Chapter 7: The Choice of Delivery Models for Legal Aid


No single issue in the history of legal aid, both in Ontario and elsewhere, has inspired as much intense debate as the choice of delivery models for legally aided services. Much of this debate in the past has focused on the advantages and disadvantages of a pure judicare model and those of a pure staff model. These differences in viewpoints are reflected in the equally divergent approaches of these delivery models. In terms of the allocation of legal aid resources within Canada, Ontario, Alberta, and New Brunswick utilize predominantly the judicare model, whereas Prince Edward Island, Nova Scotia, and Saskatchewan rely predominantly on the staff model. Quebec, Manitoba, British Columbia, the Northwest Territories, and the Yukon have adopted a mixed model, with major judicare and staff elements. In jurisdictions beyond Canada, similar divergences in choice of models are also observable.

Until recently, many of the debates surrounding the choice of delivery models were absolutist in character. Proponents of one model viewed it as superior to the other model in almost all circumstances. These claims and counter-claims were grounded on both ideological and empirical considerations. Proponents of the judicare model argued that a private, decentralized bar guaranteed clients maximum freedom to choose their counsel; ensured that the principal loyalty of the lawyer was to his or her client, whose interest would be defended or advanced as vigorously as would that of a private fee-paying client; and, in combination with vesting of the governance of the Legal Aid Plan in the profession itself, ensured maximum independence from state interference or influence, particularly in cases where the state was directly implicated as a party (e.g., criminal prosecutions, child protection cases, immigration cases, and state benefit or entitlement claims).

In contrast, proponents of a pure staff model, in which lawyers and support staff are hired directly by the legal aid administration, argued that it was less costly than its judicare counterpart because of economies of scale and specialization, particularly in processing high-volume, more routine cases; was likely to yield more consistent quality of service; and could be made more accessible to clients through location of Staff Offices in client communities. Proponents of the community legal clinic version of the staff model, in which staff are hired by each clinic and are accountable to a community-elected board responsible for the clinic's governance, claimed that a less radically individualistic and more communitarian approach to the delivery of legal aid services was superior to the judicare model-first, in being responsive to the priorities and concerns of the community in which these offices were located and, by which, to a greater or lesser extent, the offices were governed; second, in being more proactive in undertaking community outreach and sensitizing disadvantaged and often ill-informed citizens of their legal rights, and assisting them to take active steps to enforce them; and third, in promoting and advancing law-reform initiatives and test-case litigation designed to address systemic deficiencies in the law, to which an uncoordinated, decentralized legal-service delivery model (such as judicare) was unlikely to assign a significant priority.

At the empirical level, proponents of the two models (three, if community clinics are viewed as significantly distinct from the pure staff model) typically made sharply contrasting claims on behalf of their preferred model, based to a large extent on anecdotal evidence and intuition, and to a lesser extent on systematic empirical data relating to the relative performance characteristics of the two models in terms of cost, quality, accessibility, and social impact.

As Susan Charendoff, Mark Leach, and Tamara Levy note in their research paper prepared for the Legal Aid Review, more recently the discussion about the preferable legal aid model has refocused on a more productive set of issues:

There is now a growing consensus among commentators on legal aid that defining the right array of service components-varying with type of law, client need, case priorities, type of service being offered (e.g., bail hearing as opposed to a jury trial), or the collective characteristics of the needs of certain groups of clients, such as Aboriginal people or the physically, or mentally handicapped-is far more useful.

With this reconceptualization of the issues, it has also become clear that the choice of delivery models is not a simple binary one between the judicare and staff models. For example, on the judicare side, important choices must be made about the structure of the judicare tariff, for example, hourly rates, fixed fees per case, maximum hours per case, and graduated rates reflecting greater levels of experience or expertise. In addition, contracting-out to private law firms or private counsel through a competitive tendering process of blocks of cases for fixed fees per case or a total sum for the block is an option that has begun to be explored seriously, and indeed experimented with, in other jurisdictions in various, although limited, contexts. On the staff-model side, there is a need to differentiate between delivery modalities where the legal and support staff are employed directly by the Plan, and community legal clinics which are independently incorporated as non-profit organizations with their own community governance structure and which employ their own staff and pursue their own mandate. Within the community legal clinic model, which is more strongly developed in Ontario than in any other jurisdiction, most clinics are all-purpose, providing a broad array of legal and related services in the general domain of "poverty law", for example, income maintenance, including employment insurance, Canada Pensions, workers' compensation, welfare and family benefits; work-related issues, including employment standards, occupational health and safety, wrongful dismissal; housing problems; and consumer and debt problems. However, other community legal clinics have a much more specialized focus and cater to the legal needs of particular constituencies, for example, children and youth; the mentally and physically handicapped; Aboriginal people; the African-Canadian community; injured workers; and environmental concerns. Some direct staff models provide a full array of legal services within a defined legal domain (e.g., criminal law), while others involve counsel only providing summary advice or handling preliminary proceedings (as in a duty counsel model). Within a duty counsel model, duty counsel could be full-time employees of the Plan or could be retained from the private bar on a per-hour, per-diem, per-week, or some other basis.

In mixed models, the systems' different delivery methods may, in some respects, be complementary to one another and, in other respects, competitive with one another. For example, community legal clinics largely offer "poverty law" services that, for economic reasons, the private bar has little interest in providing and that, in many cases, are not suitable for discrete assignment to private counsel under a judicare system. Where a client at a clinic requires major criminal or family law services, clinics can, and commonly do refer the client to a judicare lawyer. Similarly, duty counsel handling predominantly first appearances and adjournments in criminal proceedings, and providing summary advice to individuals, in many cases deliver services that are complementary to those of private counsel who may subsequently assume carriage of the matter under a legal aid certificate. A Staff Office handling refugee and other immigration matters and employing specialized paralegal staff with close acquaintance with the countries from which immigrants have arrived may have the capacity to undertake country-specific research that can be shared with private counsel acting for refugees or immigrants in subsequent proceedings, thus minimizing the need for duplicative, and perhaps less expert, research.

On the other hand, in mixed systems, different delivery models may directly compete with one another. This is true of the staff models in Quebec, Manitoba, and British Columbia, where clients have a choice between using a lawyer employed in a Staff Office and using private counsel under a legal aid certificate. For obvious reasons, debates over choice of delivery models tend to be intensified to the extent that these models compete with rather than complement one another. As in most walks of life, whatever virtues one might attach in the abstract to competition at either the systemic or the individual-provider level, when one's own economic self-interest is threatened by the prospect of such competition it is easy to be persuaded that competition works better in other areas of economic life than in the context at hand.

These preliminary observations lead us to the general orientation for this chapter, which comprise three basic elements. First, the Ontario legal aid system in the future should adopt, explore, and experiment with a much richer array of delivery systems than those that have characterized the system to date. While it is often claimed that the Ontario legal aid system is mixed because of the co-existence of the judicare and clinic systems, in fact 80 to 85 percent of the resources of the Plan are devoted to the judicare system and only about 10 percent of the Plan's resources to the clinic system. In terms of direct staff models, they receive a minuscule percentage of the Plan's resources (primarily duty counsel programs in selected criminal and family law courts), and no attempt has been made to experiment with contracting-out blocks of cases.

The second element in our general orientation, which we do not develop at length in this chapter but do so in chapter 15, is that, in the future, a governance structure is required for the legal aid system that renders it amenable to open-minded, highly motivated, and innovative utilization and exploration of a rich array of delivery models, and systematic and objective periodic evaluations of their performance. Obviously, such evaluations (which should be undertaken by independent review and made public) should employ a settled and well-defined set of predetermined criteria reflecting performance objectives, including cost, quality of service, accessibility, and social impact.

The third element in our general orientation-performance evaluation and quality assurance-should be assigned a much more prominent role in Ontario's future legal aid system than has been in the case in the past. While comparing the costs of different delivery models (e.g., cost per case) has proven an intractable task, given problems of standardizing or otherwise taking account of differences in the nature of case-loads, problems of evaluating relative quality performance under different models has proven even more intractable, given the much more subjective, context specific, and non-observable dimensions of quality that are often-probably typically-involved. Relatedly, beyond simply measuring or evaluating relative quality of service, as we argue more fully below, serious efforts need to be undertaken to ensure base-line levels of quality assurance in the judicare system, the community legal clinic system, the direct Staff Offices that we recommend be created, and in any experiments with block contracting. Without such efforts to minimize the often wide variability that currently exists in the quality of legal services provided under the Plan, the interests of clients are not well served and the taxpayers of Ontario are not getting full value for their money.

In section 1 of this chapter, we briefly review the existing empirical research on relative performance characteristics of alternative delivery models. In section 2, we analyze from an economic perspective the markets for legal services in general, and the market for legally aided legal services in particular, focusing on systemic factors under different delivery models that are likely to affect two key performance variables: (a) supplier-induced demand; and (b) quality decline. In the light of this analysis, we propose in section 3 a series of quality-assurance measures designed to mitigate these tendencies. In section 4, we offer our conclusions and, in section 5, summarize the recommendations developed in this chapter.


A handful of comparative studies in Canada and the United States have attempted a systematic evaluation of the relative performance of various legal aid delivery models. These have been extensively reviewed in earlier studies, and in a research paper prepared by Susan Charendoff, Mark Leach, and Tamara Levy for this Review. We do not reiterate the details of these studies here. The most striking finding from most of the studies is that no one delivery model exhibits performance characteristics that are systematically superior to those of other delivery models in all contexts. We review the empirical evidence briefly under specific performance characteristics.


