This paper focuses on delivery models, in particular the cost and quality associated with different delivery models. The type of delivery model has been shown to have a major effect on the cost of delivering legal aid. The issue of relative cost of different delivery models has been an important debate in Canada for nearly twenty years.

The table below shows the basic staff lawyer and private bar delivery models as they currently exist in legal aid plans in Canada. This reflects the traditional mixed model concept of staff and judicare delivery options. Later on in this paper, the concept of a complex mixed model will be introduced to describe more recent developments in legal aid delivery models. As Table II shows, the legal aid plans fall into three main categories with regard to delivery systems. Some legal aid plans, notably Ontario, New Brunswick, British Columbia and Alberta have predominantly private bar or judicare delivery systems. A judicare system is one in which lawyers in private practice are issued certificates to provide legal aid to clients. Some jurisdictions; Saskatchewan, Prince Edward Island, Nova Scotia, and Newfoundland have mainly staff lawyer delivery systems. A staff lawyer system is one in which the service is provided by salaried lawyers employed directly by the legal aid plan. Some jurisdictions utilize both staff lawyers and private bar lawyers. Quebec, Manitoba, Yukon and the North West Territories have mixed delivery systems.

The delivery model pattern becomes somewhat more complex when type of service is considered. Among the four legal aid plans that are usually categorized as mainly judicare delivery systems, both Ontario and British Columbia mainly use staff lawyers to deliver civil legal aid other than in family law. This is usually termed poverty law service. In Quebec, this is known as income security law. This area of legal service includes legal assistance in matters related to social welfare, employment insurance, housing, and debt. The other judicare plans, New Brunswick and Alberta, do not have significant poverty law service. All of the judicare plans use private bar lawyers exclusively, or nearly so, to provide legal assistance in young offender, criminal, immigration, and family matters.

Poverty law tends to be delivered mainly by staff lawyers in all legal aid plans. In Ontario poverty law service is provided entirely by staff lawyers working in seventy community legal clinics throughout the province. The community clinic system is funded and administered separately from the certificate service, but still within the Ontario Legal Aid Plan. The Legal Services Society of British Columbia is a complex structure consisting of Branch Offices managed directly by the Society, and of Community Legal Offices and Native Community Legal Offices which operate under contracts with the central Legal Services Society. In British Columbia, poverty law service is provided almost entirely by staff lawyers in CLO’s and NCLO’s. In Newfoundland, poverty law service is provided entirely by staff lawyers. In the mixed delivery plans, Manitoba and Quebec, poverty law service is provided mainly by staff lawyers.

In the other service delivery areas, young offender, adult criminal, family and immigration there is consistency of delivery models within plans. Legal aid plans that are mainly judicare or mainly staff lawyer systems tend to be consistently so across these four types of service. Quebec is more balanced between private bar and staff lawyer delivery, and at the same time more varied than other mixed model plans. Staff lawyers are used mainly for family legal aid. Private bar lawyers are used mainly for immigration and adult criminal legal aid. Young offender legal aid is equally divided between the two.

Canada has a fairly extensive body of empirical literature about the cost and quality of staff lawyer and judicare delivery. The empirical literature is instructive with regard to the relative cost and quality of these two delivery models. This literature will be summarized in more detail in the next section. The body of research represents a debate about the merits of these two delivery approaches that has persisted in Canada for almost twenty years. The history of the delivery models debate is, in certain respects, as instructive as the results of the empirical studies.

The delivery models debate began with what might be considered the classic study comparing staff lawyer and judicare delivery, the Burnaby study. As described below in detail, the Burnaby Study was the first in a series of Canadian studies to show that staff lawyers could deliver legal aid at less cost and with a similar quality of service compared with judicare in criminal legal aid. A few years later, the evaluation of the Nova Scotia legal aid plan was published. The Nova Scotia study reported lower costs for staff lawyers compared with judicare lawyers. However, the comparisons were not controlled for degree of case complexity. These studies were followed by a series of evaluations of legal aid plans, the British Columbia evaluation in 1984, the Manitoba evaluation in 1987, and the Saskatchewan evaluation in 1988. The general findings of these studies were summarized in Patterns in Legal Aid (2nd edition) as follows:

• staff lawyers spend less time per case than private lawyers
• staff lawyers tend to plead clients guilty more often than do private lawyers
• similar proportions of staff and private bar clients are convicted
• staff lawyer clients draw fewer jail terms than private lawyer clients.

The Patterns II report comments that the traditional perspective of the private bar that staff lawyer delivery in inferior ‘is simply wrong’ Further, the report concludes that Canadian evaluations have generally found that cases referred to the private bar cost more that staff lawyer cases, even when the there is no difference between the complexity and gravity of cases handled by private lawyers and staff lawyers. These differences cannot be explained in terms of differential case outcomes. Staff lawyer clients are convicted no more often than are the clients of private bar lawyers, and staff lawyer clients tend to receive fewer jail sentences.’

In 1987 the Canadian Bar Association produced a report on legal aid delivery models. The CBA report cautioned against drawing inter-jurisdiction conclusion because of the absence of comparable data from the various legal aid plans. However, the Canadian Bar

Association report conceded that ‘in the criminal field where some hard data on quality exist, albeit crude, it appears that the staff model is capable of delivering the same outcomes for lower costs than the judicare model, or slightly better outcomes for the same cost.’

The delivery models debate has, nonetheless, continued despite the mounting empirical evidence and the CBA report. On the surface the continuing debate about delivery models would seem somewhat paradoxical. Although very recently some developments in the development of delivery models in Canada have begun to show a shift away from the traditional debate (these will be discussed in the section below on complex mixed models), the delivery models debate has continued; an ideological debate played out on empirical grounds. A few of the recent episodes illustrate well the nature of the debate.

