcely given to him to play them all at
once’. Therefore, another instrumental justification of the adversarial method is that the
labour is divided: the lawyers are in charge of identifying with the parties and making
their best case, whereas the judge is focused solely on judging the parties. This is a far
more achievable task for the judge, which increases the likelihood that she would
provide a just result.
Protecting individual rights
The standard conception of the lawyer’s role in the adversarial system consists of the
following two complementary requirements:
Full advocacy: A lawyer has an obligation to zealously defend her client according to her
client’s interests within the boundaries of the law. (Fuller and Winston, 1978: 382–384;
Luban, 1988: 11, 57, 62–63)
Non-accountability: A lawyer is not accountable for her client’s actions, should not be
normatively condemned for representing her client, and is not accountable for the result
6Politics, Philosophy & Economics XX(X)

of the trial, regardless of whether it is just or not. (Luban, 1988: 52–57; Wertheimer, 1988:
310)
Owing to this standard conception, the adversarial system provides maximum pro-
tection of individual rights. Two main arguments support this statement.
The ‘unconstrained defence’ argument. Full zealous advocacy provides the best protection
for an individual’s rights, simply because the more restrictions on the lawyer’s actions
there are, the fewer tools the lawyer has to protect her client. Moreover, full advocacy
prevents people or institutions (especially the state) from interfering with or influencing
one’s defence in court, as it ensures that the lawyer is solely bound to the client’s
interest.
15
Zealous advocacy is one way to ensure this; non-accountability is another.
It ensures that the lawyer can faithfully do her job, without worrying about being socially
condemned for her performance in court, e.g. for being responsible for the release of a
guilty man. Thus, this principle ensures that lawyers provide maximum protection to
clients’ rights.
The ‘division of judgement’ argument. The two requirements result in a division of ethical
judgement: the lawyer represents, and the judge judges (Luban, 1988: 78–80). The
difference between this argument and the ‘division of labour’ argument is the following.
The ‘division of labour’ argument focuses on the positive element of the adversarial
system, dividing labour according to roles in order to better achieve just results. By
contrast, the ‘division of judgement’ argument focuses on defending individual rights by
preventing such a division from collapsing as a result of ‘controversial clients’. The fear
of social condemnation provides an incentive for a lawyer to either ‘judge’ a client when
asked to be hired and thus refuse to work for controversial clients, or ‘judge’ a client
during representation, which results in inadequate representation.
The non-accountability requirement serves as a counterweight to such fear. It equips
lawyers with an ‘ethical shield’ that justifies their actions. Thus, it is instrumentally
desirable, because it increases everyone’s ability to receive maximum protection of their
rights by overcoming social condemnation.
These arguments may not seem to be direct arguments for the adversarial system, but
rather arguments for why, within an adversarial system, the lawyer’s role should have
certain characteristics. However, the formation of the lawyer’s role in the adversarial
system is unique and is celebrated as one of the system’s greatest advantages. If the
lawyer’s role were constructed differently, the normative appeal of the system as a
system which aspires to protect individual rights would be undermined. Thus, it serves
as a powerful independent justification for the system as a whole – not just as an
argument for what the lawyer’s role should be, having decided on an adversarial system
with representation.
The ‘formal justice’ justification
Before elaborating on the ‘formal justice’ justification, I would like to make the follow-
ing distinction. The justifications I have presented so far are intra-case justifications,
Agmon 7

focused on the normatively appealing features of the adversarial system within the
framework of a single case. However, they do not focus on the normative elements of
the comparison between cases. The following justification does just that, and therefore I
call it an inter-case justification.
As mentioned, the legal system is an instance of imperfect procedural justice. Its
results are considered as just if and only if they are in accordance with an independent
criterion. However, the procedure itself must be just as well. An adequate justice system
is one that does not subject individuals to unfair procedures, not only because they are
not likely to produce correct outcomes, but also due to their intrinsic inadequacy. In a
system of procedural justice, fair procedure matters independently of correct outcome.
One can have a valid grievance if subjected to unfair process, even if the outcome was
correct. On the other hand, one can lay no blame at the foot of the court in the case of
incorrect outcome produced by a fair procedure.
16
Such defendants, it seems, have a
valid grievance, not with respect to their conviction (which, per hypothesis was correct),
but for having been subjected to an unfair procedure. From a doctrinal perspective, the
norms associated with due process, embody the idea that fair trial is not merely an
instrument for optimal decisions.
Just institutions or systems are evaluated using the concept of formal justice. Rawls
(1999: 51) defines formal justice as the impartial, consistent, and equal application of the
law. In the process of adversarial adjudication, like cases should be treated alike, and
irrelevant factors should be excluded. Nonetheless, fair competition requires not just fair
rules and impartial referees, but also equal opportunities. As Brian Barry (1990: 99) put
it, procedural fairness must be supplemented by what he calls background fairness:
‘procedural fairness rules out one boxer having a piece of lead inside his gloves, but
background fairness would also rule out any undue disparity in the weight of the boxers
[or] sailing boats or cars of different sizes being raced against one another unless suitably
handicapped’. Following Barry, Rawls (1999: 63) argued for fair equality of opportunity,
as opposed to merely formal equality of opportunity, because:
those who are at the same level of talent and ability, and have the same willingness to use
them, should have the same prospects of success regardless of their initial place in the social
system ...The expectations of those with the same abilities and aspirations should not be
affected by their social class.
Fair equality of opportunity is necessary in order for the distribution produced by
market competition constrained by the difference principle to be just (Rawls, 1999: 63).
17
Analogously, the expectations of individuals in the same legal position (in terms of crimes
committed, criminal responsibility, available evidence, etc.) should not be affected by their
class, or financial resources.
I take it that the plausibility of such arguments and the common-sense intuition in
favour of the normative desirability of formal justice allows me to assume that if a
legal system does not comply with such requirements – impartiality, consistency, and
equality – it is a pro tanto reason to deem it unjust (Rawls, 1999: 52). Accordingly, there
is no need to elaborate on the specific ways in which the adversarial method adheres to
such normative aspirations (appointing impartial judges; strict procedural rules, etc.). It
8Politics, Philosophy & Economics XX(X)