Measuring the costs of different legal aid delivery models is often problematic because of the number of variables involved. Indigent Defenders, a study sponsored by the U.S. National Center for State Courts, described this dilemma well:

Efforts to arrive at a meaningful cost comparison across courts or on a cost per case basis are fraught with difficulties, uncertainties and hazards. Continued research along these lines, in fact, is of questionable value because of a lack of a relevant standard against which cost-per-case can be judged.

Other factors, such as the differing referral patterns, the client's freedom to choose counsel, and the size and structure of the tariff, add to the difficulty of comparing costs across jurisdictions.

Many of the Canadian, single-jurisdiction, federal/provincial legal aid studies undertaken in provinces that use a staff model conclude that that model is the less expensive delivery mode. However, most of the controlled, comparative studies completed in both Canada and the United States conclude that there is no significant difference in cost between a staff and a judicare mode of delivery.

Despite these contradictory findings, several studies argue that the staff model should have one significant advantage over the judicare model in terms of costs. This difference is the result of a staff lawyer's ability to specialize in legal aid cases. Generally, lawyers who accept legal aid certificates also have paying clients; there are relatively few whose practice is made up exclusively of legal aid case work.

Staff lawyers, at least in the field of criminal law, spend each day dealing only with legal aid cases and, in the course of doing so, often develop close relationships with the Crown Attorneys in their area. As a result of this close relationship, staff counsel are more successful plea bargainers. The Burnaby, British Columbia, study found that, although the rate of guilty outcomes was essentially the same for judicare and staff clients, clients of the Staff Office who were charged with only one offence (which represented about half the sample) went to jail less often than did single-charge judicare clients. For the clients with multiple charges, jail results were similar. The ability of staff counsel to negotiate with the Crowns at an early stage in the proceedings can lead to a more rapid disposal of cases. Although this may not always be the result, the cost savings achieved through specialization is frequently held to be one of the benefits of the staff model. The perceived benefits of the staff model's ability to specialize are discussed further below.

If close relationships are relevant, however, it may be that judicare lawyers whose practice is devoted to criminal law would produce the same benefits. On that basis, the appropriate comparison would not be between staff lawyers and judicare lawyers, but between staff and judicare lawyers who specialize in criminal law, and judicare lawyers whose practice is a mix of criminal and other matters. There may be further subcategories, because in smaller communities many lawyers are considered criminal specialists and possess solid, professional relationships with Crown Attorneys when only a portion of their practice is devoted to criminal law. Thus, it may not be the mode of payment that matters, by the year or by the case, but the concentration of a lawyer's practice.

Another conceivable ground for cost differentials flows from the way that cases are assigned to staff lawyers, based on the assumption that the public defender will "staff" a specific courtroom, act as duty counsel as well as representing individuals, develop a rapport with the Crown Attorney and judge, and thereby become a part of the administration of the criminal court. But even this cost difference would flow from the way staff are assigned work rather than the way they are paid. On the other hand, this may well not be true for the more serious and expensive cases, where there would have to be some choice of counsel, or the counsel assigned would have to follow the case to whatever court it goes.

It is worth stressing that the Burnaby Report, which compared a staff criminal law office and judicare, dealt only with the most routine criminal cases in which volume might be expected to produce cost savings (though, in fact, it did not). In Ontario's recent legal aid cost-control plan, implemented on April 2, 1996, legal aid for the low-end, more straightforward cases was reduced. Now a greater proportion of the criminal legal aid case-load than before is made up of high-end, more complex cases.

Of course, there are places in which the demand for legal aid is insufficient to sustain full-time staff. In small, rural, or remote communities, a judicare capacity may always prove to be less costly.

A further issue relating to the cost of delivery models is predictability. From the legal aid administrator's perspective, a staff model may be easier to manage because a set number of lawyers can be hired at set salaries, and a budget can be devised accordingly. However, this possibility leads to the concern that when cuts to government spending are made or when demand for legal aid increases, no new staff counsel will be hired, case-loads will increase, and the quality of representation will suffer. It is important to recognize that cuts to legal aid spending may have dramatic effects on service delivery regardless of the model in place; clients will either get poor service (if staff or judicare case-loads rise to unmanageable levels) or no service (if tariffs are reduced to a level which causes too few private lawyers to accept certificates or if the number of certificates is reduced).

Finally, the complexity of attempting to assess accurately the costs of one legal aid delivery model relative to another, not to say the incidence of disagreement among evaluators regarding the results, may lead one to consider whether there is a more productive basis for measuring the appropriate and necessary level of expenditures on legal aid. One option for evaluating criminal legal aid would be to measure the relative costs of state-funded defence work and those allocated to state-funded prosecutions. This is the approach taken in the U.S. National Center for State Courts study. Its methodology demonstrated across nine sites that legal aid matched up reasonably well in terms of compensation, but that prosecutors had significantly greater access to investigators and expert witnesses. This finding might be the basis for a more balanced and more practical evaluation of criminal legal aid costs than has been attempted previously in Canada.


Generally, the findings of the comparative studies demonstrate that there is little difference in the quality of services delivered by the different models. However, as is the case with costs, qualitative evaluations of delivery models are extremely difficult to conduct. There are many variables to control for. First, there is the difficulty of determining from whose perspective quality will be measured-that of the client, other lawyers, judges, legal aid managers, government, or funder. Even within those groups there may be variation in opinion as to whether a lawyer provided a good service for his or her client.

The second difficulty is determining which elements of service will be used to measure quality. Included in the list of options might be time spent per case, case outcome, whether or not the client received a custodial sentence, the sophistication of legal research and argument, the impact of the lawyer's work on the community as a whole, or how satisfied the client was with the lawyer's service. Each of these measurements has been used in one study or another, but the lack of consistency and the numerous variables make any conclusions regarding the quality of a particular delivery model questionable. That said, the results documented below do reveal some indication of the quality of service afforded by different delivery models. However, these results cannot be taken as conclusive.

One point worth making is that staff programs, as compared with judicare models, have the greater potential for quality-control. Since the majority of lawyers who participate in judicare schemes do so on a part-time basis, and thus the numbers are large (5,000 to 6,000 throughout Ontario), establishing and enforcing standards of quality can be difficult task. The more structured, employment-based operations of Staff Offices, on the other hand, lend themselves to case assignment based on experience, supervision, performance review, training programs, and client feedback.

Case Outcome

Case outcome does not necessarily reflect quality of service, but is one of the most common forms of measurement. It is interesting to note the general patterns that have emerged in studies to date. Almost all studies conclude that, although the rate of guilty findings is the same for both staff and judicare clients, staff counsel enter more guilty pleas at an early stage in the proceedings, and staff clients have lower incarceration rates. However, the Burnaby Report found that judicare clients received significantly more absolute discharges.

Continuity of Representation

An important aspect of legal representation is continuity. Once a client obtains a lawyer, the possibility that the solicitor-client relationship may end at any given point is no greater in a judicare model than in a staff model. However, the point at which the solicitor-client relationship may begin is likely to be earlier in a staff model. The experience of both the Burnaby Report and the Alberta Youth Offices Evaluation was that staff counsel made earlier contact with clients, often first meeting them in a duty counsel role and then working with them through disposition. Similarly, the Burnaby Report found that staff counsel remained with their clients through different stages of the proceeding more often than did judicare counsel.


One of the arguments frequently made in studies of legal aid delivery models is the perceived greater ability of staff counsel to specialize in a particular field of law. Clearly, if staff counsel spend all of their time doing legal aid case work, they will develop an expertise in this type of work. Crime may have "poverty law" aspects that permit a lawyer's specialized expertise to produce benefits for the client and for the system. Examples of such benefits include representation on bail reviews or knowledge of and access to community supports for non-custodial dispositions. Those benefits may be seen more clearly from the needs perspective of an alleged young offender than from that of an adult accused, given the greater diversity of resolution options and community supports available for youth.

In Manitoba, the three Winnipeg Staff Offices specialize in different areas of law: one specializes in criminal work, the second focuses on prison law and immigration cases, and the last does only "youth work"-young offender and child welfare cases. One purpose of this focus is to enhance the quality of service through specialization. Only a handful of private lawyers see the same number of legal aid clients annually. Again, it is worth emphasizing, as discussed earlier, that a private bar criminal defence lawyer with a mixed, private/legal aid practice might produce the same results because of his or her specialization in criminal law.

Moreover, the impact of specialization may not be as great in criminal matters as in family legal aid, where the whole dynamic changes when the client has a low-income. Purely family legal issues often form only one aspect of the shelter, health, educational, employment, safety, and social assistance issues which confront women, in particular, who seek redress within the legal system. Family law specialization may therefore be of vital importance to the quality of legal aid services.

In measuring staff and judicare models against time spent per case, the Burnaby Study found that one advantage staff counsel have over judicare counsel is the ability to achieve economies of scale and save travelling time. Other studies have suggested that lawyers from Staff Offices in a given community spend less time travelling between their offices and court because they are often able to handle more than one case per attendance at court. Another reason for underlying differences in time spent per case may be that staff duty counsel are able to negotiate more settlements than are judicare counsel, leading to fewer trials. The clearance rates demonstrated by staff duty counsel in the young offenders offices in Edmonton and Calgary also offer evidence of early resolution capacity. As was discussed above, this may be the result of staff counsel's ability to develop a better rapport with Crown Attorneys as a result of staff's regular attendance in the same court. On the other hand, critics may argue that excessively close rapport creates the danger that defence counsel's vigour and independence may be compromised.