In 1991, the Law Society of Upper Canada commissioned a critique of the Legal Aid Manitoba evaluation. The Manitoba evaluation had produced the most direct evidence of the greater cost effectiveness of staff lawyer delivery to that point. The Pristupa report was an unrelenting critique of the ‘unvalid, unreliable, irrelevant, and fatally flawed’ methods and findings of the Manitoba study. According to the report, the ‘fundamental errors in questionnaire design as well as in the reporting of results and statistical inference render the report inappropriate, and inconsequential for any discussions about the costs and benefits of a Public Defender System.....the Manitoba Report has marginal or no application for Ontario.’ Three years after the critique of the Manitoba Evaluation, the same author, under contract to the Law Society of Upper Canada, prepared a critique of both Patterns in Legal Aid and Predicting Legal Aid Costs. Both of these Department of Justice reports had supported the relative cost-effectiveness of staff lawyer delivery compared with judicare delivery. The second Pristupa report claimed that ‘the inferences and conclusions drawn from the data in Patterns are inappropriate and statistically invalid, that ‘no change in (Ontario) government policy should be considered that is based on the research methodology in Patterns,’ and that ‘no changes in (Ontario) government policy should be considered that is based on the research methodologies used in Predicting.’ The Department of Justice commissioned two reports to answer the critique leveled by the Pristupa Report on Patterns and Predicting, despite the fact that Department recognized that the issues were not open to resolution on empirical grounds.

In 1993 a committee of the New Brunswick Department of Justice recommended a staff lawyer delivery model for that province. This followed several years of relatively high costs per case reported by Legal Aid New Brunswick. Early in 1994 the Law Society of New Brunswick responded to the proposal for staff lawyer delivery with its own proposal to ‘allow the participating lawyers the opportunity to demonstrate to the public and the funders of the program that they, as a group, are prepared to make the sacrifices necessary to preserve the positive aspects of the Judicare system.’ In April 1994 the New Brunswick legal aid plan implemented a global budgeting system in order to contain costs under the judicare delivery system. Among other administrative measures, global budgeting involved a 40 per cent hold back of payments on certificates to the private bar in order to assure that expenditures would remain within a fixed budget. This measure has preserved the judicare system in New Brunswick within a ‘hard-capped’ budget.

In 1992 an influential report on legal aid in British Columbia was published. A main focus of the Agg Report was what the report considered underfunding of legal aid services. This set in motion a series of discussions between the Legal Services Society and the government not only about funding, but about achieving efficiencies in a number of ways. One important aspect of this discussion was the implementation of a mixed model for service delivery. With the exception of poverty law services, legal aid in British Columbia was delivered almost entirely by private bar lawyers at that time. In 1994, a Reform Package was introduced which included a number of major elements; a tariff reduction, a client contribution program, and a proposal to move to a 50 per cent staff lawyer system. The private bar reacted strongly to the mixed model proposal. In July 1994 the Association of Legal Aid Lawyers announced a withdrawal of services as a protest against the mixed model proposal. In September of 1994 a tri-party Accord was reached between the Law Society of British Columbia, the government, and the Legal Services Society to limit the number of staff lawyers to a total of 90 (amounting to about 15 - 20 per cent of total service delivery, and to require an evaluation of the mixed model within six months. A staffing freeze, which has been in effect since mid-1995, has effectively limited the staff lawyer component at its present level.

Another recent attempt to experiment with service delivery innovation in British Columbia was the proposal to deliver criminal legal aid services under a contracting model. The original proposal developed in 1996 was to contract out approximately 4000 criminal legal aid cases in blocks of 50. This would have amounted to about one quarter of all criminal legal aid cases in the province. The private bar reacted with caution, but not with opposition, in the beginning. Despite the size of the proposed contracting project, it was referred to as a pilot project. In preparation for the main pilot project, a limited pre-pilot contracting project involving six contracts for blocks of 50 cases was undertaken in Victoria and Vancouver in 1997. During the pre-pilot phase of the project, private bar opposition grew to the point where the project was discontinued. An evaluation of the pre-pilot phase of the project indicated that a 19 per cent cost saving was achieved, a successful contract administration relationship was maintained, and there were not attracting experienced counsel who provided quality service.

These events in the history of the delivery models debate in Canada demonstrate the continuing opposition of the private bar to staff lawyer delivery. This shows that a delivery model is not only a technical-administrative mode of service delivery. There is a politic to delivery models reflecting a set of vested interests, typically very strong vested interests on the part of influential actors in the system. Importantly, what has continually fallen victim to the politics of delivery models is the ability of legal aid plans to experiment with innovative service delivery. Thus the ability of legal aid plans to develop more cost effective delivery approaches, or possibly to develop delivery approaches which address aspects of clients’ problems other than those which ‘would normally be provided by a lawyer’ are compromised by the politics of the delivery models debate.

This observation will come as no surprise to those in other western countries. Forty years ago in Britain, the private bar described the proposed use of staff lawyers to deal with uncontested divorces as ‘a fungoid growth on the profession’. In the United States the controversy over delivery models has been fraught with controversy, with the combatants accusing each other of bad faith. An Australian observer has commented that the delivery models debate in that country has been inconclusive not only because the lack of empirical evidence but also because of the partisan debate. Finally, the Canadian Bar Association has acknowledged that the delivery models debate is one in which the ideology and the rhetoric have outweighed the results of empirical research. ‘[T]here are few topics which appear to arouse such strong and varying opinions as the choice of delivery model. Ideology and personal experience come together on this topic, allowing most lawyers to hold and advocate positions with great conviction.’

The political, cultural, and other relevant circumstances in countries which are beginning to develop legal aid plans are certainly different from those in Canada or other countries with developed legal aid systems. It would not be expected that the same factors would shape debates about delivery models as has been the case in Canada. However, the mixed model debate in Canada does provide a useful cautionary tale. The choice of a delivery model may bring with it difficult political dynamics and impediments to future development. To the extent that these can be foreseen and avoided, improvements in cost-effectiveness and service delivery will be easier to achieve.