is sufficient to assume that just like any other legal system, one of the adversarial
system’s normative justifications is that it ensures formal justice.
Part III: Undercutting legal justice – Civil and criminal cases
In this section, I argue that the fact that legal representation is allocated by a market
mechanism undercuts the normative appeal of the adversarial method. As there are
relevant differences between civil and criminal cases, I discuss them separately. I start
by discussing the clash between each justification of the adversarial system and the
market in legal representation in civil cases. Then, to avoid repetition, I discuss the
aspects of criminal cases that relevantly different.
Different types of civil cases
In arguing that the legal representation market undercuts the adversarial civil law sys-
tem’s normative appeal, I refer to three types of possible civil cases:
Poor vs. Poor Case (PvP): neither party has sufficient resources to fund adequate legal
representation, if any. I do not purport to define the exact amount of money for each case,
but I do assume that a minimal amount for adequate representation can be set. I also do not
assume that low-price lawyers are incapable of providing adequate representation. By poor I
mean people who cannot acquire a minimal level of legal resources which is needed for their
case. A good low-price lawyer cannot be of help, in some cases, without sufficient legal
resources.
Rich vs. Poor Case (RvP): one party has sufficient resources to fund adequate representa-
tion, whereas the other does not. This also includes cases in which both parties have
‘enough’, but the gap between them is significant enough to tilt the scales of justice in
favour of the richer party.
Rich vs. Rich Case (RvR): both parties have high and fairly equal resources for funding their
legal representation.
Undercutting the ‘providing just results’ justification (civil)
The ‘invisible hand’ argument. This argument relies on a qualitative assumption that each
party acts according to its self-interest and provides the best legal argument possible, as
well as on a comparative assumption – that the different arguments ‘cancel out’ each
other’s biases. The qualitative assumption conflicts with the legal representation market,
because people who do not have the resources to buy adequate representation (i.e., a
skilled lawyer with adequate legal resources) on the market would usually be unable to
provide the best legal argument (Rosen-Zvi, 2010: 718–719). Even in PvPs, although no
party has a material advantage over the other, both sides would provide poor arguments.
Thus, this problem applies both to RvPs and to PvPs.
The comparative assumption also conflicts with the legal representation market in
RvPs, because market-driven inequalities result in a situation in which due to unequal
legal resources, the legal arguments provided by the parties tend not to mutually ‘cancel
Agmon 9

each other out’, since they are not of the same quality. The perspective of the rich is more
dominant.
18
Both conflicts – the quantitative and the comparative – make it harder for the judge to
extract the truth of the matter, and thus undermine the invisible hand argument in favour
of the adversarial system.
The ‘division of labour’ argument. According to the division of labour argument, the adver-
sarial method is normatively appealing because it offers the most appropriate framework
for the delivery of just results. This argument does not directly conflict with the market in
legal representation. Even if there are disparities between the parties, or if the quality of
representation is poor, the division of labour might still be of value.
However, it does conflict indirectly with a potentially valuable response to the con-
flict between the market in legal representation and the invisible hand argument. Judges,
according to this response, could ‘reimburse’ the poor party in RvPs for the ‘influence’
that the rich party’s wealth has; or, alternatively, ‘improve’ both sides’ arguments in
PvPs, and thus avoid both the qualitative and comparative problems arising from the
invisible hand argument. Such ‘reimbursement’, according to Posner and Yoon’s survey
(2011: 335), already takes place in reality. Nevertheless, this response fails to vindicate
the method’s appeal. To start with, it does not apply to adversarial systems with juries –
citizens who do not possess the legal expertise required for the ‘improvement’ of legal
arguments. Second, the normative appeal of the adversarial system stems from the
superiority of its labour division in providing just results. Having a judge make argu-
ments for one of the parties (or both), combines the two roles into one. Thereby, it
undermines – and effectively, eliminates – the division of labour in an unappealing way.
If the involvement of the judge is necessary, an institutionalised mechanism allowing her
to lead the investigation seems more likely to provide just results than a vague ‘reim-
bursement’ by a passive judge in the adversarial system.
Undercutting the ‘protecting individual rights’ justification (civil)
Unconstrained defence argument. The standard conception of the lawyer’s role, which
consists of the ‘full advocacy’ and the ‘non-accountability’ requirements, is meant to
protect individuals’ rights. Although it is mostly regarded as a justification for criminal
cases, it applies to civil cases too. Guaranteeing that irrelevant factors (like powerful
corporations’ influence, or public condemnation) will not influence legal representation,
and thus jeopardise the protection of individuals’ legal rights, is as relevant in civil cases
as in criminal cases (Rhode, 2004: 8–10).
However, under the legal representation market, this standard conception leads to
different results. With regard to the full advocacy requirement, two problems arise. The
first is that, under the legal representation market, zealous advocacy is the exception
(Rhode, 2000: 55–56). Most people do not have the resources to pay for zealous advo-
cacy, and therefore do not enjoy the quality representation needed for protecting their
rights. This problem is evident in RvPs, where the quality disparities endanger the
disadvantaged party’s rights.
10 Politics, Philosophy & Economics XX(X)

It is hard to think about a similar problem in PvPs. In such a case, the parties do not
endanger each other, for both suffer from poor representation. Nonetheless, in PvPs it is
more difficult for the judge to provide just results. Think of a case in which the court’s
decision is reached by a coin toss. Ostensibly, there is no inherent unfairness in the act.
At the same time, there is no reason to believe that the result will be the right one. The
same applies to PvPs. Considering both sides are equal, no unfairness is evident, at least
from an intra-case perspective. Nevertheless, there is no reason to believe that the just
result will be reached by the court if both sides are poorly represented. Thus, both sides
are endangered by an unjust result that might violate their rights. In any case, it is clear
that one’s ability to protect oneself is undermined by poor representation.
The second problem arising from the full advocacy requirement is that, while most
people cannot afford quality representation, this requirement encourages the purchase of
zealous advocacy by those who can afford it, based on the false assumption that any
biases will cancel out. Thus, instead of protecting individual rights, zealous advocacy
gives rise to a violation of rights by encouraging the wealthy to pursue their self-interest
regardless of justice, at the expense of the disadvantaged.
The ‘division of judgement’ argument. The heart of this argument is that the non-
accountability requirement ensures that the normative and legal judgement of the case
is in the hands of the judge, not the lawyer. The legal representation market does not
conflict with this justification directly. On the contrary, it might appear that it serves as
an additional incentive for lawyers not to judge clients, but rather accept whoever can
pay. As Friedman (2002: 109) argues, the market is blind; it does not discriminate
between people – the only ‘filter’ in use is money. However, on closer inspection, the
filter introduced by the market is no better than the lawyer’s own normative judgement.
Under the market, those who do not possess the necessary means are ‘sentenced’ by their
financial status.
19
This is a paradigmatic example of the ‘access to justice’ problem, on
which I do not focus in this paper. Nevertheless, it is essential to bring it up in this
context, to show that the legal representation market undermines yet another argument in
favour of the adversarial method.
Undercutting the ‘formal justice’ justification (Civil)
Formal justice requires that the law and its enforcement are applied impartially, con-
sistently, and equally. The market in legal representation does not conflict with judges’
impartiality: I assume that judges rule according to what they believe is just, even if it is
against the interests of the rich. With consistency and equality, things get more compli-
cated. Assume the law says that X leads to consequence Y. Presumably, a judge will
always rule Y when faced with X. Therefore, consistency – in this narrow sense – is
maintained under the market. The same applies to equality: any person found responsible
for X would have to bear consequence Y, just like any other person found responsible
for X. The conflict between formal justice and the market in legal representation arises in
a wider context, namely, in the process of arriving at the conclusion that the facts of the
matter should be interpreted as X to begin with.
Agmon 11