It should also be noted that the measurement of time spent per case may be skewed by the way in which an evaluation is conducted. For example, if time spent per case for the judicare model is measured by the number of hours billed to legal aid, the results may not be entirely accurate. Patterns in Legal Aid 1994 reported that two provincial studies and one federal study of legal aid programs in three provinces found that judicare lawyers typically bill for the maximum allowable hours for each type of case. This finding could be interpreted to reflect either that judicare lawyers do spend more time per case or that their billing practices overstate the number of hours spent per case (particularly if tariffs are low).


The most frequently voiced concern about a staff model of legal aid delivery is that it is less flexible, and therefore more vulnerable to heavy case-loads, than are other delivery models. The concern is that counsel will be able to spend less time per case, and quality of service will suffer. There is an assumption underlying this concern which should be examined closely-namely, that less time spent per case necessarily means that the client is receiving an inferior service. This may not always be true; it may simply be that a staff delivery system is more efficient in certain areas of practice than are other models of delivery. Client evaluations have shown that, even where Staff Office case-loads are high, more guilty pleas are entered, and less time is spent on each case, clients still seem to be satisfied with the quality of service delivered.

A related issue is that, considering the case-load of the legal aid system as a whole, a judicare model may be perceived to more easily adjust to increasing demand for service than a staff system. Some argue that, in a judicare model, increased demand is easily absorbed by the legal community because there are enough lawyers to take on the work, whereas in a staff system adding lawyers to an office or opening another Staff Office is operationally more onerous. This argument is not wholly persuasive. In both situations, whether or not the system can absorb increased demand will depend on how much funding is available for legal aid delivery. In each system, if funding is available, more lawyers could be paid to do the required work. Meeting an increased demand for legal aid services is predominantly a question of resources, not of choice of delivery model. In this respect, one advantage of a mixed model of legal aid delivery is that legal aid administrators could mandate that, when staff case-loads reach a certain level, cases must be referred to the private bar by administrators.

Impact on the Poverty Community

"Impact work" can be defined as work which not only helps clients to improve their current legal situation, but aims to help prevent that situation from recurring by educating clients about their rights and responsibilities. The U.S. Delivery Systems Study defined impact work as "a project's achieved or expected results in terms of long-lasting improvement or avoidance of deterioration in the living conditions of significant segments of the eligible population".

The extent to which one believes that the impact of a legal aid delivery model is a qualitative issue depends, in part, on one's general view of the purposes and goals of legal aid. If one believes that part of the purpose of legal aid services is to help to educate clients about the legal system and the laws that affect them, then the ability of a particular delivery model to achieve this end is relevant to the quality of service being provided. On the other hand, if one believes that the purpose of legal aid is only to provide legal services to the poor on a case-by-case basis in the same way as they would be provided to paying clients, "impact on the community" may not have much to do with quality of service.

In the United States, the impact on the poverty community was the only criterion under which the various models were found to have significantly different results. The highest impact rating was achieved by the existing staff system. The lowest impact rating was given to a pure judicare model. Generally, staff, and particularly clinic, systems are better positioned to carry out impact work. This type of work can be done through the use of staff paralegals, community legal workers, or social workers, or by locating the Staff Office in proximity to low-income client groups. Judicare lawyers are not compensated for this type of work, and therefore tend to restrict themselves to more traditional legal work. Private-bar lawyers do little to help raise client's awareness of their legal problems. The Review of Legal Services in British Columbia suggested that a general consensus exists that private-bar lawyers had difficulty understanding issues of poverty, race, and gender. However, it has to be remembered that a great deal of impact work is carried out by members of the private bar either individually or through lawyer organizations such as the National Association of Women and the Law and the Criminal Lawyers' Association.


In its broadest sense, the phrase "access to legal services" could be interpreted to mean access to a legal aid delivery model that meets every legal need with full service. In the extreme, implementing anything less than the full range of legal aid delivery models could be seen as limiting client access. Given the limits to government funding for legal aid, the possibility that this full range of delivery systems can be implemented is unrealistic. The challenge is to find the mode of delivery that is most appropriate for particular communities of consumers or for particular areas of law. Increased use of paralegals, diversion, mediation, law students, and self-help mechanisms could go some way to increase people's "access" to legal services.

Each of the basic models of legal aid delivery has strengths and weaknesses which make it more or less appropriate for serving different client communities and different classes of legal problems. Some of the strengths and weakness of different delivery models with respect to access are set out below.


Although in some respects the judicare model may seem to be the most accessible to clients because it provides them with a wide range of choice in obtaining counsel, the judicare model has some practical drawbacks in serving legal aid clients.

First, the tariff rate will have a significant effect on the quality and experience level of counsel who will accept legal aid certificates. If tariff rates are low, the choice available to legal aid clients may become limited if more experienced, specialist counsel refuse to participate in legal aid or, even if registered on a legal aid panel, turn down many cases.

In addition, client access to lawyers in a judicare model stems, not from what the system is able to provide for the client, but from how the client perceives the service provided. The judicare model is based on the principle that legal services should be available to low-income clients in the same way that they are available to those who can afford them. This premise is based on a questionable assumption, that legal aid clients need the same type of lawyer and same type of legal service as non-legally aided clients. This is often not the case, particularly with respect to young offenders or family clients, where income level and the range of resolution options demand a wider and perhaps more creative response. Not only do the areas of law differ, but the type of lawyer or legal worker required to assist low-income clients can differ as well.

For first offenders or first-time users of legal services, and for clients facing a host of other difficulties in their lives, visiting a traditional lawyer in a traditional lawyer's office can be a daunting experience. The office may be in a different neighbourhood from that which the client is used to; it may be in a large corporate sky-rise; it may be business-focused and formal. Each of these factors can contribute to a client's discomfort with a judicare model and may be a barrier to clients' access to the legal service provided.

One benefit of a judicare model with respect to access to legal services is the relative ease with which such services can be provided to remote communities. For practical reasons, a judicare system seems to be the best way to provide legal services to communities which may not require them on a constant basis. For smaller, remote communities which may require the assistance of a lawyer only once a month, a full-time Staff Office or clinic is an unreasonable expectation. The practical answer may be to have lawyers doing legal aid work in nearby towns who are willing to fly into the community, as required, to address legal problems which may arise.

Staff and Clinic Models (with Supervised Paralegal Component)

As distinct as the structure of staff and clinic models may be with respect to the degree of managerial and service input available in the low-income community, the two models have some significant characteristics in common. Both are offices located in particular communities, both have the capacity to include the use of paralegals or community legal workers, and both use salaried lawyers to provide legal services. The National Council of Welfare report, Legal Aid and the Poor observes that decentralized Staff Offices are psychologically and physically more accessible to low-income people, especially if the office has strong community links. For example, the office may be located in the client's community, or in a nearby community with which the client is familiar, making it easier to get to physically and heightening the community's awareness of the existence of legal aid and the office. Establishing the office as a community service rather than as a service which many communities may use (judicare) may go some way towards increasing clients' use of or access to the service. Also, staff and clinic models are more accessible to legal aid clients because the atmosphere is often less formal than in traditional legal offices. A client's comfort level with the setting in which the service is provided will often lead to or increase that client's willingness to use the service. Finally, although this will vary with type of Staff Office, part of the mandate of community legal clinics is to engage in public legal education through community outreach work. Educating people about their rights often provides them with the confidence to exercise those rights. The ability of clinics to do this work increases clients' access to legal services because it makes them more aware of when those services might be required and useful.

Access and Other Legal Aid Initiatives

It is important in thinking about access issues to consider the numerous ways in which access to legal services could be improved. True access means that services will meet the varying needs of different legal aid clients. Young clients have different needs from elderly clients, clients with psychiatric difficulties have different needs from clients without such concerns, family law clients may require a different mode of legal service from criminal law clients. It is erroneous to assume that all legal aid clients will require or be able to use the same type of service. For example, assisted self-help legal services may work well with simple uncontested divorce cases. However, self-help services would not be appropriate for young offender matters or for women who have been the victims of domestic abuse. In the latter two examples, legal services combined with some element of social work would better serve clients' needs. Family law Staff Offices in Manitoba and youth offices in Alberta have recently been experimenting with the use of social workers or case workers in the delivery of legal services.


A frequently raised concern in discussions about the use of a staff model is that staff lawyers have less legal independence than do private counsel because the lines of control and payment by the government may be more direct.

Ideally, all lawyers should act independently, regardless of who is paying them or the nature of that payment. The threat to independence exists in any legal aid plan; a third-party, the government, is paying for the delivery of legal services to those who could otherwise not afford it. The nature of the payment may vary between the systems, but the fact of government funding remains the same. It is the fact of third-party payment, rather than the nature of the payment, that creates the potential for conflict when government funding is involved.

The area of law may also affect the perception of legal independence. For example, in family law, the government usually has no direct stake in the issues of the parties, but in criminal law, some would argue, there is a direct link between the client's fate, which rests in the hands of an employer within the broader public sector, and the Crown Attorney, also a public-sector employee. This potential conflict arises not only in the criminal area, but in most areas of "poverty law" and in a significant proportion of family law cases.

In a contract model of legal aid delivery, there is also potential for legal independence to be compromised. Contracting law firms will generally want to ensure that they will "obtain" the next contract. The element of competition in a contract model could make it vulnerable to political considerations, and compromise legal independence. Accordingly, with each of these models, it is necessary to build in protections of independence.


Legal aid commentators have observed that, particularly in Ontario's legal aid context, an individual's right to choose a lawyer is viewed as sacrosanct. The Law Society and many members of the legal community argue that the freedom to choose one's own counsel is one of the most important virtues of a judicare system.