This section first presents research results relating to the relative cost of staff lawyer and private bar delivery. ‘Raw data’ from legal aid plan data bases are not presented here, although these data are used in the Department of Justice Patterns in Legal Aid and in the Canadian Bar Association Legal Aid Delivery Models report. Data from management information systems do not have controls for complexity of cases. Thus intra-plan comparisons between staff lawyers and private lawyers may be flawed because of different referral patterns to private and staff lawyers, in which staff lawyers may tend to get the less complex cases. Inter-plan comparisons may be flawed because of different definitions of a case, and different coverage provisions may result in comparisons of non-equivalent cases.

The first carefully designed study presenting evidence on the relative cost-effectiveness was the Burnaby Public Defender study published in 1981. That study revealed that the average cost of staff lawyer cases was $235. compared with an average cost of $225 for Burnaby private bar lawyers and an average cost of $264. for Vancouver private bar lawyers. (Burnaby is a suburb of Vancouver.) A major concern with these findings revolved around productivity. About twenty per cent of the staff lawyers’ time was spent on duty counsel work. Without the duty counsel work, staff lawyers could have increased their case loads by about fourteen percent. If there had been an increase of only four cases per month, the average cost per case for staff lawyers would have dropped to $192.

The evaluation of the Manitoba legal aid plan produced what is perhaps the best data on comparative staff and judicare costs. The overall average cost per case for staff lawyers reported in this study was $197., compared with $307. for private bar lawyers. Costs were further analyzed in terms of average cost by quartile thresholds of the case load as one way to control for possible difference in case complexity. Staff lawyers completed the first 25 per cent of their case load for an average cost of $48. or less, compared with $201 for the private bar. Staff lawyers completed 50 percent of their cases for an average cost of $100. or less, compared with $263. for private bar lawyers. The point at which 75 per cent of all cases were taken into account produced average costs per case of $241 for staff lawyers and $310 for private bar lawyers.

The Manitoba evaluation report also presented average costs per case for specific types of offenses, comparing staff and private bar lawyers. The following table shows some of these comparisons.

As well as differences in cost per case, the Manitoba study identified differences in hours spent per case by staff and private lawyers. Staff lawyers spent an average of 3.9 hours per case compared with 8.2 hours per case by private lawyers for assault cases. For weapons offenses staff lawyers used 5.2 hours, compared with 9.7 hours used by private lawyers. For break and enter cases, staff lawyers spent an average of 3.2 hours compared with an average of 6.9 hours spent by private lawyers. Thefts over $200 required an average of 3.4 hours to complete the case compared with 7.5 hours for private lawyers.

An evaluation of the Saskatchewan legal aid plan was conducted at about the same time as the Manitoba evaluation. In connection with the Saskatchewan evaluation, a costing substudy was conducted to examine any possible advantage of moving to a mix including more private bar delivery. At that time, about 98 per cent of all legal aid was delivered by staff lawyers. That study estimated that the total cost of legal aid would increase by thirteen per cent if one third of all legal aid were delivered by the private bar, and by sixty-four per cent if all legal aid cases were referred to the private bar.

In 1993 the Alberta Legal Aid Society implemented a three year pilot project to deliver legal aid to young offenders by means of a staff lawyer clinic approach. Two clinics were established, one in Calgary and one in Edmonton, the two largest cities in the province. The evaluation of the staff lawyer clinics demonstrated that staff lawyer delivery was more cost effective than private bar delivery. In 1996, the average cost per case for staff lawyers was $353 compared with $500 for private bar lawyers. The private bar costs were about 30 per cent higher than for staff lawyers.

The staff lawyers working in the clinics performed duty counsel work as well as full service representation. The evaluation concluded that staff lawyers were more effective at resolving matters at the early stages of the justice process than were private lawyers. The cost savings that resulted from duty counsel resolving matters early in the process were estimated at $2.4 million over the entire period of the evaluation. This was because of about 4800 certificates that were not created because matters were resolved by duty counsel.

The empirical research that has been conducted to date in Canada points consistently to the conclusion that staff lawyer delivery is less expensive than judicare delivery. This raises the quality issue. Staff lawyers may be less expensive, but is the quality of service comparable to that provided by private bar lawyers?

The conclusion that there is a cost advantage in staff lawyer delivery is frequently countered by two main arguments; that the types of cases dealt with by staff lawyers tend to be less complex, thus explaining the lower cost; and that the quality of service provided by staff lawyers is inferior. The available evidence does not support either of these contentions.

One of the perennial arguments presented by critics of staff lawyer delivery has been that staff lawyers provide an inferior service. One way to assess quality of service is in terms of outcomes achieved for their clients by private bar and staff lawyers. Canadian research has shown that clients of staff lawyers and private bar lawyers are convicted at about the same rate. However, the clients of staff lawyers tend to receive fewer jail terms than the clients of private lawyers. In the Burnaby project, clients were convicted in about sixty per cent of cases. This was about the same rate as for legal aid clients of private bar lawyers. However, about 40 per cent of the clients of private bar lawyers were sentenced to jail, compared with thirty per cent of the clients of staff lawyers. Staff lawyers tended to plead clients guilty slightly more often than private bar lawyers, but with no apparent effect on disposition or sentence.

The evaluation of the British Columbia legal aid system in 1984 produced the same patterns. Clients of staff lawyers and private bar lawyers were convicted at similar rates. However, staff lawyer clients were sent to jail thirty per cent of the time, while clients of private bar lawyers were imprisoned forty two per cent of the time.

The evaluation of Legal Aid Manitoba again produced a similar pattern. Seventy-two per cent of both staff lawyer and private bar lawyer clients were convicted. However, only twelve per cent of the clients of staff lawyers were imprisoned, compared with twenty-three per cent of the clients of private practitioners.

