Evidence of co-production in public service provision: the case of the administrative arbitration centre in Portugal

Co-production includes all actions where citizens assist, as volunteers, in the provision of services by public agencies in order to increase the efﬁciency and efﬁcacy of the public services provided. This practice, known as co-production, is being adopted by governments in the resolution of conﬂicts, particularly those regarding administrative and ﬁscal matters. However, is co-production a more efﬁcient and effective way of settling disputes in administrative and tax areas than the traditional administrative model? And why? In Portugal, the Administrative Arbitration Centre was created in 2009 with the aim of resolving disputes between public administration and taxpayers/service users by means of co-production. The available data support the thesis that efﬁciency and efﬁcacy are higher under the co-production model. Nevertheless, users are not totally satisﬁed since the costs associated with the use of this service provision model are also higher


Introduction
The engagement of citizens in the production of public services has recently become an important topic in the study of public policy and public management (Eijk & Steen, 2014).This engagement has taken on the concept of co-production, which here we assume to be all actions in which citizens assist public agencies on a voluntary basis in order to improve public service provision (ibid, p. 2).These actions on the part of the citizens include any contribution in terms of time, effort and information to provide public services or produce goods (Alford, 2009).The term co-production first appeared in the work of Oström and her colleagues in the 1970s.This concept was first used to describe the possible relationship between those who deliver services (civil servants) and the users of these services.These users thus contributed their knowledge to improve the services they used.In this sequence, "by co-production, (. ..) (we mean) the process through which inputs used to produce a good or service are contributed by individuals who are not 'in' the same organization" (Oström, 1996(Oström, , p. 1073)).
The current situation results after the influence of New Public Management which promoted privatization and contracting of public services to the private sector (Silvestre, 2010).Although there are several criticisms of this movement, its ideas still find a lot of support nowadays among the governments of various countries and their defenders (Joshi & Moore, 2004).Originally, this approach promoted power in professional management; the use of explicit measures of performance; greater emphasis on controlling the product; the breakdown of units; increased competition; emphasis on management styles practiced by the private sector; and greater discipline and parsimony in the use of inputs (Hood, 1991).What is certain, however, is that no one knows specifically which organizational arrangements are best suited to providing high-quality, efficiently devised public services whose results can be substantiated (Alford & O'Flynn, 2012;Hodge, 2000).In general it can be stated that "the general belief of the experts is that most countries are agnostic" (Joshi & Moore, 2004, p. 31) regarding the rights and wrongs of this approach.
The approach proposed earlier was based on the susceptibility of choice of service users who, by their action in selecting which goods and services they would like to use, would increase competition among public organizations operating in the same sector of activity (Araújo, 2013).With greater competition, one could then make better use of scarce public resources and thereby reduce organizations' costs while increasing services quality.At the same time, there would be even more effectiveness when users evaluated the performance and measured the quality of organizations that provide public service.These organizations would theoretically be committed to a comprehensive reform both in terms of their structure and of their culture, as is the case in the private sector (Jung, 2010) where customer satisfaction is paramount.Citizen thus became citizen-consumers, or clients, exercising their choice in the consumer society (Clarke, Newman, Smith, Vidler, & Westmarland, 2007).
One of the major criticisms of New Public Management concerns its theoretical basis, i.e., it is badly defined and built (Araújo, 2004).According to Radnor and Osborne (2013) the movement's theoretical influence was based on the experience of a generic management theory whose contributions essentially resulted from the experiences of the private sector and whose activity, in turn, was placed at the manufacturing level, where the product was considered crucial (Radnor & Osborne, 2013).According to the authors, public services deal primarily with services rather than products, so the theoretical support employed in public sector reform is, under this theoretical perspective, poorly considered.They also clam that while products are tangible, services are intangible; hence the observed inconsistency, because the main judges of services should be their users not those who produce them.This inconsistency can be seen in the reform proposed, which valued issues related to efficiency and effectiveness, because a reduction in public spending was required (Osborne, 2013).This occurred through the use of tools and models originating from the private sector, such as the LEAN model, Six Sigma, Process Reengineering and Total Quality Management.Although the focus of these models and tools tends to promote the service user as a key part of the organization's action, in reality they were overlooked in providing those services.Thus, instead of listening to users, this reform gave priority to the evaluations of the civil servants in the operational base -those who directly provide those services to users -and with achieving the goals decided upon by their superiors (Radnor & Osborne, 2013).In the end, the emphasis was on the mode of operation of existing public organizations and on trying to change the way they operated internally, rather than on the intention and necessity of meeting the real needs of their users.