The arguments about the lack of choice in Staff Offices are well known. In criminal cases where the client is faced with deprivation of liberty and the stigma of a criminal record, the quality of the relationship between lawyer and accused is vitally important; as it is in many family law matters, given the often sensitive and intimate nature of the issues involved. The full confidence a client must have in his or her lawyer can be best be assured, it is argued, if there is the ability to freely choose the most appropriately skilled and compatible counsel. To have to choose one of a limited range of staff counsel could compromise the result.

However, these arguments about choice of counsel assume that a judicare model provides clients freedom to choose the lawyer whom the client believes will best represent his or her interests. Many lawyers do not participate in the legal aid program and, of those who do, many accept a very limited number of cases. Moreover, clients may actually be disadvantaged in their process of choosing counsel by their relative lack of familiarity with the legal community. Low-income earners more so than middle- and upper-income earners tend not to be informed and skilled consumers of legal services. They tend to lack familiarity with professional relationships, and much of their lives is spent on more fundamental challenges of day-to-day survival. If a client has had some contact with the legal system before, he or she may know of lawyers who may be appropriate representatives in his or her current situation. In both the criminal and the family contexts, word-of-mouth in detention centres, or referral lists provided by women's shelters, may provide some assistance in deciding which counsel to choose. However, even with these recommendations, clients have a limited basis on which to judge whether a particular lawyer is best for them. There are few objective criteria to which judicare lawyers are held which would help clients assess whether a lawyer will adequately represent them in their current predicament.


In our view, many of the above delivery-model comparisons are highly suspect in terms of their generalizability, primarily for the reason that the features of interest being compared are not preordained but endogenous to the design of the systems in question, so that these design choices largely predetermine the outcome. For example, a Staff Office is likely to appear more expensive on a cost-per-case basis, at least in the short run, if legal aid tariffs for judicare lawyers are set unrealistically low. Conversely, if judicare tariffs are set very high, a Staff Office will appear cheaper. Similarly, if staff lawyers are paid very generously, and are not asked to carry a substantial workload, the Staff Office is likely to seem more expensive than a judicare delivery model. Conversely, if staff lawyers are paid badly and expected to carry an excessive case-load, at least in the short run, the Staff Office is likely to seem cheaper than a judicare system (depending on the tariff structure).

Thus, a legal aid authority, in choosing how to design the system, can, at least in the short run, generate any outcome it wishes in terms of the relative cost of alternative delivery models. However, in the long run, it is not as unconstrained. Here it needs to be sensitive to the complex interaction of several variables. In setting judicare tariffs, the legal aid authority obviously must be sensitive to the opportunity costs of lawyers undertaking legal aid cases, and hence forgoing privately paid legal work. If the tariff is set significantly below the level of these opportunity costs, then mostly very young lawyers or older and unsuccessful lawyers with low opportunity costs are likely to be attracted into the judicare segment of the legal aid system (while recognizing some role for altruistic motivation in undertaking legal aid work), having predictable effects on service quality. We note here, in passing, that the legal aid tariff in Ontario has not been increased since 1987, and in many cases has been sharply reduced, raising important questions as to the long-term sustainability of current tariff levels, without inducing the predicted quality effects.

Similarly, in setting staff lawyers' salaries and benefits, and determining case-loads, the legal aid authority should be sensitive to the opportunity costs faced by staff lawyers in forgoing opportunities in the judicare market or in privately paid practice, making appropriate adjustments for greater job security, less risk of income fluctuations, and perhaps differences in working hours. Setting staff salary and benefits too low and workloads too high will mean that staff positions will be attractive only to individuals with low opportunity costs, which will have equally predictable long-run effects on service quality. Yet again, staff legal aid lawyers' salaries and benefits cannot be set without taking into account Crown Attorneys' salaries, which in turn need to take into account the opportunity costs of their alternative career options, if these positions are to be attractive to high-quality service providers.

If appropriate financial relationships in a mixed legal aid system can be fashioned among the private legal services market, the judicare market, and the staff market, in the sense that lawyers of comparable quality are attracted to the three sectors, then it becomes possible, in principle, to make cost comparisons that are not an artifact of arbitrary or inappropriate determinations of the legal aid tariff or staff salary and benefit levels, but of underlying relative efficiencies associated with different delivery models in different contexts. These comparisons would focus on the relationship between units of inputs of different kinds required to achieve a given quality and quantity of outputs under different delivery models in different delivery contexts. Given the financial interrelationships that should obtain among these three markets (the private, paid legal services markets; the judicare market; and the staff market), it would be surprising if, in general, relative cost differences were large. This seems consistent with the data reviewed above.

For similar reasons it would also be surprising if major general differences in quality of services emerged. This view reinforces an observation made at the outset of this chapter, that analysis of relative efficiencies of different delivery models is highly context-specific. For example, it seems obvious that in handling large volumes of routine cases, for example, first appearances, adjournments, and bail hearings in criminal proceedings, economies of scale and specialization are likely to render some form of duty counsel/Staff Office more cost-effective than a judicare system, given the obvious transaction costs, public and private, entailed in handling these matters on a discrete, contracting-out (certificate) basis. For larger, more complex, and unique legal proceedings, it would be surprising if these cost advantages of a Staff Office were to persist; furthermore, for these kinds of proceedings, quality disadvantages may afflict the Staff Office in terms of access to private counsel with proven expertise in the particular case in question, although Staff Offices are more amenable to systematic quality-control mechanisms, and in practice (as noted above) the effectiveness of the judicare system is affected by availability of counsel and the types of information problems facing the client. However, those differences are largely conjectural on our part and await confirmation or refutation from real observed experience with different delivery models in different contexts.

In part because of this caveat, we favour setting up Staff Offices with a full range of responsibilities for service delivery in particular areas of the law (e.g., criminal law, family law, and immigration law), in competition with the judicare system, so that conjectures can be tested as to where the comparative advantages and disadvantages of Staff Offices really lie. In addition, in order to attract high-quality personnel to a Staff Office operation, it is important that there be significant potential for career development through the opportunity for younger and less experienced lawyers to work with senior and more experienced lawyers on more complex cases. Senior lawyers, in turn, are unlikely to be attracted into a Staff Office without the ability to undertake more complex cases, and younger lawyers will feel that their career-development path is likely to be undesirably truncated if they are confined to working on large volumes of routine cases. In the end, quality of service, motivation, and commitment may be undermined in a Staff Office that, as a matter of policy, is permanently confined exclusively to large-volume, routine matters. The challenge for institutional design then becomes to define the range of large volume, routine matters in which clients may not be given the choice of counsel, owing to the superior cost efficiencies of a Staff Office in relation to such matters, and to establish some appropriately defined (by context) seriousness or complexity threshold above which clients may be offered the choice between Staff Office lawyers and judicare lawyers.


In choosing among delivery models, and designing their major features, an economic perspective is helpful in illuminating the incentive properties of alternative design choices. While this perspective has rarely been applied to legal aid systems, it has been frequently and fruitfully applied to health care and other social programs. We sketch the focus of this perspective below.


The basic economic analysis of market structure assumes a homogeneous good or service whose quality is standardized, or at least easily observable. But many goods and services cannot be evaluated in this way. Automobiles and houses are merely two examples of goods whose quality, though of critical importance to the buyer, is not readily observable. Similarly, the effectiveness of professional services and of employees depends largely on the effort expended by the seller, that is, by the professional or employee himself or herself.

This type of situation creates what economists call a "principal-agent relationship". The words "principal" and "agent" are used here in the economic sense, not in the legal sense. For economists, a principal-agent relationship arises whenever one individual (or organization) is supposed to act in another's interest. The relationship is problematic whenever the agent has both the incentive and the ability to choose actions that are not in the principal's interest. If the agent's interest is perfectly aligned with the principal's, or if the effect of the agent's actions on the principal's interests can be perfectly observed, then there is no agency problem. But if their interests are not perfectly aligned, and if the principal is not fully informed about the agent's behaviour, then a principal-agent problem arises. Asymmetric information between the principal and the agent creates the opportunity for the agent to depart from the principal's interest; and, when these departures occur, the market will generally come into equilibrium at some point other than the optimal price/quantity combination that would occur if agency was perfect. As we will see, both asymmetric information and differences in incentives are typically present in the legal aid context.


Suppose that the government or the legal aid authority has established a set of criteria for determining which cases should be funded and to what extent, and has chosen a mechanism for funding those cases. The authority might, for instance, have determined that funding will be allocated only to criminal cases with sufficiently serious consequences on conviction or only to family law matters that have a certain degree of complexity. The authority wishes to obtain a given quality of service for legally aided clients while staying within its budget. What is to prevent the authority from simply announcing its criteria to prospective clients and to the legal profession, and observing whether the criteria are being respected?

If every legal matter came with an unambiguous mark of its seriousness or complexity, then the legal aid authority could indeed just sit back and watch its budget be spent wisely. But legal matters are not like that. Before the event, no one really knows for certain whether a given case is going to be serious or trivial, complex or simple; the best one can do is look for indicators of the underlying characteristics of the case. Thus, for instance, the funding criterion for criminal cases cannot be that the accused will go to jail if convicted, because, given the great range of circumstances of accused persons and the discretion of the sentencing judge, before a trial it is impossible to know whether imprisonment will be the result of conviction (unless the offence has a statutorily required minimum term of imprisonment); rather, the funding criterion must be based on an indicator of the seriousness of the consequences of conviction, such as the nature of the offence charged, or a likelihood of imprisonment or loss of livelihood on conviction.