The Saskatchewan evaluation produced similar results. Only fourteen per cent of the clients of staff lawyers were convicted, compared with 32 per cent of private bar clients. In this case, however, it must be kept in mind that Saskatchewan is a 98 per cent staff lawyer system. The private bar provides service in only a small number of cases where conflicts of interest exist, or where staff lawyers are not available to handle the case. There may, therefore, be differences in referral patterns that could account for these differences.

Finally, the evaluation of the Alberta young offender staff lawyer project showed findings which are generally consistent with the previous research. The Calgary office did not show statistically significant differences in disposition or sentencing compared with the private bar. The staff lawyers in the Edmonton office produced better sentences for their clients compared with the private bar.

These results from studies conducted in different places and at different times all point to the same conclusion. Staff lawyers achieve similar conviction rates compared with private bar lawyers. Staff lawyers usually get fewer custodial sentences for their clients. This suggests that the quality of service provided by staff lawyers is at least equal to that provided by the private bar.

Opponents of staff lawyer delivery frequently argue that the cost advantage of staff lawyers arises because of different referral patterns. Dramatic differences in costs per case between staff lawyers and private bar lawyers, such as the difference of $122. for staff lawyers versus $729 for private bar lawyers reported in the Nova Scotia legal aid evaluation which are not controlled for case complexity may well reflect differences in referral patterns and should be interpreted with great caution. However, in the Burnaby Study all cases were assigned to the staff office until over a period of weeks the work load optimum was reached. At that point the office would accept no more cases, and all cases were assigned to the private bar. This process achieved a type of random assignment of cases. Thus the cost difference could not be attributed to referral patterns.

The Manitoba evaluation attempted to partially control for case characteristics by comparing staff lawyer and private bar costs and outcomes for specific types of offenses. As summarized above, staff lawyer costs per case were consistently lower for all categories of offenses.

Both of these studies suggest that the cost advantage of staff lawyers can not be dismissed because of differences in referral patterns. The dramatic differences that appear based on ‘raw data’ that do not control for case type and complexity may to some degree reflect differences in referral patterns. The research that does exist on the subject suggests that after controlling for type and complexity of case, the cost advantage acknowledged by the Canadian Bar Association in the Legal Aid Delivery Models report remains.

It is frequently remarked that staff lawyers are not independent of the justice system as are private bar lawyers. The Canadian research shows that staff lawyers tend to plead clients guilty more often than private bar lawyers. The evaluation of the Alberta youth staff lawyer project reported that staff lawyers tend to resolve matters earlier in the justice process than private lawyers. Yet in terms of outcomes, staff lawyer clients get similar levels of conviction and fewer custodial sentences compared with the clients of private bar lawyers. At a minimum, it does not appear that the greater tendency for pleas of guilty has any practical consequence for the quality of service, as measured in terms of outcomes. Henry and Fleming remark that ‘[t]he independence question is based on ideological concerns, predominantly of the legal profession, and is not, at present, backed up by any scientific empirical data.’

As pointed out above, private bar lawyers do tend to spend more time on cases than staff lawyers. Data from the Manitoba evaluation was presented above to show that private bar lawyers spent more time per case. Even more detailed data from the Manitoba study show that staff lawyers spent less time per case regardless of the plea, the sentence, or the disposition. The table below is reproduced from the Manitoba report. The data shown in Table IV relate to four offense categories reported for combined time differences generally; assault (indictable), break and enter, theft over $200, and weapons offenses. These data demonstrate conclusively that private bar lawyers spend more time per case than staff lawyers. Further, it seems clear that the additional time spent by private lawyers achieves no advantage for clients in terms of the outcome of the case.

Some very similar results emerged from the Alberta young offenders staff lawyer evaluation. The table below compares the amount of time spent by staff and private bar on various aspects of cases.

The Canadian data show that private lawyers tend to spend more time on cases than staff lawyers. The data from Manitoba suggests that the additional time does not result in better outcomes.

The explanation for the additional time spent by private lawyers may relate in part to the nature of payments to the private bar on a tariff system. A consistent complaint by the private bar is that legal aid tariffs are too low compared with the ‘market value’ for legal services, and that they are thus inadequately compensated for their services. This may lead to an incentive on the part of lawyers to bill to the maximum allowable limit. According to Patterns, ‘Alberta provincial evaluators have referred to this practice as strategic billing; Quebec provincial evaluators have called it bill padding.’

It has already been observed that one of the criticisms of staff lawyer delivery is that staff lawyers plead clients guilty more than do private bar lawyers. The implication is that the quality of the defense provided by staff lawyers is therefore less. This might also partly explain differences in time spent per case.

Data from the British Columbia legal aid evaluation did show a difference in the percentage of guilty pleas between staff and private lawyers.

However, the British Columbia evaluation also found that among those who did go to trial, equal proportions of staff lawyer and private bar clients received acquittals, stays or withdrawals of charges.

The Alberta youth staff lawyer evaluation compared the timing of guilty pleas entered by private lawyers and staff lawyers. The data from that research show that staff lawyers tend to plead their clients guilty earlier in the process than private bar lawyers.

This table illustrates what may be a difference in approach or strategy between staff and private lawyers in the Alberta project. Staff lawyers show a much greater tendency than private lawyers to plead guilty before the trial date. Private bar lawyers show about an equal tendency for guilty pleas before the trial date and on the date of the trial. Pleading guilty on the date of the trial gives the appearance of a strategy to wait until the last possible moment in the hope that some unforeseen or hoped for event will swing the case in favour of a client who is otherwise without a good defense. There may, of course, be other explanations. This finding suggests that the practices of staff lawyers and private bar lawyers doing legal aid cases may be different. The data on guilty pleas and outcomes, and time per case and time per case and outcomes, suggest that the differences in approach do not affect outcomes.