Disparities in the quality of representation may distort this process in two ways. One
is that in some cases wealth may be needed in order to unveil the facts of the case, e.g.
legal experts, labs, investigators, etc. Thus, two objectively identical cases could be
treated differently in court due to epistemic disparity: the rich use extensive resources
to reveal the truth, while the poor cannot afford to do the same.
Another way that representation disparities affect the interpretation of cases has to do
with the comparison between identical cases (an inter-case conflict). In cases where the
law is unclear and can be interpreted in different ways, parties that are better represented
are more likely to convince the court to prefer their reading of the law. Thus, C
1
and C
2
–
two identical cases – may receive different treatment from the court, not due to an
asymmetry in their factual basis, but due to the different interpretation of the same set
of facts in each case.
Different types of criminal cases
I now move to argue that, though there are some differences between the criminal and
civil adversarial systems, they are insufficient to imply that the justifications for the
adversarial system are not undercut in the criminal system. The same arguments that
apply to the civil system apply – with some minor modifications – to the criminal system.
The kind of criminal cases to which I will refer to are the following:
State vs. Rich case (SvR): a case in which the defendant has more resources than the
prosecution. Theoretically, the state would almost always have more resources, or at least
equal resources, compared to the defendant. However, a rich person may decide to invest a
huge amount of money in one case that has no public significance. Matching her resources
in every case would be unreasonable for the state, as the rich has only one case to deal with
and the state needs to allocate its resources among many cases according to their public
significance.
State vs. Poor case (SvP): a case where the defendant is being inadequately represented by
a public defender due to a lack of resources, or where she privately funds her representation
but still has significantly less resources than the prosecution.
Differences and similarities with civil cases
The main relevant difference between criminal and civil cases is that in the criminal
system we prioritise avoiding false incarceration, and to achieve this we are willing to
trade off some competence in delivering just results. As Blackstone’s ratio indicates
(1765–1769: 352): ‘better that ten guilty persons escape than that one innocent suffer’.
This difference in priority results in two of the criminal system’s unique characteristics,
both of which are supposed to create a structural imbalance in favour of the defendant
(when compared to the parity between parties in civil cases). One is that the public
prosecutor (at least theoretically) is neither obliged nor expected to zealously defend the
victim (if there is one). To be sure, the state is a party to the trial, and the prosecutor
argues its case passionately. But the prosecutor represents the public – and the public’s
12 Politics, Philosophy & Economics XX(X)

‘self-interest’ is justice. So the state pursues ‘justice’, while the defence pursues their
own interests.
20
The other is the ‘beyond reasonable doubt’ standard of proof, compared
to the ‘balance of probabilities’ standard in civil cases, which means that in criminal
cases the burden of proof lies with the state.
21
Despite these differences, the fact that the basic adversarial logic is supposed to
function similarly in both kinds of cases means that, by and large, the legal representa-
tion market in the criminal system undercuts the system’s normative appeal in similar
ways to those presented with regard to the civil adversarial system.
First, the market means that many defendants are inadequately represented (not-
withstanding the fact that some states provide some funding to poorer defendants).
For example, in the United States most court-appointed lawyers do not have the
resources to defend their clients, and some do not even try – many defendants file
‘ineffective assistance’ claims about unprofessional or negligent lawyers. In addi-
tion, many poor defendants are provided with inexperienced lawyers (Rhode, 2004:
11–13, 122–137). Therefore, many SvPs are similar to PvPs and RvPs in that the
insufficient level of legal representation provided to the poor undercuts some of the
system’s justifications. So my argument about the market undercutting the ‘invisible
hand’ (qualitative assumption: the poor cannot provide the best legal case for them-
selves), ‘division of labour’ (we do not want the judge to reimburse for low quality)
and ‘unconstrained defence’ (the poor will not be zealously defended) justifications
– as well as the resulting inter-case formal justice justification, concerning like cases
being judged alike – apply. Only the division of judgement justification avoids being
undercut, because some level of representation is usually provided to defendants in
criminal cases.
However, intra-case disparities, on first glance, might seem less problematic in crim-
inal cases for two reasons. One is that the structural imbalance in favour of the defendant
in criminal cases might justify the claim, as Wertheimer (1988: 313) has argued, that
people should be able to get ‘the best legal representation that money can buy’ in order to
defend themselves. The second is that, in theory, the prosecutor has no incentive to use
her material advantage to tilt the scales of justice in order to win, so disparities in SvPs
are of less normative significance than in RvPs. After all, the prosecutor’s goal is
nominally justice, not winning (theoretically, of course; I return to this ‘naive’ assump-
tion below). I will take up each of these concerns in turn.
Concerning intra-case disparities in SvRs, as the system is already designed to
favour the defendant, allowing the rich to tilt the scales even more is unjustified,
since it could shift the balance in favour of the defendant to such an extent that it
might prevent the system from providing just results at all. Thus, intra-case dispa-
ritiesinSvRsincriminalcases,causedbythefreemarketinlegalrepresentation,
also undercut the ‘providing just results’ and ‘protecting individual rights’ justifica-
tions, as presented with regard to RvPs in civil cases. A plausible response to this
argument would be that in some cases the current institutional mechanisms, that are
supposed to ensure the imbalance in favour of the defendant, are ineffective and that
unlimited funding could be needed to achieve the right balance between the state
and the protection of the individual. However, if unlimited funding is needed to
achieve legal justice, the poor should also be entitled to it. But this, of course, would
Agmon 13