In addition to uncertainty before a trial, there is imperfect information after its conclusion. Typically, the lawyer is in the best position, after the event, to determine whether a given case actually did meet the criteria established by the legal aid authority. Importantly, the lawyer's information on this issue is superior both to the client's and to the legal aid authority's. The client is unlikely to be able to determine the nature of the case with certainty. The authority, though presumably more sophisticated than the client, must also rely on the lawyer's assessment of the nature of the case. There are at least two reasons for this reliance. First, the authority must to a large extent respect the confidentiality of the lawyer client relationship; and, second, even if the lawyer could be required to share all of the case information with the authority, it would be prohibitively expensive for the authority to review each and every file to determine compliance with the criteria.

Thus, in the legal aid setting, the lawyer has two principals, the client and the legal aid authority (and arguably a third-the Law Society-with respect to professional standards). Furthermore, each of these principal-agent relationships has its own informational asymmetry. The legal aid lawyer is, in effect, a "double agent". On the one hand, the lawyer has both a duty to represent the client's interests and superior information about how to serve the client's interests: the lawyer's framing of the pros and cons of any given course of action is likely to be very influential in the client's decision about whether to pursue that course of action. This sort of authority problem is not limited to the legal aid setting: it arises whenever a lawyer wants to proceed with the case in a way that the client, if fully informed, would not choose. On the other hand, the lawyer is the agent of the legal aid authority in that he or she has a responsibility to respect the funding criteria established or enforced by that authority: the legal aid authority may not wish to fund every procedural step that the lawyer would have recommended to a private client, but the authority is largely dependent on the lawyer's assessment of whether any given step is warranted in any given case. Importantly, as is true in other third-party-payer systems (such as medicare), this second authority problem arises even if there is no information asymmetry between lawyer and client: both have incentives to seek funding for more extensive representation than the legal aid authority, if fully informed, would be prepared to fund.

An agency relationship creates room for behaviour that is not in the principal's interests. The excess costs that arise because of imperfect agency are often called "agency costs". Agency costs have been extensively studied in corporate governance, in employment relationships, and in the provision of medical services, although much less so in the provision of legal aid. The two main categories of agency costs relevant to legal aid are supplier-induced demand and quality decline.

Supplier-Induced Demand

As noted above, economists generally think of demand and supply as separate functions that interact to determine the price and quantity of a good. But when the supplier of a good is also the agent of an imperfectly informed customer, demand and supply are no longer clearly separated: where the customer's decisions depend on the supplier's advice, there is scope for supplier-induced demand.

Supplier-induced demand has been defined in various ways, but the central idea is that a professional advisor can, because of his or her informational advantage, increase the amount of services that the client would like to use. The mere fact that the impetus to use the services comes from the professional and not from the client is not in itself necessarily a cause for concern; we assume, by and large, that professionals will honestly advise clients as to the appropriate course of action. Thus, services obtained through supplier-induced demand should be considered a cost only when they exceed the quantity of services that have been deemed appropriate, according to some normative criteria. In the context of legal aid, the appropriate normative criterion to apply is not the competitive market outcome; presumably, we have a system of legal aid because we believe that the market, left to itself, does not supply enough legal services to low-income earners. Rather, the question is whether the legal aid authority, if fully informed, would have funded the service. If not, then the authority's demand for the service is supplier-induced.

The literature on supplier-induced demand in health care is extensive. A substantial body of studies supports the position that supplier-induced demand is pervasive in medical and dental care. This sort of supplier-induced demand obviously raises serious problems for cost containment and for economic efficiency.

In a legal context, it is hard to imagine that the number of matters, or the number of clients potentially in need of advice or representation, could be affected by the availability of legal aid; therefore, supplier-induced demand for legally aided services derives primarily from two sources. First, lawyers may play an active role in persuading the legal aid authority that a given case, or class of cases, should be funded. This source of supplier-induced demand will be particularly important where opportunities for lawyers to practise privately are declining (or increasing at a slower rate than the supply of lawyers). Economists would normally expect competitive forces to push lawyers' fees down under these conditions, but if lawyers can persuade the legal aid agency that the existing tariff is necessary to provide proper representation, and that additional cases should be funded, this downward pressure is reduced. Second, lawyers have an important influence on how much work is done on any given case. It is important to note that, in both of these sources of supplier-induced demand, the client's interest is the same as the lawyer's; while it may occasionally happen that the client would be better off with fewer legal services or with no representation or advice whatsoever, in most cases the client will want the lawyer to do the best job possible at the legal aid authority's expense.

Quality Decline

One of the central manifestations of imperfect information in markets of all sorts is the problem of observing quality. The quality of many goods is not observable until the customer has had time to use the good (or, in the case of some goods, has actually consumed it). While there are many mechanisms, ranging from reputational devices to formal warranties of quality, to compensate for the difficulty of observing quality, there are also many markets in which these mechanisms are likely to be ineffective. Where a principal depends on an agent to obtain a good or a service, the problem of controlling quality is exacerbated: the agent may have an incentive to provide lower-quality work than the principal is expecting to receive.

The quality of a lawyer's work is largely under his or her control; thus, a second type of agency cost that may be important to legal aid is the possibility that the quality of work a client receives may not be at the level the legal aid agency is paying for. But the quality of legal services is difficult to measure. Quality includes not just the result achieved or the technical quality of the legal work performed (e.g., proper drafting of documents or competent representation in court), but the lawyer's ability to inform the client and to respond to his or her instructions. A lawyer who promptly returns necessary phone calls is a better lawyer than one who has to be hounded by his or her clients; a lawyer who enters a guilty plea only when the client is fully prepared to admit guilt is a better lawyer than one who participates, out of laziness, indifference, or financial motives, in a plea of convenience entered by an accused who privately maintains his or her innocence. From the legal aid authority's perspective, both guilty pleas look the same, and the only available measure of the quality of representation on the plea may be whether the client got a reasonable sentence. But, in reality, the second plea was obtained through low-quality service, even though the result is the same as in the first plea. The second plea may result in the lawyer being paid by the legal aid authority for high-quality work without having performed it.

In a market setting, there is likely to be some measure of control over such discrepancies in service quality. Clients will not remain with lawyers who do not treat them with respect or who disregard their instructions; sophisticated or repeat clients may have a basis for comparing the quality of service that they have received on different occasions. Again, as with supplier-induced demand, such control is unlikely to be perfect. But, in the legal aid relationship, where the lawyer is a "double agent" and where both the client and the legal aid authority may lack the ability to observe quality, even these controls may be lacking, and low-quality service may be a severe problem.


In this section, we discuss how different delivery mechanisms tend to control or to give scope for different agency costs.


The judicare model is commonly assumed to be the most amenable to supplier-induced demand. Economic analysis supports this assumption. There are, as noted above, two distinct points at which supplier-induced demand might be relevant to judicare. First, the number of lawyers willing to accept legal aid certificates is likely to increase the number of certificates issued. One might expect that more certificates will be issued when clients have an easier time finding a lawyer; for example, clients who have previously obtained legal aid and had satisfactory service will be more likely to apply for legal aid for subsequent legal problems. More important, one would expect lawyers to encourage prospective clients to apply for legal aid certificates, again causing more certificates to be issued. In a normal market, economists would expect an increase in the supply of lawyers to reduce the price of legal services; but, in the market for legal aid, customers do not have the usual incentive to seek out the cheapest supplier of the good, so that the usual effect produced by an increase in supply on price will not be forthcoming. The legal aid authority could attempt to take advantage of market conditions by lowering the tariff, but this is more difficult to do when the tariff is centrally determined, particularly if the authority is a branch of the Law Society, than in a decentralized, competitive market. Second, in any given case, the judicare lawyer has an incentive to take whatever procedural steps recommend themselves on legal grounds, or at least to attempt to persuade the legal aid authority to fund such steps.

The tariff structure may also have an important influence on the amount of supplier-induced demand under judicare. Roughly speaking, judicare could have three basic tariff structures. First, lawyers could be paid purely on an hourly basis. This tariff structure obviously creates the greatest scope for supplier-induced demand, though it may be necessary in lengthy or complex cases to permit lawyers to bill by the hour. Second, lawyers could be paid block fees so that the lawyer rather than legal aid authority bears the cost of any work beyond the value of the block fee. Supplier-induced demand may be somewhat mitigated by block fees, in that there is no financial incentive to work additional hours when compensation is fixed, although block fees may create incentives for suppliers to economize on quality. Further, block fees do not remove the incentive to get new clients into the legal aid system. Third, lawyers might be paid by the hour, with a cap on the number of hours for each type of case. This tariff structure provides no incentive not to work the maximum number of hours, but it does mitigate some types of supplier-induced demand, in that the lawyer and the client must decide how to spend the hours that are available for each case, rather than trying to persuade the legal aid authority to fund every step. On the other hand, to the extent that lawyers are essentially expected to do the work beyond the cap without remuneration, quality decline becomes an issue.

The consequences of supplier-induced demand when the legal aid authority's budget is fixed are simply described: legal aid cases will be funded according to the ability of lawyers to persuade the authority that their cases meet the criteria, rather than according to the criteria themselves. When the authority's budget is exhausted, some cases that ought to have been funded will not be.