However, the data presented above in Table VIII call to mind one of the main criticisms of staff lawyer schemes; that they are not independent. The Canadian Bar Association report on Legal Aid Delivery Models raises the possibility of potential conflict between the best interests of the client and the interests of the system ’when the paymaster is the public purse’. Within an adversarial legal system, the principal goal of the defense bar is to provide full and fair defense of the client’s interests. Defense schemes should be judged on how well they do this. These data suggest that more empirical evidence should be gathered on how staff lawyers and private bar lawyers defend clients, and how any differences might impact on the best interests of the clients.

Based on the Canadian research, it can be concluded that staff lawyer delivery can be less expensive than private bar delivery, with no compromise with respect to quality of service. These conclusions are based on several studies carried out in different places and at different times, in legal aid plans which are different in many respects. Each study may have some methodological shortcomings, as one expects in all research into complex issues. However, the strength of this body of research is in the consistency of the findings regardless of differences in research approaches and settings.

Legal aid delivery is not a simple, one-dimensional issue. A delivery model must provide the best service possible, in the most cost-effective manner, in ways that address a number of major aspects of service delivery. Legal aid service is provided in different areas of law, to diverse client groups, in different geographical areas, and involving cases that vary from simple to very complex. These and other factors make legal aid delivery a complex and multidimensional problem, not a simple and unidimensional one. It stands to reason, then, that neither private bar lawyers providing service on a individual fee-for-service basis nor staff lawyers providing a similar service as salaried employees will necessarily be the best solution to all delivery problems.

Legal aid delivery in Canada is moving beyond the simple mixed model of staff lawyer and judicare delivery that has framed the debate about cost and quality in Canada for the past twenty years. What may be termed complex mixed models are emerging in Canadian legal aid, in which a variety of delivery modes are being developed to target specific service delivery needs. The following section describes some recent innovations in legal aid delivery which point toward the development of complex mixed models.

This principle has been recognized with regard to legal aid delivery for different types of service for some time. For, example, the Ontario Legal Aid Plan has traditionally delivered poverty law service through a system of clinics employing staff lawyers, while delivering criminal, family, and immigration legal service using private bar lawyers through the certificate system. Similarly, The Legal Services Society of British Columbia has delivered poverty law service through quasi-independent Community Legal Offices and Native Community Legal Offices which operate under contract with the central Legal Services Society. The community office concept is intended to maximize community input into identification of needs and the delivery of service. In order to accommodate a sparse population in a vast geographical areas, the North West Territories legal aid plan has delivered all types of service through eight legal aid offices and five clinics located in remote communities throughout the territory, in addition to the head office in Yellowknife.

Contracting out legal aid cases to law firms is an alternative to either staff lawyer or judicare delivery. Contracting out has so far been used for very specific purposes in Canadian legal aid, rather than for mass casework. Contracting out was first introduced into the Legal Aid Manitoba system in 1992 with the Portage Legal Services Initiative. The Portage experiment, named for the town of Portage LaPrairie in the area, was an attempt to provide legal aid services in a cost-effective way to the rural and sparsely populated Interlake Region of central Manitoba. Law firms in the area were invited to submit bids to provide duty counsel and full representation services for that area. Under a contractual agreement with a local law firm, Legal Aid Manitoba was able to provide legal aid services at a lower cost than would have been the case with either staff lawyers or private bar lawyers using individual legal aid certificates. This contracting arrangement remains a regular part of the Legal Aid Manitoba delivery system. It serves as an example of how contracting may be used as an alternative to the traditional options of staff lawyer or private bar delivery to provide cost-effective service in rural areas.

The following year, in 1993, Legal Aid Manitoba began contracting out blocks of 50 young offenders cases in the Winnipeg area. With the proclamation of the Young Offenders Act the legal aid plan experienced a large number of legal aid applications with mandatory eligibility under Section 11 of the Act. Most of these matters were relatively simple, and in most cases would not have received service under the normal financial eligibility requirements and coverage provisions that applied to adult criminal offenders. Since that time several law firms in Winnipeg have been accepting contracts for blocks of 50 young offender cases, and this has since become a regular part of the service delivery system. There has been no formal evaluation of this delivery mode. Legal Aid Manitoba management has concluded that considerable savings have been achieved by contracting out young offender cases, and that the quality of service is satisfactory. This project illustrates the successful use of contracting out as an alternative to staff lawyer or private bar delivery to provide service to a specific client group, in this case, young offenders.

The British Columbia Legal Services Society carried out a successful experiment contracting out both young offender and adult criminal legal aid in 1997. As was explained above in the section discussing the history of the mixed model debate, the original intention was to experiment with contracting on a mass casework scale. The large scale experiment was not carried out because of political reasons. However, the evaluation of the smaller pilot projects, two young offender pilots in Victoria and two adult criminal pilot projects in Vancouver, showed that the projects produced an estimated cost saving of nineteen percent over private bar delivery, and that no problems attracting experienced lawyers to deliver quality service were encountered.

One issue that arose in the planning of the British Columbia contracting project was to avoid where possible situations in which contracting would conflict with the work of the staff lawyers. This reflects a broader issue with regard to complex mixed models. The various delivery modes in a complex mixed model should be complementary, and overall build toward an integrated delivery system comprised of delivery modes that are targeted toward specific delivery problems, and which function in a complementary manner.