To rectify the theoretical deficiencies identified in the public management model proposed by New Public Management, there emerged the governance model (Araújo, 2013;Osborne, 2010), particularly co-production of public services as a mechanism for providing those services.Co-production, as was already mentioned, has become an important topic in this area of knowledge because it requires the engagement of citizens in the provision of public services (Eijk & Steen, 2014;Alford & O'Flynn, 2012).This approach results from the inefficiencies of the previous model, one that did not consider user requirements although theoretically that premise did exist.Based on the traditional model of public services provision, in this case the administrative courts, we can ask the following research question: Is co-production a more efficient and effective way of settling disputes in administrative and tax areas than the traditional administrative model?And why?
The research question tends to connect three central concepts in this study: co-production in regard to efficiency and effectiveness in providing public services.The analysis will focus on a research design of the case study type (Yin, 2009).According to Yin (2009), case studies are very useful to analyze the regularity of a particular social phenomenon.To this end, the aim of this study, in a comparison with the traditional administrative model, seeks to describe and understand whether co-production is more efficient and effective in the provision of public services when based on the participation of its users.In this particular case, this description consists in demonstrating how individual and collective citizens' participation has evolved in regard to the resolution of conflicts with the tax authority in Portugal.It should be noted that this study focuses on the Administrative Arbitration Centre that was created in Portugal in 2009 with the aim of resolving tax-related disputes between the Portuguese government and users of the services.This option is justified by the lack of studies of similar cases.After a description of the phenomenon, we will seek to understand the reasons that explain the relationship between the concepts now under analysis, by means of a survey among the arbitrators of the abovementioned arbitration centre.This paper starts by presenting the framework of coproduction, namely the conceptual characterization of this approach.Next, the methodology seeks to define the choices of the research.Then we will show the results and discussed them based on the approach that is used, concluding the paper with a final reflection.

Co-production of public services
As mentioned above, the concept of co-production used in this study is one that considers all actions in which in which citizens assist public agencies on a voluntary basis in order to improve public service provision (Eijk & Steen, 2014).The collaboration of citizens in providing public services has existed for many decades in the public sector (Bovaird, 2007).According to Pestoff, Osborne, and Brandsen (2006), the term refers to a citizen's participation in the production of public services, where that participation is delimited.Although the authors do not mention it, co-production has commonly begun to be used in the production and provision of services by private organizations.
Unlike public organizations, private organizations have for a long time kept up a long-standing "a relationship of exchange that is affordable, voluntary and direct" with the users of their services because users are not obliged to in fact purchase those services (Jung, 2010, p. 442).This relationship is based on the interests of consumers who, in turn, are willing to demand more and better conditions from their service providers (Hilton & Hughes, 2013).According to Hilton and Hughes (2013) these requirements stem from consumers' needs and expectations, such as speed, convenience and affordability when providing the service, as in the case of automatic payment at supermarkets, self-service cheek-in when boarding aeroplanes (Hilton & Hughes, 2013) or banking operations online or at ATMs.The authors would like to note that in these cases, such practices depend on two requirements in particular: on consumers' demand for better services and at the same time being willing to use new technology, for example; and also on their ability to acquire the equipment necessary to access these servicesconsidering here the services that require information technology (IT).At the end of the process, customers who use IT will become partial collaborators because they help private organizations to improve their services based on the former's experience.There is, however, one aspect that stands out from these experiments: the provision of services is regular and occurs through a supplier-service professional relationship (whatever the sector may be), where all parties contribute with their resources (Bovaird, 2007).