As noted above, an impressive body of research supports the proposition that supplier-induced demand for medical and dental services does occur. We are not aware of any properly conducted quantitative study of supplier-induced demand for legal aid. The rapid increase in Plan costs from the mid-1980s to the mid-1990s, including dramatic increases in costs per case for most classes of cases, may seem to indicate that supplier-induced demand has occurred in Ontario, particularly since, during this period, the legal profession was growing more rapidly than the opportunities for private retainers, and the tariff was basically constant (so that the growth in costs reflects an increase in the quantity of legal aid services provided, not an increase in the price of those services). But, even here, other forces are involved. The increasing complexity of criminal matters (due to the Charter of Rights and Freedoms) and family law matters, the explosive growth of certificates for immigration matters, and the effect of the recession on the demand for legal aid services were undoubtedly responsible for much of the growth in legal aid costs. Further detailed empirical analysis would be required to sort out the relative importance of supplier-induced demand and the other forces at work.

The other type of agency cost discussed above was quality decline. There are competing views about quality decline under judicare. On the one hand, one of the normal market mechanisms, the client's ability to exit, will operate under judicare. To the extent that clients are able to choose among lawyers willing to take a certificate, clients will tend to select those lawyers who appear to be offering high-quality service (assuming such lawyers are willing to take legal aid cases). While the nature of the agency relationship between lawyer and client means that the client will never be able to assess quality perfectly, some dimensions of quality can be assessed quite readily even by first-time clients (e.g., Does the lawyer return my phone calls? Do I trust this lawyer with the intimate details of my life?), while other, more technical, dimensions of quality can probably be assessed to some degree by those clients who already have some experience with the legal system. On the other hand, if judicare lawyers overload their practices to the point where they cannot spend adequate time on any one case, and if clients are unable to exit, because, for instance, they are unable to make any informed assessment of the lawyer's ability, or few lawyers in the area take legal aid for that kind of case, quality is likely to decline precipitously. This type of quality decline is more likely to occur with block fees than with hourly payment. The degree of quality reduction under judicare depends on the relative magnitude of these two effects.

Staff Models

The staff model encompasses at least three methods of delivering legal aid services, each model having very different institutional characteristics. These are:

  1. The pure or direct staff model, in which lawyers carry cases much as in the judicare model, but are salaried employees of the legal aid authority. Typically, under this model, clients have little or no ability to choose counsel, and lawyers have little or no ability to choose cases. The public-defender offices in many U.S. jurisdictions are examples of the pure staff model.
  2. The duty counsel model, in which lawyers do not carry cases and typically do not offer representation at trial, but instead offer summary advice and/or representation in non-trial court appearances. Duty counsel offices are typically run by the legal aid authority and are located in or near the courts; lawyers and clients typically do not establish a traditional solicitor-client relationship. The lawyers who work in duty counsel offices may be full-time employees, or may be private practitioners who are on a rotation to supply services through the office on some form of retainer basis.
  3. The community legal clinic model, in which legal representation in criminal and family matters is not the focus. While the community legal clinic will always have lawyers on staff, much of the work of advising and assisting clients will be carried on by non-lawyers (in Ontario, community legal workers), who will develop some knowledge and expertise in areas of law that are relevant to the clinic's clientele. In addition, community legal clinics are expected to carry out forms of advocacy that go beyond the representation of individuals' problems to more systemic concerns.

The differences among these three models are obviously important, but for the purposes of a general discussion of agency costs it is not necessary to distinguish them. We will therefore describe the effects on agency costs of a pure staff model.

Supplier-induced demand is not likely to be a serious problem under the staff model. Since staff lawyers are on salary, they normally have no incentive to seek out clients or to do unnecessary work on any given case. This conclusion might not hold if, for some reason, a particular clinic were overstaffed; the staff might then have an incentive to seek out work in order to justify their complement. But this situation is unlikely to arise in the current fiscal climate.

On the other side of the coin, rather than supplier-induced demand, one might expect that staff lawyers would resist increases in their workloads, given that their compensation is unlikely to be tied closely to output. Given an acceptable level of quality, the staff lawyer and the legal aid authority may have different views about the appropriate number of cases for the staff lawyer to carry, and this difference can be thought of as another source of agency costs.

Quality decline may be a greater concern under the staff model, for two distinct reasons. The first is the monitoring problem that has been much studied by agency theorists. Neither the client nor the legal aid authority can easily assess the quality of the work done in any given case; the client lacks legal expertise, while the authority has neither the resources nor the information on which to base a judgement in each and every case. Second, even if straightforward shirking can be contained through a combination of a sense of professionalism, quality-control mechanisms, and benchmarking of the performance of Staff Offices and personnel against one another, lawyers in the staff model may be prone to "burn-out" if burdened with excessive case-loads or mostly high-volume, routine, professionally unchallenging work roles. The legal aid authority has a fiscal interest in getting as much work as possible out of staff lawyers, so there will be a tendency for their work to be intensified. Salaried lawyers are likely to respond by reducing the amount of work done on each case to cope with the increased volume. This process will presumably continue to the point where the lawyers, taking into account their earnings, their hours, their opportunity costs, and their sense of professionalism, are indifferent about staying at the Staff Office and leave to practise elsewhere. As a matter of principle, there is no reason to think that this point will be consistent with the quality of work desired by clients or by the legal aid authority itself.

It has been suggested that staff lawyers are likely to develop a degree of specialized expertise that enables them to do better work in a shorter time than judicare lawyers. Further, it may be appropriate to give clients a degree of voice in the staff model, by representation in the office's governance structure or through some procedure for dealing with complaints. Moreover, performance evaluation and quality-control mechanisms may be more easily implemented in a more hierarchically structured Staff Office than in more individualized service provision under judicare. These features may well reduce agency costs in the staff model. On the other hand, the inability of clients to choose counsel in some staff models removes one of the standard competitive devices for controlling agency costs. Nor can professionalism or expertise alone prevent burn-out; indeed, professionalism may exacerbate burn-out, in that the lawyer's first response to an intensification of work may be to attempt to maintain quality in handling all matters that he or she is responsible for.


A third model, which has been tried in limited ways in Manitoba and the Yukon, and is an important component of proposed reforms in England and Wales, is the contracting-out of cases. The former U.K. government's report on legal aid, Striking the Balance, proposes that most future legally aided services in England and Wales should be provided under long-term (three- to four-year) block contracts under a competitive tendering process with private law firms. In addition, classes of activity (e.g., duty counsel services) would also be contracted-out under a competitive bidding process. Under the first variant of this model, the legal aid authority bundles together a block of cases with similar characteristics and seeks bids from private law firms. The authority will presumably accept the lowest bid that it believes is consistent with desired levels of quality. The clients in these cases will then be represented by lawyers from the firm whose bid is accepted by the authority. A block contracting system can be thought of as combining features of both judicare and a Staff Office-or a time-limited Staff Office with private lawyers compensated on a piecework basis.

Contracting-out may be attractive to the legal aid authority because it addresses some of the agency costs created by the relationship between the authority and legal aid lawyers. A rational law firm will presumably tender a bid that is high enough to cover its expected costs of doing a good job on the average case, but low enough to obtain the contract. If the market for legal services is reasonably competitive, the firm must therefore reveal to the authority the true value that it places on the opportunity to take these cases. Furthermore, since the firm bears the risk that the particular bundle of cases it is assigned could be more costly than average, but receives a benefit if the bundle is cheaper than average, the rational bid will also reflect the firm's degree of risk aversion. The authority can therefore control its costs by placing this risk on the firms that are most willing to bear it; less-risk-averse firms will tend to put in lower bids. Thus, it is most unlikely that there will be any supplier-induced demand in a system of contracting-out.

In contrast, the problem of quality decline is not eliminated by contracting-out. Once the rational law firm has the contract, its incentive is to treat every case as a simple one (e.g., in criminal cases, to encourage guilty pleas): since the firm's gross income from this group of cases is now fixed, it can improve its profit only by reducing costs, and perhaps economizing on quality. This agency cost will be exacerbated by the client's inability to change law firms.

But contracting-out need not always imply shabby service. A key feature of contracting-out is that it is not a permanent arrangement; the legal aid authority will regularly have new blocks of cases to auction off, and will presumably not grant contracts to firms that have persistently failed to meet prescribed levels of service quality. Although the legal aid authority will never be able to measure the quality service perfectly, there may be some indicators, ranging from the results achieved by the firm for the clients to the responses of clients, that the authority can use to estimate whether the firm has been providing good quality. In other words, the firm's interest in maintaining its reputation with the authority (not to mention with private clients and with other firms) acts as a constraint on shirking and other forms of quality decline.

Another method of controlling quality in contracting-out might be to make the contracts non-exclusive. Rather than having one firm solely responsible for a given block of cases, the legal aid authority might award the contracts to several firms and allow clients to choose among them. This approach would capture the virtues (if any) of choice of counsel in controlling quality, while retaining the competitive character of contracting-out; but it would likely be possible only in larger urban centres, given that blocks of contracts need to be large enough to permit realization of economies of scale and specialization.

Given the relative paucity of experience with block contracting in the legal aid context, we believe that utilization of this delivery model should proceed cautiously and on an experimental basis. In contrast to the former U.K. government's proposals, we do not see block contracting as a panacea for existing deficiencies in the Ontario legal aid system; it may well prove yet one more futile attempt to discover a single "holy grail", in terms of delivery models, for the provision of legally aided services. Empirical evidence suggests that, in general, staff models and judicare models perform comparably in most dimensions, and that each is superior to the other in some contexts. We have no similar evidence to rely on in a wholesale shift to block contracting. In particular, the modalities of the contracting-out process need to be carefully worked through. For example, the legal aid authority, in soliciting tenders on blocks of cases, could pre-set the price per case and invite service providers to submit competing bids based on quality assurance commitments that would, of course, require prior evaluation for appropriateness and post-service monitoring for compliance. Alternatively, the authority could specify in its calls for bids various quality-assurance conditions that it deemed appropriate and invite bids that would compete primarily on price, although the authority would still need to satisfy itself that the quality commitments being made were credible, given the size, experience, and so on of the bidders, and would need to engage in post-service monitoring to ensure compliance with these quality conditions. A third option would be to call for bids on blocks of cases with no price or quality conditions predetermined, with winning bids selected on the basis of the authority's judgment of which bidder presented the most attractive cost-quality trade-offs. Again, as in the previous bidding modalities, prior evaluation of the credibility of quality commitments would be required, as well as post-service monitoring of adherence to these commitments. However, because the selection criteria under the third option, in particular the trade-off calculus between cost and quality, are not specified in advance, the authority is likely to be vulnerable to charges of subjectivity, favouritism, or worse, in its allocation of contracts.