A novel concept in duty counsel service which reflects the complex mixed model approach is expanded duty counsel. This idea was piloted in criminal legal aid service in Manitoba from 1994 to 1996. Following the implementation of an automatic charging policy for domestic violence in Manitoba, and the establishment of a specialized domestic violence court in Winnipeg, Legal Aid Manitoba experienced a dramatic increase in legal aid applications in this area. The cost of these additional cases was substantial. In 1994 the legal aid plan implemented the expanded duty counsel project, mainly in response to that situation. Expanded duty counsel was implemented in the domestic violence court and in the over-night custody court. One year later, the expanded duty counsel program was implemented for the non-custody docket. All clients encountered at first appearance receive service, regardless of financial eligibility. The main feature of expanded duty counsel is continuity; continuity of lawyers assigned to the same court for an extensive period of time, and continuity of the relationship between the lawyer and the client which allows the lawyer to retain a client first encountered in first appearance court. Expanded duty counsel might well be called a disposition model of duty counsel. The basic objective is for the duty counsel lawyer to dispose of as many simple cases as workload permits, as early as possible in the criminal justice process. This approach to service delivery rests on the assumption that a large proportion of criminal cases are quite simple. There are no complex points of law. The facts of the case are straightforward. If the duty counsel lawyer determines that there is a defense and that the case should go to trial, and if the accused person is financially eligible, the case will be transferred to a regular staff lawyer or a certificate will be issued to a private bar lawyer. However, if the duty counsel lawyer feels that the charge will result in a guilty plea or a stay of proceedings, (s)he will retain the case and attempt to resolve it in the best interest of the client. Usually this involves negotiation with the Crown Attorney. Expanded duty counsel lawyers retain cases that they feel can be resolved without a trial. However, depending on their workload, and on the number of days per week that they had to be in first appearance court, the expanded duty counsel lawyers can handle a small number of brief day or half day trials.

The evaluation of the pilot project concluded that Legal Aid Manitoba achieved substantial cost savings by having the expanded duty counsel program dispose of cases, compared with the likely cost of private bar or regular staff lawyer service, with no compromise in quality of service. The expanded duty counsel cases were assessed for coverage and financial eligibility, and were divided into certificate and certificate non-equivalent cases. Three indicators of seriousness and complexity; prior offenses, other related charges, and breach of a judicial order were determined in order to assess the degree of case complexity and seriousness. These were compared with the costs of staff lawyer and private bar cases for the same offenses. Overall the data showed that expanded duty counsel cases cost considerably less than either staff lawyer or private bar cases involving the same offenses. Such comparisons were very crude, however, because comparability was difficult to achieve. The private bar and staff lawyer cases would have included some more complex cases that the expanded duty counsel lawyers would not have dealt with. As well, the lawyers in the expanded duty counsel project may have handled cases differently than roughly identical cases would have been handled by private bar lawyers. It was, nonetheless, concluded that the expanded duty counsel service was cost effective.

Expanded duty counsel became a major component of the criminal legal aid delivery system in Manitoba following the pilot project phase. In April 1997 expanded duty counsel was adopted province-wide. The impact of adopting expanded duty counsel province-wide has not been assessed. The management information system at Legal Aid Manitoba has not yet been restructured to capture a full range of data on expanded duty counsel. However, data for criminal and youth legal aid, for both staff lawyers and private bar lawyers, show that expanded duty counsel is removing many cases from the full service certificate stream, and increasing the average cost per case for full staff lawyer and private bar certificate service. The table below shows the changes in the volumes and average costs of cases since expanded duty counsel was extended province-wide.

The volume of cases has declined and the average cost has increased in each category following the implementation of the expanded duty counsel program province-wide. This has occurred as expanded duty counsel has disposed of the simpler cases at the early stages of the criminal justice process before transfer to the certificate stream or to a staff lawyer for full service is required. This clearly shows how expanded duty counsel has become a separate delivery mode, and a main component of the legal Aid Manitoba service delivery system along with the staff lawyer and private bar delivery modes.

In 1995 Legal Aid Manitoba opened an Urban Aboriginal Legal Aid Clinic in Winnipeg. The clinic had fewer clients than expected during the first year of operation. This was probably because the expanded duty counsel program was removing many cases from the system, and thus from the specialized clinic. This points out how elements of a complex mixed model can come into conflict, and points to the necessity for integration of the components of a complex mixed model. A staff lawyer clinic designed specifically to serve a special needs clientele is, in itself, an illustration of the complex mixed model approach.

Duty counsel services in some other jurisdictions appear to operate in a similar fashion to expanded duty counsel. In the North West Territories, staff lawyers servicing circuit courts in remote villages tend to blend duty counsel and trial work as circumstances require. Legal aid lawyers are reported to routinely resolve cases at the duty counsel stages of proceedings. A ‘presumed eligibility’ system operates which ignores financial eligibility requirements unless a matter is sufficiently complex to require a legal aid certificate.

Legal Aid New Brunswick reports that approximately thirty-five per cent of all duty counsel contacts result in completed or disposed matters. No formal study of duty counsel services has been carried out in that province. The nature of the matters resolved by duty counsel is not known. Duty counsel in New Brunswick is provided by private bar lawyers. It is not clear what similarities there might be between the New Brunswick and Manitoba duty counsel services.

Expanded duty counsel is a new approach to legal aid delivery that is designed to deal with that relatively large proportion of the overall case load that are of minimal complexity. These simple matters can be resolved competently by expanded duty counsel lawyers, retaining clients for a week or two until a second appearance, with a minimal amount of time, and at much lower cost than if certificates were issued to private lawyers for the same cases. Cost-effectiveness of the overall delivery system is achieved by utilizing this particular delivery mode targeted at the resolution of relatively simple legal aid cases at the early stages of the justice process. More complex cases are dealt with by private bar or staff lawyers.

The complex mixed model concept is an emerging approach in service delivery. Nowhere is this trend more apparent than with the certificate side (adult criminal, young offender, immigration, and family legal aid) of the Ontario Legal Aid Plan. In 1996 the Ontario Legal Aid Plan sought and received approval from the Law Society of Upper Canada, the governing body of the legal aid plan at that time, to launch an extensive series of pilot projects to test alternative service delivery methods. The pilot project initiative consists of 15 pilot projects in total, some which will be carried out in multiple sites. The pilot projects that will be put in place are as follows; in criminal legal aid: staff lawyer offices, expanded duty counsel, contracting; in immigration legal aid: an extension of the staff lawyer immigration pilot project to encompass detention hearings, special immigration panels, contracting; in family legal aid: staff lawyer offices, contracting, expanded duty counsel, unbundled family law services; in young offenders legal aid: staff lawyer offices, youth court counsel, contracting for court-appointed counsel cases.