render the system unfeasible: we cannot provide all with unlimited resources. There-
fore, either the current institutional mechanisms to favour the defendant work (an
assumption I make for the purposes of my argument), or better mechanisms should
be put in place. Either way, market-based allocation that allows the rich to get the
best legal representation money can buy does not lead to the reestablishment of the
normative appeal of the system – on the contrary.
Moving to SvPs, the claim that disparities are less acute in criminal cases and
thus do not undercut the normative appeal of the adversarial system is objectionable
for several reasons. First and foremost, even though the prosecutor does not zeal-
ously defend a client, it does not change the fact that when she has more resources
within a legal competition her argument and the poor defendant’s argument will not
to mutually ‘cancel each other out’, as has been assumed in the comparative
assumption of the ‘invisible hand argument’, since they are not of the (roughly)
same quality.
Furthermore, if, in response, one would argue that the prosecutor can somehow
‘reimburse’ the poor’s insufficient level of representation, then one would face the same
problem of ‘undercutting the division of labour’ argument, as presented regarding civil
cases. Only this time, the division of labour problem seems to be much worse, since it
leaves the fate of the defendant not in the hands of the well-intentioned impartial judge,
but rather in the hands of the prosecution. Benevolent as the prosecutor may be, putting
the fate of the poor defendant in her hands disaffirms the fundamental assumption of the
criminal adversarial system – namely, that the facts of the matter will emerge out of the
legal competition between the state and the defendant. Therefore, disparities in SvPs, as
in RvPs, undercut the ‘providing just result’ justification.
As for the ‘protecting individuals’ rights’ justification, allowing inequality of
resources in favour of the state goes directly against it, because it gives the prosecutor
an institutionalised advantage over the poor defendant. If the whole point of this justi-
fications is to prevent the state from abusing its power against individuals, putting the
fate of individuals in the hands of the state’s lawyer seems unintelligible. So disparities
in criminal cases undercut this justification as well. It is true, however, that since the
prosecutor is not obliged to zealously defend the victim’s interests, in some cases where
the prosecutor is benevolent the results would not be as bad as in civil cases. But the fact
that disparities in criminal cases would sometimes not lead to results that are as bad as
the results in civil cases, is not good enough a conclusion to reestablish the appeal of this
justification.
Finally, even if in theory the prosecutors are supposed to pursue justice and not their
self-interest, in reality this is hardly the case. Either because they are being evaluated by
incarceration rates, or because they believe that they pursue justice, public prosecutors
are as zealous as any other lawyers.
To recap, due to the reasons specified above, although there are different normative
rationales underlying civil and criminal cases, the legal representation market in both
SvRs and SvPs, as in civil cases, undercuts the system’s normative appeal. There are
different existing institutional mechanisms (e.g. the ‘beyond reasonable doubt’ standard)
that are supposed to address these differences: the legal representation market is not one
of them.
14 Politics, Philosophy & Economics XX(X)

Part IV: Reestablishing the system’s normative appeal
The unstated requirements: (equal opportunity for) sufficiency and equality
The conflict between the legal representation market and the normative appeal of the
adversarial system, both in civil and criminal cases, stems from two problematic con-
sequences of that market:
The qualitative problem: Insufficient quality of representation of the badly off.
Thecomparativeproblem: Significant disparities between parties in the quality of
representation.
22
These suggest that sufficiency and equality in legal representation are unstated underlying
normative requirements of the adversarial system.
23
To be sure, perfect sufficiency and
equality are unattainable: even with sufficient and equal legal resources, one’s representation
could still be insufficient, because, for example, the lawyer made an unintentional human
error, or unequal if there are minor differences in talent between lawyers. This is where
opportunity comes in. Addressing the ways in which the market undercuts the normative
appeal of the adversarial system requires setting it up in such a way that each receives the
equal opportunity for sufficient level of representation, and the equal opportunity for equality.
So what should be done? I do not purport to provide a detailed alternative institutional
design for adversarial systems, which would require sophisticated empirical and eco-
nomic analysis. Instead, I will discuss two – relatively abstract – general ways to change
the current market-based system. One was presented by Wertheimer, and the other I offer
as a preferable alternative.
24
Such a discussion, and the normative conclusions I draw
from it, should provide a normative benchmark for future consideration of more detailed
proposals for reform.
I start by discussing the proposals with regard to civil cases, and then move to discuss
specific considerations for criminal cases (which are slightly different).
First option: Levelling down (civil cases)
Wertheimer (1988: 304–305) argues that in some RvPs we should level down the resources
of the rich to match those of the poor. The merit of such a suggestion is twofold. First, it
solves the intra-case comparative problem: the rich would not have the resources to tilt the
scales to their side. Second, it acknowledges the fact that different cases require different
legal resources, and does not impose a single cap on costs, but varies by case. However, it
is deficient for two reasons. First, it does not solve the qualitative problem. Levelling down
in RvPs, where the poor cannot afford to adequately represent themselves, would mean
that both parties would have inadequate representation.
25
Second, since the qualitative
problem remains unsolved, it is not clear whether levelling down solves the inter-case
conflict at all. Under a Wertheimerian legal system, some cases would be RvR (rich vs.
rich). Hence, there would be no need to level down. Other cases would be PvPs. In PvPs,
where the legal representation of both sides is equally poor, it would be harder for the court
to extract the truth of the matter, and consequently harder to apply the laws consistently
and equally. Thus, there would still be inter-case disparities between PvPs and RvRs.
Agmon 15

Furthermore, in such a system it would be hard to justify levelling down in RvP to the rich
party, since levelling down renders them unable to receive adequate representation. Thus,
Wertheimer’s intra-case levelling down solution seems unsatisfactory.
Second option: Levelling down þtype-based-floor (civil cases)
In contrast, the combination of levelling down and setting a resource ‘floor’ solves, at least
theoretically, the conflicts presented so far. This solution satisfies both the sufficiency ideal
(i.e. addresses the qualitative problem) and the equality ideal (i.e. addresses the comparative
problem).
26
In each type of case, a minimum level of resources would be set, to which each
party would be entitled, in order to ensure that all have an adequate level of representation
(the state would be responsible for allocating resources to those who cannot afford the
minimum). Such a floor cannot be universal: some kinds of cases require more resources
than others. Thus, there would need to be different thresholds for different types of case.
27
Levelling down is supposed to ensure that parties would be unable to turn their market
advantage into legal advantage.
28
In such a system, we are left with either AvAs (adequate
representation vs. adequate representation) or RvRs, since when disparities emerge, level-
ling down takes place.In both cases, the intra-case qualitative and comparative problems are
solved: there are no intra-case disparities and no qualitative deficiency.
A standard levelling down objection (hereafter: LDO) to my Levelling Down þType-
Based-Floor proposal is as follows. If one can improve the quality of one’s representa-
tion without harming anyone else, there is no reason for requiring one to level down just
for the sake of equality. Such an objection fails with respect to intra-case levelling down,
because one’s advantage in the adversarial system is necessarily the other’s disadvan-
tage. Therefore, levelling down is not done merely for the sake of equality, but also for
the benefit of the disadvantaged.
29
Moreover, contrary to Wertheimer’s suggestion,
under the system I propose, the rich person whose resources are levelled down would
not be able to complain that she is prevented from receiving adequate representation. She
would be able to complain that she could get better representation if she faced a richer
rival. However, under such a system, in which the rich person can fund her representa-
tion as she pleases in RvRs and get adequate representation, this complaint pales in
comparison to the disadvantaged person’s claim to receive an equal chance for justice.
Another objection to restricting the amount of money one could spend on legal
representation is that the disparity in the willingness to pay may reflect justified pre-
ferences, and not just the litigants’ ability to pay: a person may care greatly about a case,
and this may affect how much they spend. In such a case – the objection might run –
fairness could require that the one who cares more should be able to invest more in their
legal representation without being levelled down.
Before responding to the argument, it should be noted that even if this objection is
correct, it does not carry much weight. In reality, since defending oneself in court is a
high stakes affair, the limiting factor on spending for almost everyone would be the
ability to pay rather than willingness to pay. Therefore, the key question here is how
many actual cases are there in which willingness to pay actually tracks the true intensity
of people’s preferences, and not their ability to pay. This is an empirical matter, but I
think I can safely assume that the number is limited.
16 Politics, Philosophy & Economics XX(X)