The provision of public services is not exclusive to public administration.There could be countless combinations of forms of service provision that were especially developed in recent decades (see Savas, 2000).Whichever combination is analyzed it is common to identify numerous actors, apart from public authorities, engaged in providing such services, especially the "engagement of service users as co-producers of the services they receive" (Pestoff et al., 2006, p. 592) who in different ways (collaborating in the production process, offering their time and/or information) contribute to the provision of a public service or the production of a good.According to the authors, the reason for this lies in the need to consider the users of public services as a fundamental element.This happens because they will be the ones who define what they want to use, as well as the quality of what they are provided with.Without their consultation or engagement, it is difficult to consolidate this.For example, in the case of health services the collaboration and participation of the patient throughout the process is critical to the success of medical procedures.The same goes for recycling of household waste, where the collaboration of citizens' collaboration in separating waste is critical to the success of this policy.

The co-production of public services: concept
The concept of co-production has an inaccurate and ambiguous meaning in the literature.The truth is that co-production can take on various facets and moments, for instance: cogovernance, where third-party actors participate in planning and providing public services; co-management, where the thirdparty actors collaborate with the State in providing public services; or co-production (not in a narrow sense), which refers to citizens' participation in producing, at least in part, the public services they will benefit from (Pestoff et al., 2006).According to the authors, the latter type can include services where the relevant public authorities fund and regulate them but are not directly involved in providing them.
In Joshi and Moore (2004)'s view, co-production is defined as any action that involves service users.In general, this involvement is not perfectly typified, since it is carried out on an informal basis between users and State agencies.When this informality ceases to exist due to the formalization of a specific law, it becomes institutionalized co-production.This institutionalized co-production in turn requires that the engagement of users in the provision of public services be continued between public agencies and users, so it is not informal as a result of the long-term relationship between the parties.They also suggest that this does not require a contractual relationship between the parties involved, as is claimed by the advocates of New Public Management.This is because contracts proscribe voluntary contribution and this cannot be confused with commercial types of arrangements such as public-private partnerships, for example.
According to Eijk and Steen (2014), the ability and willingness of citizens to participate in improving the provision of public services has been shown to be essential for this model to emerge as a credible and feasible alternative, and is dependent on the human and social capital of the participants.Human capital is based on socioeconomic variables such as the level of education or financial capacity of its members, while social capital depends on each member's availability to contribute (Eijk & Steen, 2014;Joshi & Moore, 2004).
In the same sequence, Bovaird and Löffler (2013, p. 100) base their opinion on Governance International (2011), for whom co-production aims at "the public sector's use of the assets and resources of users and communities to achieve better impact and lower costs".According to them, the participants in the process should possess some essential characteristics, namely that service users, thanks to the actual contact they have with those services, know how to identify the features that can be improved; and if they are able to do so, their contribution may help to improve the service.If the contribution is dependent on the willingness of users and there is no contractual basis between the parties, the organization will benefit from this knowledge without having to offer a financial reward for it.The citizen can influence and improve the service, but also contribute to the improvement in the quality of life of the entire community he belongs to.Finally, the rule is broken according to which public organizations should adopt a paternalistic relationship with the users of their services because they become participants; partners and not just mere recipients of public services.
The position of a user of public services in this model takes on different contours and features compared to those the former model.In the previous model, those who made use of the service should be considered 'clients', as is the case in the commercial relationship that individuals have with private organizations, whose main purpose is profit (Silvestre, 2010).The term 'client' means that there is a direct relationship, by means of a payment, between those who consume a certain product or service and the organization that provides it; therefore, there is a direct relationship with the organization's profit (Radnor & Osborne, 2013).Because of their importance, 'clients' own the empowerment as a result of the financial consideration involved, of selecting what they will make use of, having the choice reject or consume a certain good or service (Jung, 2010).Instead, citizens who are users of public services should take a different approach because they may be interested the rewards arising from their participation, the solidarity enjoyed with other members of their community, or just the satisfaction of contributing to a cause that they feel is just (Eijk & Steen, 2014).