This ambitious pilot project initiative was developed in anticipation of the release of the McCamus Report on legal aid in Ontario, which was a government sponsored review of the Ontario legal aid system. The McCamus Report recommended, among other things, experimentation with alternative delivery models.

As of late 1998, some of the pilot projects in immigration, youth, and family law are under way, or are about to be implemented. The Law Society decided to delay the implementation of the criminal legal aid pilots, since these would be more controversial among the politically powerful and conservative criminal bar. All of the pilot projects will be monitored and evaluated.

In their pilot project form, the projects are being designed and located so that they do not interact or compete with one another. However, it is clear that the projects that prove successful will have to be implemented in such a manner that they become coherent and interdependent elements of an overall delivery system.

The Ontario Legal Aid Plan has moved a considerable distance from being the champion of private bar delivery in the early 1990’s to having launched the most ambitious program of pilot projects on innovative service delivery in the history of legal aid in Canada. Clearly, the program of pilot projects is aiming beyond the simple mixed model of staff lawyer and judicare delivery toward a complex mixed model comprised of several complementary delivery modes.

The delivery of refugee and immigration legal aid in Ontario is of special interest with respect to complex mixed models. During the past three years, the Ontario Legal Aid Plan has experimented with a refugee staff lawyer clinic in Toronto called the Refugee Law Office. The evaluation of the staff lawyer office demonstrated that the productivity of the clinic was not sufficient to make the clinic cost-effective, although the quality of service was high. The reason for the low productivity was the choice, taken in order to respect right to choice of council in the strictest manner, not to assign applicants to the clinic. Applicants were informed about the clinic, but only as an option along with private bar lawyers. Many clients continued to gravitate toward private lawyers through informal networks of friends and family or other contacts. Thus too many clients continued to gravitate toward private council through other referral channels. In the second phase of this pilot project, which is part of the pilot project initiative described above, the staff lawyer office is to assume responsibility for detention hearings. This is a previously unmet need that will augment the work of the refugee law office. Other measures are also being taken to increase the case load of the staff lawyer office.

Another immigration legal aid pilot project which is getting under way is the Specialized Refugee Panels project. Immigration cases from two particularly high volume countries, Mexico and Nigeria, will be assigned to special panels of private bar lawyers who have extensive experience dealing with refugee matters from those countries. Thus what is emerging in refugee legal aid service, on an experimental basis, is a two-fold approach to service delivery in this area. The Specialized Panels are designed to deal with refugees from the high volume countries, with issues and problems unique to those countries of origin. The Refugee Staff Lawyer Office will continue to handle the more diverse case load from a larger number of countries, plus the detention hearings. This emerging arrangement demonstrates the essential feature of a complex mixed model; different delivery modes targeted at specific requirements within the overall service delivery system.

The Alberta Youth Staff Lawyer Office, which has been discussed extensively above, is an example of the development of a complex mixed model in that province. It reflects the complex mixed model idea because this is a staff lawyer office that is targeted toward servicing a specific clientele, young offenders, in cities where there is a high volume of those clients. The two staff lawyer youth offices in Calgary and Edmonton provide a variety of innovative services to meet the special needs of youth. These reach beyond the services normally provided by lawyers. These include: baby sitting provided by a downtown church so that female young offenders with children will have a place for them while appearing in court; a program, of providing public transit tickets to youth to overcome transportation problems that would result in failures to appear in court; recreational activities, employment counseling aimed at crime prevention; and mental health assessments and counseling to assist young offenders to take advantage of solutions to the difficulties that may be underlying their legal problems.

Assisted self-representation combines public legal information with summary advice and, possibly, some limited legal assistance. This type of limited service is designed to provide some assistance to applicants whose applications for legal aid are refused, because the legal matter does not fall within coverage provisions, or because the applicant fails, possibly by a very narrow margin, to meet the financial eligibility requirements. Assisted self-representation is a response to the increasing number of people who do not quality for legal aid, as legal aid plans continue to cope with limited budgets.

One assisted self-representation project was attempted on an experimental basis by the Legal Services Society of British Columbia. A number of booklets informing clients how to defend against a criminal charge were prepared by the Public Legal Education Department of the B.C. Legal Services Society. This material was provided to rejected applicants in five branch offices and community legal offices throughout the province. The results of the study showed that the legal information material did not prepare people to defend themselves effectively. However, the material was extremely useful in several other ways. The material was judged by clients to be very helpful in alerting them to the potential seriousness of their situation. A large percentage of the sample indicated that reading the pamphlets convinced them to borrow the funds to hire a lawyer. The legal information material provided useful information about sources of advice and assistance, other than legal aid, that might be available. The research indicated that an area not covered well by the original material, but a very useful type of information to be included in the future, was information about alternative measures and how to effectively present an alternative measures option to the Crown Prosecutor. The study concluded that the public legal information was useful in a number of ways, but not primarily to prepare people to effectively represent themselves in court.

The Ontario Legal Aid Plan Pilot Project initiative includes an ‘Unbundled’ Family Law Services Project. Unbundled services is an assisted self-representation project in family law. In this case, applicants will be issued a limited legal aid certificate for a few hours of service from a private bar lawyer to get advice or assistance in drafting documents, or to advise them about how to represent themselves. At the time of this writing, the Unbundled Family Law Services project is just getting underway. There are no results to report at this point in time.

These two projects represent another delivery mode than is emerging in Canadian legal aid. It is intended to deal with the lower priority legal matters for which legal aid plans have had increasing difficulty providing coverage in the past few years. It is a delivery mode that combines with the others to extend limited services to clients whom the legal aid plan could otherwise no longer afford to serve.