As for the objection itself, it is unclear why one’s degree of ‘caring’ is relevant in the
context of providing legal justice in an adversarial system. First, if the equality and
sufficiency requirements are correct, then the system is set up in a way that already
assumes allocation of legal representation should be done according to principles that are
not responsive to preferences. Putting it differently, the system is designed to provide
justice – not to satisfy preferences. So one is at risk of proving too much by claiming that
preferences should matter for determining the amount of spending allowed, as this
undermines not only my proposal but the appeal of the system as a whole. More gen-
erally, it is also unclear why preferences should matter in relation to legal justice. To
illustrate, consider the following cases:
The Caring Rich: a rich person is charged with theft and faces trial. She has an extremely
strong preference not to go to jail, and legitimately decides to invest everything she has in
legal representation.
The Non-Caring Poor: a poor person from a minority group faces trial. As in the Caring Rich
case, she is charged with an identical case of theft. She does not have the will power to
properly defend herself, as she thinks the system is rigged against her, and decides not to
invest anything in legal representation.
It is likely that the court will reach different decisions in each case – despite the
charges being identical – owing to the rich woman’s better legal representation. Accord-
ing to the objection, this disparity is justified, since the rich cared more than the poor. I
find this result counterintuitive, to say the least. It is unclear to me why the fact that the
result of the case is more important to one party than the other should matter in a trial.
Quite the reverse, in cases where people do not wish to defend themselves, there might
be a justification to compensate for that, and not to reward the more ‘caring’ party. To
illustrate, imagine that the non-caring poor not only does not have the enthusiasm to
defend herself, but also insists on not being legally represented at all. In such a case, it is
clear that the adversarial method will not work as there are no two competing sides. To
achieve justice, it might be justified to force the defendant to be legally represented, or
conduct her trial according to a different, non-adversarial method.
Now, I am not trying to defend the rather strong claim that preferences should not
matter at all in the context of adjudication. I am also not trying to argue, at least in
this article, that we should force people to be legally represented if they do not wish
to be. My point is that the examples above show that in addition to the fact that the
adversarial system is already supposed tobesetupinawaywhichisnon-responsive
to preferences, the more general claim that preference-satisfaction should matter in
the context of providing legal justice also seems controversial. This is not a con-
clusive response, but I think it provides a pro tanto reason to consider this objection
not especially forceful.
Having established the justification for intra-case levelling down (with a resources
floor), the question remains: why allow RvRs at all? Why not force all to get equal
adequate representation and remain with a system in which all cases are AvA and
thereby solve the inter-case problem?
Agmon 17

The LDO does apply to this proposal. In the system I suggest, a rich person’s high-
quality representation can never harm the non-rich, since in each instance of intra-case
disparities the richer is required to level down. How can we justify levelling down the
rich from RvR to AvA, when no one is being harmed?
One answer might be that due to the requirements of formal justice, the differences in
results between RvR and AvA are unacceptable. However, assuming adequate represen-
tation, the differences should not be acute, since judges should be able to extract the truth
out of fairly strong arguments and evidence even in an AvA. Thus, there is a trade-off
between absolute formal justice and further restricting people’s freedom to fund their
legal representation as they please.
For those who are not strict egalitarians, the anti-levelling down stance is strength-
ened by this trade-off. Heavily funded legal representation can (and maybe does) lead to
the development of the legal profession, as it incentivises talented people to become
lawyers – funding that the state cannot provide. Second, it is unlikely that there would be
similar AvA and RvR cases. The usual legal needs of the rich differ immensely from
those of the poor (Hadfield, 2000: 998–999). Consequently, formal justice requirements
are unlikely to be violated, since there are few similar cases available to be treated
differently. Thus, there may be no need for inter-case levelling down in the first place.
Having said that, there is one reason to level down in RvRs. Investing many resources
in cases where there is no need for such an investment might lead to wasting resources
from the legal system – a fact which affects the public as a whole (Forer, 1984). For
example, cases with overinvestment in representation could take longer (good lawyers
can take advantage of the system and ‘drag’ the case for a long time). They could also
require more work from the legal system’s employees, who already face challenging
workloads and suffer from insufficient resources. It is true that this is a different a type of
reason, which is not directly related to the qualitative or comparative problems. Never-
theless, it is still a valid and important reason that should be weighed against the reasons
mentioned above for not levelling down in RvR cases. Thus, the question of whether the
rich should be levelled down, even in RvR cases, remains open for the time being.
Levelling down þtype-based-floor – Implications for criminal cases
In general, the Levelling Down þType-Based-Floor suggestion applies to criminal cases
as well. There are, however, two differences between the cases that should be consid-
ered. First, notice that in contrast with civil cases, there are no ‘RvR’ criminal cases.
Hence, levelling down in SvR would mean de facto that inter-case inequalities will not
exist: the state will have no reason to spend more than needed (and will be levelled down
if needs be) and the rich will have to level down.
Second, unlimited funding of the rich in criminal cases might be more objectionable
as it entails severe public costs. If justice is to be achieved, the state would need to
increase its resources to the level of the rich. Thus, spending substantial resources when
facing citizens of great means, so as to not constrain their right to use their money, seems
unreasonable – especially when levelling down means that the rich are still protected by
the structure of the system and by adequate representation. Thus, levelling down the rich
in SvR is justified, provided that the other components of my suggestions are fulfilled.
18 Politics, Philosophy & Economics XX(X)