Some of the prior screening and post-service quality-verification problems noted above might be reduced for all three options under a system of franchising, as is currently being implemented in England, where only firms which had precommitted to adherence to a specified set of quality-assurance conditions would be eligible to enter bidding competitions for contracts. Franchising is conceptually distinct from block contracting in that meeting franchise conditions may be a prerequisite for participation in the provision of legally aided services under any delivery model (analogous to setting conditions to be met for registration on a legal aid panel in Ontario), or for more favourable terms of participation. However, under the former U.K. government's proposals, franchising would become an integral feature of block contracting. Franchise conditions for block contracts would cover "inputs" (requirements related to the qualifications, experience, and training of service providers); "structures" (requirements relating to the office environment and management, systems, including case management, strategic and financial management, and personnel management); "process" (requirements about the steps, or transactions, to be taken on the file); "outcomes" (requirements related to the level of client satisfaction with the services, the cost and time taken per case, and the case result). Poorly performing firms may lose their franchise status and be disqualified from bidding in future competitions, and hence face strong incentives to perform at acceptable levels in order to minimize or avoid this risk. Another complexity in the contracting-out process is likely to arise in relatively thin markets, where a firm has won an initial contract and developed or reinforced its ability to realize economies of scale and specialization by virtue of the contract; at contract renewal time, the legal aid authority may find no or few competing suppliers in the market, rendering it vulnerable to "hold-up" by the incumbent firm by being forced to continue dealing with it on its terms. Another unresolved question with contracting-out models is the potential scale of transaction costs to the authority in administering the contract-allocation and compliance process relative to the transaction costs of administering a certificate system.



In the course of public consultations undertaken by this Review and in the course of the research undertaken for it, opinions were often expressed by judges, practitioners, and others as to the high degree of variability in quality of legal services provided by lawyers under the judicare program in various areas of law (e.g., criminal, family, and particularly immigration law), and to a lesser extent by community legal clinics. Most of the evidence supporting these opinions is anecdotal and impressionistic. Nevertheless, we are persuaded that the new legal aid system must assign a high priority to developing effective performance measures and quality-control mechanisms for all delivery models employed by the system.

Defining "quality" is not as simple as it may seem. First, "quality" is, in substantial part, a function of the goals adopted for the legal aid system and the particular delivery models within it. For example, if an important part of the mission of the system is effective communication with clients of varying cultural backgrounds, community outreach, public legal education, and law-reform initiatives, these must be included in any quality-control and performance criteria. Obviously, these objectives may have little or no relevance to a staff duty counsel operation, where "quality" would need to be defined more narrowly to apply to case handling and summary advice at preliminary stages of proceedings, or to the judicare model, where quality again almost certainly takes on different connotations, depending upon the context.

Even if "quality" can be defined in particular legal aid contexts by reference to agreed-upon criteria reflecting the Plan's objectives in each of these contexts, observing, measuring, and evaluating performance is itself a challenging exercise. Requiring judicare lawyers, Staff Officers, community and legal clinics, or block-contract lawyers to report case outcomes so that these can be compared with those of other providers of similar services is one option, although standardizing for the different nature of case-loads from one service provider to another (reflecting the fact that one service provider may be handling more difficult or problematic cases than another) is problematic. Another problem is that, in some areas of law, proceedings do not yield easily measured outcomes (e.g., in many family law matters). Another approach is to rely on consumer-satisfaction surveys, but it is the nature of things in many areas of law that consumers are often ill informed and unsophisticated, and their judgment about service quality must be accepted with considerable caution. For similar reasons, relying on a reactive, client-driven complaints-investigation system is unlikely to be an effective response.

Notwithstanding these problems of defining and measuring quality of service, it is possible to state, as a general proposition, that problems with quality of service may result from: (a) the lack of access to services (eligibility criteria); (b) the wrong mix of services (e.g., exclusive reliance on legal services); (c) service providers who lack appropriate training and expertise (education, experience, judgment, or cultural sensitivity); (d) negligent performance (careless or inadequate case preparation); (e) deliberate wrongdoing (e.g., charging extra fees to clients above the legal aid tariff); (f) inability to identify individual or systemic problems (ineffective reporting, supervision, or monitoring systems); (g) inability to correct known problems (ineffective training, mentoring, or sanctions).


Existing quality-control measures for the provision of legal services are primarily those adopted generally by the Law Society, the self-regulatory body for all lawyers practising law in Ontario, and are not specifically adapted to the provision of legal services under the Plan. The principal components of the quality-control measures adopted by the Law Society entail entry standards for admission to the profession; rules of professional conduct; and the complaints and discipline process. While these quality-control measures have the advantage of entailing no direct costs to the public, they are subject to severe limitations in a legal aid context. First, entry requirements are of a quite general character and do not test for competence to undertake specialized areas of practice. Second, rules of professional conduct primarily focus on ethical issues, and only to a minor extent on defining "best practices" to be followed in particular areas of law or particular proceedings. Third, the complaints and discipline process is largely reactive to complaints by consumers or other service providers and for the most part focuses on various forms of misconduct (e.g., misappropriation of trust funds) or egregious forms of delinquency, rather than failure to adhere to competent standards of practice. Fourth, the Law Society's quality-control measures do not extend non-lawyers (e.g., paralegals, interpreters, or translators) who often play important roles in the delivery of legal aid in various contexts.

With respect to the community legal clinics, only recently has the Clinic Funding Committee begun to evolve a quality-assurance program and the process of evolution and adoption has been slow and contentious. The main components of this program are:
(a) peformance standards for five key work processes:
1. board governance and overall management
2. understanding the community
3. program planning, development, and evaluation
4. communications
5. services (legal-file management, summary advice, law reform, and community development);
(b) a system to monitor compliance with the standards (periodic reports, audits, interviews) and to assess performance against customer expectations and program;
(c) a system to respond to identified quality problems (education, mentoring, sanctions); and
(d) identification of best practices and continuous improvement through incorporation of best practices into performance standards.


Effective quality-control measures aspire to provide performance standards for all legal services; provide a system for monitoring performance on an ongoing basis; identify best practices and include them in performance standards and make continuous improvements to them; and apply the standards to all service providers. Potential disadvantages of such a regime are that it relies heavily on the accuracy and appropriateness of performance standards and may be insensitive to deviations from these standards required by the particularities of given cases; may entail significant administrative costs in defining and revising standards, monitoring adherence to them, and enforcing remedial responses or sanctions in the event of deficiencies; may have negative and unintended impacts on professional practice (e.g., by requiring lawyers to spend too much time filing outcome or compliance reports, or completing checklists); and may be more difficult to apply to service providers outside an institutional setting, in particular, private lawyers acting on a case-by-case basis under the auspices of the judicare system.

Notwithstanding these difficulties in the implementation of an effective quality-control and performance-evaluation regime, we strongly believe that a major priority of the new legal aid system should be to implement, over time, such a regime. The basic elements of such a regime would seem to entail the following:

  1. Improved standards for qualifying as a provider of legal aid services. In our view, this would entail the legal aid system administration specifying minimum professional qualifications for registration on legal aid panels in particular areas of law. It is important here not to be too rigid or unidimensional in specifying these minimum requirements. Qualifications for registration on a legal aid panel might be obtained in a variety of ways. In order not to create unnecessary barriers to younger and less experienced practitioners becoming active in the Plan, weight could be attached to service in legal aid clinics during law school or to specialized courses or training programs taken during or after law school. For other practitioners, years of practice, with some minimum percentage of practice activity allocated to the area where registration on a legal aid panel is sought, courses taught or taken, or complex trials handled might be appropriate criteria. Some form of point system might be devised that enables the requisite number of points for qualification for registration on a legal aid panel, to be acquired in various ways. Graduated criteria might be adopted for registration on a panel, with higher standards required for eligibility to undertake more serious cases of a particular type. Continuing-education requirements might be imposed as a condition for maintaining registration on a panel. Similar requirements should, in principle, apply to lawyers providing legal aid under different delivery models (e.g., Staff Office, community legal clinics, and block contracts).
  2. Improve the system's ability to detect quality problems. This is likely to entail the collection of better information from service providers on billing procedures and case outcomes; expanding client surveys; expanding the consumer complaints system; and conducting random file audits to ensure compliance with prescribed standards. Substantially superior information technologies to those at present possessed by the Plan will be required if the system is to acquire effective monitoring capacities, including monitoring abnormalities in billing practices or case outcomes that may be symptomatic of supplier-induced demand or quality deficiencies.
  3. Improve responses to known or suspected problems. This is likely to entail the provision of peer review and mentoring; requiring education and remedial education and training in response to identified practice deficiencies; and more effective sanctions for persistent deficiencies, such as removal from legal aid panels and a more coordinated complaints and disciplinary process between the Plan and the Law Society.
  4. Experiment with the development of detailed performance standards. The legal aid system should begin to develop detailed performance standards for particular areas of law and particular classes of proceedings, in consultation with members of the bar, the judiciary, specialized legal associations, and client organizations. These standards should apply to Staff Offices, community legal clinics, lawyers and law firms providing legal services under block contracts, and judicare lawyers.