These examples drawn from recent developments in service delivery illustrate the emergence of an approach to delivery models that goes beyond the tradition bipolar concept of a mixed model based on the staff lawyer - private bar options. The traditional concept is a simple mixed model, as distinguished from what may be termed the complex mixed model. A complex mixed model is an integrated set of delivery modes (staff lawyers, private bar lawyers on fee-for-service certificates, expanded duty counsel, contracting) and structures (clinics) that are targeted at specific service delivery problems. The complex mixed model rests on the recognition that no one delivery mode is the best for all purposes, without qualification. Complex mixed models use a variety of delivery approaches in an integrated fashion to address some particular service delivery need. The components of a complex mixed model will vary from one jurisdiction to the next depending on the circumstances specific to that place. The specific components that are employed are essentially incidental. The essential element to the complex mixed model concept is the utilization of a range of delivery modes matched to specific delivery problems.

Interest in delivery models in Canada has focused on the relative cost effectiveness of staff lawyer and private bar delivery of legal aid. The results of Canadian research carried out in several provinces points to the conclusion that staff lawyer delivery has been less expensive than private bar delivery. For the most part, the evidence further suggests that the quality of the service provided by staff lawyers is equal to that provided by their private bar counterparts.

The Canadian research leaves one area of concern about staff lawyer delivery. There is evidence to suggest that staff lawyers plead clients guilty earlier in the criminal justice process than private bar lawyers. While this does not necessarily mean that the quality of the service provided by staff lawyers is compromised, it does suggest that staff lawyers and private bar lawyers may do their work differently. Further research should examine more closely the way in which staff lawyers and private bar lawyers do the work of representing clients.

The evidence that can be drawn from carefully designed research about staff lawyer and judicare costs, as opposed to uncontrolled comparisons based on raw data, is based on research in criminal legal aid. One evaluation of staff lawyer delivery of refugee legal aid did not find that staff lawyers were relatively cost effective compared with private bar lawyers. The evaluation concluded that this was due to low productivity of the staff lawyer office. The productivity problem was caused by the manner in which cases were assigned to the staff office. Therefore the results could not be generalized.

The Refugee Law Office study highlights the importance of productivity in any conclusion about the cost -effectiveness of staff and judicare delivery. The relative cost-effectiveness of the two basic delivery modes is a function of three variables; the level of the tariff, the sum of costs of staff lawyer salaries, benefits and overhead, and the productivity of the staff lawyers. A management strategy designed specifically to achieve productivity is necessary. Ultimately, cost-effectiveness must be achieved through productivity management. It does not come about automatically through some mechanism akin to the hidden hand in Adam Smith’s economics.

The conclusion that staff lawyer delivery of criminal legal aid can be less expensive, and with no compromise to quality of service is well established. Other areas of service delivery are different from criminal legal aid, however. For example, family law cases are not so structured as are criminal cases. The issues in a family law dispute may be more complex and emotionally charged. Family law cases may be more protracted as disputes evolve over time. Legal matters in other areas of civil law would also be different from criminal law matters. The conclusions from the research in criminal legal aid delivery may not be generalizable to other areas, or at least all aspects of those types of service. Research comparing the cost-effectiveness of staff and judicare delivery in areas of legal aid service other than criminal should be carried out. The complex mixed model concept suggests, at least for heuristic purposes, that staff lawyer delivery may be better for some aspects of service delivery than for others. This specialization of delivery modes should be explored in any further research.

The delivery model debate that occurred in Canada for nearly two decades reflects an interesting history. One of the lessons that emerges is that a delivery model represents more than simply a technical delivery option. In Canada, and elsewhere, delivery models have become burdened with a set of vested interests that have obscured the issues of cost and quality of service with rhetoric and ideology. The delivery models debate would be interesting if it were merely a ‘dust up’ between ideological foes driven by professional ethics and self-interest. It has been more than that, however. The proponents of private bar delivery have made arguments that cannot be dismissed out of hand. They have argued their case on the basis that accused persons should have a right to legal counsel of their own choice. Right to choice of counsel, it is argued, will result in the best quality of service because clients will choose the best available counsel, and that the solicitor-client relationship will be better by virtue of the trust that the client will place in a lawyer of his or her own choice. As well, it is argued that despite typically low legal aid tariffs, a private bar lawyer is more likely to mount a more aggressive defence on behalf of the client, since the private bar lawyer is free of any potentially compromising systemic relationships with the state. Nonetheless, the manner in which the delivery models debate has played itself out in Canada has stifled innovation and has been an impediment to the development of legal aid. Efforts by legal aid plans and policy makers to implement innovative pilot projects were sometimes resisted by the proponents of the existing delivery model.

Of course, one can not know if conditions in countries that are beginning to develop legal aid systems might develop in a manner so as to create the same constraining effect on innovation as has been the case in Canada. At least, the Canadian debate offers a ‘cautionary tale’ for those at the early stages of developing a legal aid system.

Even as the old debate about the advantages of staff lawyer versus judicare delivery lingers on, interesting new developments are emerging in delivery models. The simple mixed model of staff lawyer and judicare delivery that has framed the debate for the past twenty years is becoming eclipsed by the concept of the complex mixed model. The complex mixed model embodies the concept that neither the staff lawyer mode nor the judicare mode is categorically the best delivery option for all delivery problems. A legal aid delivery system must respond to a variety of delivery problems, reflecting different client groups, geographical factors, and types of service. Legal aid plans in Canada have recently begun experimenting with other delivery modes; contracting, specialized panels, expanded duty counsel designed to address particular problems in legal aid delivery with more focused delivery options. The research and development agenda with respect to complex mixed models is to determine how a variety of delivery modes, targeted at specific delivery problems, can be developed into a complementary, mutually supporting, multidimensional legal aid delivery system.