In sum, although the criminal system operates with a different set of norms (‘less
justice, more protection’), my proposal stays largely the same: the poor should be
provided with adequate representation (where what ‘adequate’ means varies by type
of case), and the rich or the state should be levelled down in cases in which they possess
an advantage as a result of their wealth.
A final comment on feasibility
The normative guidelines sketched above might seem unsatisfying, since they do not
address feasibility concerns, like how an adversarial system can be economically main-
tained without a market. One objection could be that the rich, at least in civil cases, will
response to these restrictions by opting-out from the public legal system and increasing
their use of private arbitration. This would create problems in terms of inter-case equal-
ity, and could potentially lower the level of the services given to the rest of the public –
assuming the rich’s investment is crucial for sustaining the quality of the services.
30
Other objections might pose questions like how should the thresholds be set? How can a
system be funded in a manner that secures adequate representation for all? And would
this reform motivate people to overuse the legal system?
These considerations are beyond the scope of this paper. Also, following Jonathan
Wolff (2020: 260), I am not sure that philosophers have a clear advantage over other
experts in addressing them. I did make some feasibility-related assumptions throughout
my argument, but only uncontroversial ones that do not require complex analysis (e.g.
that it is impossible to provide unlimited legal resources to everyone all of the time). All
other assumptions which might be deemed feasibility-related (e.g. that it is possible to
provide justice by using an invisible-hand-based adversarial mechanism) are internal to
the justifications of the adversarial system. If they are incorrect, then the system is
founded on false assumptions and is unjustified to begin with.
One might argue that without answering these feasibility-related questions, my argu-
ment is meaningless. I dispute this claim. My purpose was to show that if we take the
normative appeal of the adversarial system seriously, the requirements of equality and
sufficiency must be somehow satisfied – and that in order for any kind of real life legal
adversarial mechanism to work, market-based allocation is not the way to satisfy it.
Thus, the Levelling Down þType-Based-Floor proposal serves as a normative bench-
mark to evaluate whether an adversarial legal system is justified or not. My discussion of
this proposal was intended to anticipate normative objections that could have potentially
stopped it from getting off the ground (e.g. the LDO). Questions regarding viability do
not undermine my proposal, but rather the possibility of establishing a (roughly) justified
adversarial system in real life. To see this, consider these three scenarios concerning the
relation between feasibility concerns and the normative stand of the adversarial system.
1. A free market in legal representation is necessary for the economic sustainability
of the adversarial method, but meeting the normative ideals of such a system is
impossible. Consequently, the method cannot be justified. That is unless one finds
other justifications for maintaining the system. For example, if there are no other
feasible legal systems that are more desirable. In this case, the trade-off would be
Agmon 19

between a non-appealing adversarial legal system, other less-appealing systems,
and no system at all.
2. A free market is not necessary for the system, but getting rid of it is too costly (e.g.
in terms of efficiency or stability). In this case, an adversarial method could be
justified (under very different circumstances), but is not.
3. There is a way to both implement the Levelling Down þType-Based-Floor ideal
and keep the system sustainable. In this case, once a reform is put in place, the
adversarial method could be and de facto is justified.
In all three cases, feasibility-related concerns either affect the possibility of ever
establishing justified adversarial legal systems (option 1), or the possibility of establish-
ing a justified adversarial system in reality (options 2 and 3). They do not, however,
impact the justification for the normative ideal I propose.
Conclusion
In contemporary adversarial systems, the services provided by lawyers are a commodity
sold and bought in the market: the more one pays, the better quality of legal represen-
tation one gets. A lot has been said and written about the problems that this institutional
structure gives rise to – the ‘symptoms of the disease’, as it were. But the disease itself –
i.e. the fact that there is a market in legal representation incorporated into the adversarial
system – have rarely been subject to normative examination. In this article, I have tried to
address this discrepancy by taking a new approach to the philosophical evaluation of our
adversarial legal systems: instead of making arguments about the role of distributive
justice or professional ethics in the legal system, I have discussed the institutional
normative limits of the market in legal representation.
After carrying out a scrutiny of this kind, my conclusion is that allocating legal
representation by the market in adversarial systems undercuts the system’s normative
appeal. In order for this normative appeal to survive, I showed that two, previously
unstated normative requirements must be satisfied: (equal opportunity for) equality in
legal representation and (equal opportunity for) sufficiency of legal representation. My
Levelling Down þType-Based-Floor proposal serves as a starting point for a reform that
would satisfy these requirements.
Acknowledgements
Earlier versions of this article were presented at the ‘State and the Market’ and the ‘Law
and Political Theory’ workshops in Tel-Aviv University (2017), The 4th Annual Con-
ference of the Centre for the Study of Global Ethics in the University of Birmingham
(2018), The Office of Legal Counsel and Legislative Affairs – Israel’s Ministry of Justice
(2018) and the OWIPT (Oxford University’s Work in Progress – Political Theory)
Seminar (2019). I am grateful to the participants of all these meetings for their insightful
feedback. Comments from Susanne Burri, Talia Fischer, Ruth Gavison, Avner Inbar,
Cristobal Otero, Alex Voorhoeve, Fredrick Wilmot-Smith and Neta Ziv encouraged me
to revise important parts of the main argument. I am especially grateful to Judge Michal
Agmon-Gonnen, David Enoch, Matt Hitchens, Yonatan Levi, Michael Otsuka, Assaf
20 Politics, Philosophy & Economics XX(X)

Sharon, an Editor at Politics, Philosophy & Economics, and two anonymous referees for
their many helpful written comments and support upon earlier drafts.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publi-
cation of this article.
Notes
1. Hereafter I will use the term ‘truth of the matter’ to refer to both discovering the facts and, and
to the accurate legal interpretation.
2. I will use, loosely, ‘lawyer’ and ‘legal representation’ interchangeably throughout this article.
3. In using the terms ‘lawyers’, ‘legal representation’ or ‘quality legal representation’, I refer
both to the quality of the lawyer herself and the quality of legal resources available to her –
depending on the financial resources of the client, as both influence the overall quality of the
client’s legal representation.
4. Of course, money enters the ‘legal system’, in a broader sense, much earlier on. The emer-
gence of disputes itself is influenced by the economic status of the opposing parties (as
demonstrated in Felstiner et al., 1980: 633–637). However, these problems have been recog-
nised and researched thoroughly, and are not unique to the legal system. Inequalities lead to a
wide range of problems, with respect to many parts of society. The purpose of this paper is to
show that the specific fact that contemporary adversarial legal systems are incorporated within
a free market in lawyers leads to a different, more constitutive set of challenges.
5. A paradigmatic example for this phenomenon is Schwartz’s discussion (1983: 543–563) of the
moral accountability of civil advocates. He acknowledges and identifies normative require-
ments for the adversarial system that could be undercut by a free market mechanism, but he
does not question or challenge the market-based allocation of legal representation. Rather, he
focuses on the implications of these requirements on the moral accountability of lawyers.
6. Wertheimer (1988) is the only scholar I am familiar with who provided a philosophical
analysis of the market in legal representation within the adversarial legal system. However,
he only discusses one justification of the adversarial legal system (providing just results), and
focuses solely on civil cases (while I am discussing both criminal and civil cases). I elaborate
on his argument in part IV of this paper. Additionally, during the publication process of this
article, Frederick Wilmot-Smith (2019) has published a new book in which he argues for the
primacy of justice over the market in legal resources. Nevertheless, he provides a somewhat
general argument which is directed at all legal systems, thereby leaving the detailed discussion
on the adversarial legal system unattended.
7. A similar philosophical approach was taken recently by Johana stark with regard to regulatory
competition law (see Stark, 2019: 125–126).
8. To clarify, my ideal for reform is therefore desirable solely in order to reestablish the
adversarial system’s appeal. Further argument would be needed, if there is any, to apply it
generally to all legal systems. Additionally, my intention is to make a fairly general
Agmon 21