The challenges entailed in devising and implementing an effective quality-control regime comprising the elements sketched above are not small. However, to date, the legal aid system and the Law Society have barely begun to embark upon this task. Failure to do so in the future is likely to leave unaddressed the substantial agency cost problems (i.e., supplier-induced demand and quality decline) described previously in this chapter. For disadvantaged clients who must have recourse to the legal aid system, legal aid is likely to remain something of a forensic lottery, rather than an assurance of equality before the law, and for taxpayers who fund the system, serious doubts will continue to be harboured as to whether they are getting value for their tax dollars.


This chapter has sought to make the case for a much more eclectic and diverse set of delivery models for the provision of legal aid in Ontario than has hitherto been the case. In particular, we have sought to make the case for a custom-designed delivery system that is tailored to the particularities of given classes of legal services, given geographic locations and the particular circumstances of individual clients. The overriding objective motivating this approach is the provision of maximum coverage for legal aid needs from a fixed budget-put bluntly, getting maximum mileage from a fixed amount of resources. This will require that the new legal aid system be entrepreneurial, innovative, open-minded, adaptive to changing needs and circumstances, and daring-that is to say, willing to take calculated risks by trying the untried and, at the same time, after giving them a fair trial, willing to cut its losses by abandoning unsuccessful initiatives and pursuing others. Operating under an open-ended budget, the legal aid system, until very recently, had few incentives to assign a high institutional premium to these characteristics. Now, in an era of fixed budgets, the legal aid system has no choice but to reconceive itself along these lines if it is to honour its commitment to the people of Ontario-both many of the province's most disadvantaged residents for whom the availability or lack of availability of legal aid is often of the most serious moment in their lives, and the taxpayers of the province who are asked to fund the system-to ensure that the ideal of equality before the law is realized as fully as our resources permit.

In choosing the elements of a substantially enriched and more diverse delivery system, the legal aid system should be guided by several basic principles. First, the legal aid system should seek to narrow the gap between full representation and no representation (an "on/off switch" approach to legal aid) by instituting a greater variety of legal services in order to assist a broader spectrum of potential clients. Full services for a small subset of needy clients, and no services at all for many others, is simply not an acceptable outcome. The range of services that might be made available, depending on the seriousness and complexity of the legal matters at issue, might range across a wide spectrum, from public legal education to preparation of self-help kits, duty counsel, Staff Offices, community legal clinics, judicare, block contracting and so on. That is to say, for financially disadvantaged clients with legal needs, it should rarely be the case that doing nothing should be the chosen option-principally by posing as the alternative the issuance of a judicare certificate, at the other end of the spectrum, which in many cases is an unsustainable option, given cost implications and budgetary constraints.

Second, the choice of delivery models must be highly sensitive and adaptive to context-the legal context in which services are required, the geographic context where they must be provided, and the special-needs context of particular client groups who require these services. As we have argued earlier in this chapter, for large-volume, relatively routine matters, duty counsel might be the appropriate response, either as part of a Staff Office or, where the size of the case-load does not warrant a Staff Office, on a stand-alone basis. While this may deprive clients of choice of counsel in this range of matters, recent budgetary cutbacks in many cases mean no counsel at all. Geographic context is also important. Obviously, in order to realize economies of scale and specialization, Staff Offices require substantial case-loads, which means that they are likely to be justifiable only in substantial urban centres. Nevertheless, in smaller communities, the legal aid system might still find it appropriate to hire duty counsel either on staff or on retainer to provide basic services in routine matters.

Client context also matters. Particular groups of clients often have distinctive legal, linguistic, or cultural needs that require tailored responses. The specialized community legal clinics with respect to children and youth, the elderly, Aboriginal people, environmental concerns, the African-Canadian community, and so on-already reflect a partial response to such needs. However, additional responses are likely to be appropriate in various contexts. For example, Jonathan Rudin, in a research paper prepared for this Review makes a persuasive case for establishing a network of Aboriginal legal service centres in major urban centres throughout Ontario where there is a significant Aboriginal population to service the legal needs of Aboriginal people living in these centres or nearby reserves. Rudin contemplates that, in many cases, these Aboriginal legal service centres would be integrated with local Aboriginal Friendship Centres and would provide culturally accommodating one-stop service for Aboriginal people in their catchment areas, including criminal and family law services. While community legal clinics have generally been opposed to assuming major responsibilities for the delivery of criminal law and family law services out of a well-founded concern that the volume of needs in these areas is likely to overwhelm their principal "poverty law" mandate, Rudin's proposals demonstrate that once again there are few absolutes or system-wide generalizations possible in choice of delivery models. Similarly, in the remote communities of Northern Ontario, a powerful case can be made for extending the number and scope of specialized full-service Aboriginal legal service centres with "fly-in" responsibilities for servicing remote communities in their catchment areas. Similarly again, the special legal needs of mentally handicapped clients-in the criminal justice system, guardianship proceedings, and committal and hospital review proceedings-almost certainly call for tailored responses. In short, context is everything in the choice of delivery models-legal context, geographic context, and client context. System-wide generalizations about the advantages and disadvantages of different delivery models cannot remotely capture these context-specific particularities.

Third, while different delivery models will often be complementary to one another, in some contexts we believe that there are substantial advantages to creating competition between different delivery systems. This is particularly true in the case of Staff Offices and the judicare system, at least in larger urban centres. We take this position for several reasons: (a) The existing empirical evidence on the relative performance characteristics of these different models in terms of cost, quality, accessibility, and social impact is so indeterminate that further experimentation is required to test the relative performance characteristics of these models against one another. This necessarily requires that the models compete across roughly the same range of legal matters. (b) We believe that intermodel competition will impose a desirable form of discipline on each of the competing models. Individual service providers within each system will be sensitive to the fact that the client choices will, in part, reflect perceptions of relative quality of service, convenience, expertise, promptness, courtesy, and so on available from providers in the other system and that their own professional prospects will eventually be affected by these client choices. At a systemic level, if the legal aid system seeks to impose excessive case-loads on Staff Offices, those offices can simply refer clients out to judicare lawyers. If judicare lawyers demand excessive tariffs for their work, these demands can be resisted, and clients redirected to Staff Offices. Threats of strikes by personnel in either system in an attempt to extract additional returns from the legal aid system will be disciplined by the existence of an alternative delivery system. (c) Intermodel competition preserves, and indeed enhances, client choice of counsel in those ranges of matters in which the models directly compete. (d) For reasons noted earlier, to attract high-calibre senior and junior legal staff to Staff Offices, their mandate cannot be confined only to large-volume, routine matters.

Fourth, in choosing delivery models in particular contexts, the legal aid system should be sensitive not only to relative cost considerations, but also to relative quality considerations. The ability to insist on and make operational an effective quality-control regime in a particular context should significantly influence the choice of delivery model in that context if agency costs in the form of supplier-induced demand and quality decline are to be contained to reasonable limits.

These general conclusions on choice of delivery models frame and shape the more particular recommendations we develop in succeeding chapters that focus on the provision of legal aid in specific contexts. As stated at the outset of this chapter, no single issue in the development of legal aid systems in Ontario and elsewhere has generated as much debate as the choice of delivery model. We believe that much of this debate has been unproductive, ideological, or self-serving, and has avoided the hard work of choosing and tailoring particular delivery responses to particular classes of legal needs. Because debates over delivery models have only recently begun to focus on this more productive issue, little hard empirical or comparative evidence exists as to which delivery model is optimal for which context. Of necessity, this means that the new legal aid system will need to adopt an innovative and experimental mind-set, involving significant elements of trial and error, as it redesigns the system for delivering legal aid across Ontario.

However, the alternative-preserving the status quo-is unsustainable, given the impact of recent budgetary cutbacks and the prospect of fixed budgetary resources for the future. The present lack of coverage of tens of thousands of disadvantaged Ontarians with pressing legal needs is an indictment of our collective commitment to equality before the law for all Ontario residents-a defining tenet of all representative liberal democracies-and provides us with no choice but to rise to the challenge of devising new and innovative ways of doing more with less.


  1. The Plan should seek to narrow the gap between full representation and no representation through provision of a much greater variety of legal services in order to assist a broader range of potential clients by invoking a wide spectrum of delivery mechanisms, including, for example, public legal education, duty counsel, supervised paralegals, Staff Offices, community legal clinics, judicare, and block contracting.
  2. The choice of delivery models must be highly sensitive and adapted to context-the legal context in which services are required, the geographic context where they must be provided, and the context of particular client groups with special needs who require these services.
  3. In choosing delivery models in particular contexts, a premium should be attached to early intervention to promote issue identification, settlement, mediation, diversion, and referral to other community agencies, rather than withholding legal assistance until disputes have become more intractable and costly to resolve, which may be counter-productive and impose greater demands in the long term on both the legal aid system and the underlying justice system.
  4. While different delivery models will often be complementary to one another, there are also substantial advantages, in various contexts, of creating competing delivery models, for example, judicare and Staff Offices in criminal law, family law, and refugee/immigration law in larger urban centres.
  5. In choosing delivery models in particular contexts, the Plan should be sensitive not only to relative cost considerations, but also to relative quality considerations, in particular the ability to prescribe and implement an effective quality-control regime in order to control problems of supplier-induced demand and quality variability or deterioration.