normative claim about the adversarial legal system as a whole. Admittedly, my argument
might not apply to some specific fields of law, and would probably need refining in order
to apply to different adversarial systems. This simply highlights that there is more work
to be done.
9. Notice that my argument is not concerned with the question whether legal representation
should be a private or public good. As shown by Anderson, goods can be privately owned
without being allocated by the market, or publicly owned and marketised (Anderson, 1993:
190–191). There might be other reasons to think that legal justice should be a product of a
completely public owned endeavour, but this discussion is beyond the scope of this paper (for
more on this see Brennan, 2017; Dorfman and Harel, 2016).
10. Indeed, according to the American Bar Association’s report (2014), most graduates from elite
universities are employed by top private firms that seldom represent the poor. Moreover, a
Legal Services Corporation (LSC) report (2009) shows that the proportion of self-
representation in court among low-income citizens is extremely high.
11. For example, a survey conducted among American judges showed that courts’ decisions are
significantly influenced by disparities in legal representation quality (see Posner and Yoon,
2011: 335–336, 343–346; see also Anderson and Heaton, 2012: 188–197). It should be noted
that James Geyner and Cassandra W Pattanyak (2012: 2175–2196) have challenged the existing
literature on the subject for ‘purporting to measure quantitatively the effect of legal represen-
tation in civil disputes’, and argued that it is extremely difficult to reach generalised conclusions
regarding the impact of quality legal representation. According to them, legal representation
does not always make a difference (for an example of a case heavily influenced by the quality of
representation see Greiner et al., 2012: 901–989). For the purpose of this paper I do not wish to
evaluate the findings presented in the empirical literature. If Geiner and Pattanyak are correct,
my argument is limited to cases where quality legal representation matters.
12. Another justification for the adversarial system on which I do not elaborate focuses on the
intrinsic value of the adversarial method. This justification reflects the notion that the lawyer’s
role in the adversarial system is intrinsically valuable, for it is an instantiation of the client’s
dignity (Donagan, 1983: 123–133), autonomy (Fried, 1976: 1072–1073) or right to be heard
(Harel, 2014: 225). Although I do not discuss this line of justification, I do believe that this
justification is also undercut by the market in legal representation, as one’s right to be heard is
dependent on one’s access to legal resources. Another useful summary of the justifications for
the adversarial legal system can be found at Wendel (2010: 17–44).
13. I do not offer a defence for these justifications. If one rejects them, then one has no reason to
believe in the adversarial system independent of the conflict between its normative justifications
and the market in legal representation (unless, of course, there is some other justification for it).
14. Another relevant question is whether the adversarial system produces just results more often
than alternative legal systems. I do not discuss such comparative arguments.
15. In criminal cases the prosecutors’ role is different, as I explain in section IV.
16. Although one might have a legitimate grievance in the case of false conviction, for example,
this is a separate issue from whether the court is blameworthy.
17. Notice that the concern is not merely optimising outcomes: ‘the reasons for requiring open
positions are not solely, or even primarily, those of efficiency. I have not maintained that
offices must be open if in fact everyone is to benefit from an arrangement’ (Rawls, 1999: 73).
Restricting access to positions, Rawls argues, is a violation of equal treatment.
22 Politics, Philosophy & Economics XX(X)

18. Recall that my argument does not encompass cases in which there is sufficient evidence and
the legal interpretation is conclusive, since in such cases the impact of quality legal represen-
tation is marginal. However, in real life, even in cases where P is a stone-cold winner, R can,
and sometimes does, use her economic advantage to settle the case out of court, so that the
cases that do go for trial are more likely to be close cases in which disparities in legal resources
are decisive (see Gross and Syverud, 1991: 319–393).
19. For instance, the LSC report (Documenting the Justice Gap in America: The Current
Unmet Civil Legal Needs of Low-Income Americans (2009)) shows that almost one
million Americans were denied legal aid by the LSC alone, and were left without legal
representation.
20. However, oftentimes this structural imbalance does not have significant influence within a
system in which the overwhelming majority of criminal cases are settled via plea-bargaining
(Pfaff, 2017). Thus, some part of this section encompasses a fairly idealised normative dis-
cussion, and there might be a need to further strengthen the poor and adequate defendants’
rights against the zealous prosecution.
21. Another difference is that in criminal cases the state usually provides its own legal represen-
tation, so there is only a legal representation market for defendants – but this does not change
much for the purpose of my argument.
22. It might be plausible that in a different legal system equality between the parties would not be
needed at all. However, this is the way adversarial systems are set. Thus, this is more of a
structural problem then an intrinsic problem of inequality.
23. A similar idea was suggested in Murray L Schwartz’s ‘Postulate of Equal Competence’ (1983:
546–548). However, he did not make the link between the postulate and the market in legal
representation.
24. I only discuss proposal which involve levelling down, as I assume that balancing the scales of
justice by levelling up people’s legal resources, so as to make sure all cases become, in effect,
RvR cases, is unfeasible, due to scarcity of public resources.
25. Recall that the definition of RvP includes cases in which the poor side can afford adequate
legal representation, but the gap with the rich is so large that it is still problematic. Only in
such cases Wertheimer’s suggestion is desirable.
26. By ‘equality’ I do not mean equality per se, but rather a balance between the parties within the
adversarial adjudication process.
27. Due to the diversity of cases, the classification of types would be general, and exceptions
would be required. However, such a problem is common to all general legal principles that
need to be applied to a diversity of cases, and therefore using it to object my claim would be to
demand too much.
28. In some cases, one party’s representation necessitates more resources than the representation of
the rival party. Therefore, equalising resources by levelling down mightlead to unjust results. To
counter this, the threshold can be set according to the party which requires more resources.
Moreover, the threshold could be a range between two limits and not a specific number.
29. This is an instance of a well-known, more generalpoint, according to which nothing like the LDO
applies in competitive settings or to positional goods (see Brighouse and Swift, 2006: 471–497).
30. For a similar objection regarding the education system see Hirschman (1970: 51–52).
Agmon 23

References
Anderson E (1993) Value in Ethics and Economics. Cambridge, MA: Harvard University Press.
Anderson JM and Heaton P (2012) How much difference does the lawyer make: the effect of
defense counsel on murder case outcomes. Yale Law Journal 122(1): 154–217.
Assy R (2015) Injustice in Person: The Right to Self-Representation. Oxford: Oxford University
Press.
Barry B (1990) Political Argument: A Reissue with a New Introduction. Berkeley, CA: University
of California Press.
Blackstone W (1765–1769) Commentaries on the Laws of England, vol. 4. Oxford: Clarendon
Press.
Brennan J (2017) Consequences matter more: in defense of instrumentalism on private versus
public prisons. Criminal Law and Philosophy 11(4): 801–815.
Brighouse H and Swift A (2006) Equality, priority, and positional goods. Ethics 116(3):
471–497.
Conte A, Newberg HB and Rubenstein W (2011) Newberg on Class Actions, 5th ed. Eagan: West.
Donagan A (1983) Justifying legal practice in the adversary system. In: Luban D (ed) The Good
Lawyer: Lawyers’ Roles and Lawyers’ Ethics. Totowa: Rowman & Allanheld, pp. 123–149.
Dorfman A and Harel A (2016) Against privatisation as such. Oxford Journal of Legal Studies
36(2): 400–427.
Emons W (2000) Expertise, contingent fees, and insufficient attorney effort. International Review
of Law and Economics 20(1): 21–33.
Felstiner WLF, Abel RL, and Sarat A (1980) The emergence and transformation of disputes:
naming, blaming, claiming ....Law & Society Review 15(3/4): 631–654.
Forer LG (1984) Money and Justice: Who Owns the Courts? New York, NY: Norton.
Fried C (1976) The lawyer as friend: the moral foundations of the lawyer-client relation. Yale Law
Journal 85(8): 1060–1089.
Friedman M (2002) Capitalism and Freedom. Chicago, IL: The University of Chicago Press.
Fuller LL and Winston KI (1978) The forms and limits of adjudication. Harvard Law Review
92(2): 353–409.
Galanter M (1974) Why the ‘haves’ come out ahead: speculations on the limits of legal change.
Law & Society Review 9(1): 95–160.
Gerber RJ (1987) Victory vs. truth: the adversary system and its ethics. Arizona State Law Journal
19: 3–26.
Geyner J and Pattanyak CW (2012) Randomized evaluation in legal assistance: what difference
does representation (offer and actual use) make? Yale Law Journal 121: 2118–2215.
Greiner DJ, Pattanayak CW and Hennessy J (2012) The limits of unbundled legal assistance: a
randomized study in a Massachusetts district court and prospects for the future. Harvard Law
Review 126: 901–989.
Gross SR and Syverud KD (1991) Getting to no: a study of settlement negotiations and the
selection of cases for trial. Michigan Law Review 90(2): 319–393.
Hadfield GK (2000) The price of the law: how the market for lawyers distort the justice system.
Michigan Law Review 98(4): 953–1006.
Harel A (2014) Why Law Matters. Oxford: Oxford University Press.
Hayek F (1960) The Constitution of Liberty. Chicago, IL: University of Chicago Press.
24 Politics, Philosophy & Economics XX(X)

Hirschman AO (1970) Exit, Voice and Loyalty. Cambridge, MA: Harvard University Press.
Jones C (2003) Theory of Class Actions. Toronto: Irwin Law.
Legal Services Corporation (LSC) report (2009) Documenting the Justice Gap in America: The
Current Unmet Civil Legal Needs of Low-Income Americans.
Luban D (1985) Political legitimacy and the right to legal services. Business and Professional
Ethics Journal 4(3/4): 43–94.
Luban D (1988) Lawyers and Justice:An Ethical Study. Princeton, NJ: Princeton University Press.
Markovits D (2008) A Modern Legal Ethics: Adversary Advocacy in a Democratic Age. Princeton,
NJ: Princeton University Press.
Mill JS (1989 [1859]) On liberty. In: Collini S (ed) ‘On Liberty’ and Other Writings. Cambridge:
Cambridge University Press, pp. 1–116.
Moorhead R (2011) Filthy lucre: lawyers’ fees and lawyers’ ethics – what is wrong with informed
consent? Legal Studies 31(3): 345–371.
Nozick R (1974) Anarchy, State, and Utopia. New York, NY: Basic Books.
Pfaff J (2017) Locked In: The True Causes of Mass Incarceration and How to Achieve Real
Reform. New York, NY: Basic Books.
Posner RA and Yoon AH (2011) What judges think of the quality of legal representation. Stanford
Law Review 63: 317–350.
Rand A (1964) The objectivist ethics. In: The Virtue of Selfishness. New York, NY: Penguin,
p. 134.
Rawls J (1999) A Theory of Justice, revised ed. Cambridge: Belknap Press.
Rhode DL (2000) In the Interests of Justice: Reforming the Legal Profession. Oxford: Oxford
University Press.
Rhode DL (2004) Access to Justice. Oxford: Oxford University Press.
Rosen-Zvi I (2010) Just fee shifting. Florida State University Law Review 37(3): 717–768.
Satz D (2010) Why Some Things Should Not Be for Sale: The Moral Limits of Markets. New York,
NY: Oxford University Press.
Schwartz ML (1983) The zeal of the civil advocate. Law & Social Inquiry 8(3): 543–563.
Stark J (2019) Law for Sale: A Philosophical Critique of Regulatory Competition. Oxford: Oxford
University Press.
Ullmann-Margalit E (1978) Invisible-hand explanations. Synthese 39(2): 263–291.
Wendel WB (2010) Lawyers and Fidelity to Law. Princeton, NJ: Princeton University Press.
Wertheimer A (1988) The equalization of legal resources. Philosophy & Public Affairs 17(4):
303–322.
Wilmot-Smith F (2019) Equal Justice: Fair Legal Systems in an Unfair World. Cambridge, MA:
Harvard University Press.
Wolff J (2020) Ethics and Public Policy: A Philosophical Inquiry. New York, NY: Routledge.
https://books.google.co.il/books?id¼d1GqjIhRejMC&printsec¼frontcover&source¼gbs_ge_
summary_r&cad¼0#v¼onepage&q&f¼false
Author biography
Shai Agmon is a DPhil candidate in the Department of Politics and International Relations at the
University of Oxford. His research interests include institutional limits of markets, prioritarianism,
intergenerational justice, legal philosophy, philosophy of economics and public policy.
Agmon 25