To sum up, we can assert that the engagement of public service users has gradually been taking place in recent decades, and that lately the dynamics have be growing in this approach.This is due to the increasing scarcity of resources, which does not allow them to be wasted (Eijk & Steen, 2014).To avoid this waste, it is important that the main beneficiaries of public services take an active part in the definition of those services.It is also a result of the failure of New Public Management to engage citizens and foster their participation in public service provision (Araújo, 2013).If users can voluntarily contribute to make this happen, it is theoretically advantageous for public authorities because by gathering expertise they can improve the quality of services.It will also aid in the construction of a more plural and participatory public administration.For these reasons, it is pertinent to examine a specific case where such a contribution has been given.

Research design and selection of the sector
As mentioned in the introduction, and based on the research question and its overall objective of study, the research design used will be of the case study type (Yin, 2009).According to the author, this design is appropriate when studies are exploratory in nature and have research questions like: What?He also maintains that such research design does not require control of events and focuses on contemporary occurrences, so it should consider internal validity (through the respondents' understanding of the answers they submitted) and external validity (whose goal is the correct connection between the tool used for data collection and the conceptual approach used) (Bryman, 2004).Internal validity is thus ensured since the respondents are within the concept of co-producers as defined by Oström (1996Oström ( , p. 1073)): the "contribution of subjects who do 'not' belong to the organization".External validity was assured too, because the questions that respondents were asked refer to the understanding of two distinct forms of dispute resolution within the administrative and tax field -arbitration courts and arbitration centres -regarding their efficiency and effectiveness.In that sense, respondents do have the knowledge from the both types of service provision selected for this analysis.
For the purposes of this study a specific type of co-production was selected: co-planning and co-use, i.e., the partnership established between organizations/subjects to create the arbitration court and use that both public and private users make of it.This option is justified by the dearth of studies on co-production outside Anglo-Saxon countries, whose social, political and economic proportions are markedly different from other realities, including the Portuguese one.Moreover, this is an exploratory study in which the analysis is performed in view of the agents of the arbitration courts involved in providing the available services.It should be noted that responding agents are real sources of knowledge, because they are experienced users as well as providers of services (they are judges of the administrative courts).Thus, these agents are able to assess whether this solution using co-production is more efficient and effective than the traditional model of conflict resolution.Based on real and specialized knowledge, it is justifiable to ask them, which is especially relevant for this study internal validation.

Data collection and sources
For this exercise to be possible, the data was initially collected from existing documentation and files (Hood, 2011) in order to describe the social phenomenon that will be the aim of the study.According to Yin (2009), the great advantage of these sources are: they are stable because they can be constantly updated; they are not obstructive since they does not result from previous case studies; they are accurate because they contain names, references and details of the social phenomenon to be asked about; and they cover a wide range of data over long time periods and various events.However, there are some weaknesses in the selection of these sources, especially because: they are difficult to access; it is possible to select information that does not contemplate or capture the entire reality; and possibly because much of this information is not available.
For the purposes of the advantages and disadvantages in choosing these documentary sources, we took special care to access the databases available on the Internet which include all legal and judicial development, justification and intervention of subjects and available data that are actually new methods of initiating study in this area of knowledge (see Hood, 2011).This way, the weaknesses were cancelled out and we contemplated a more accurate context so that the solution can be properly structured (Pollitt, 2010).The validity of the information is ensured by this procedure and we essentially sought statistical types of data.These data will be presented only to attest to the evolution of the phenomenon under discussion here.Based on the central research concepts, it was possible to access the statistics provided by the Directorate-General for Justice Policy (http://www.dgpj.mj.pt/sections/estatisticas-da-justica, last accessed on 3 April, 2016).Here we collected data that pertain to the traditional model (Supreme Administrative Court, North Central Administrative Court and South Central Administrative Court) and to Alternative Dispute Resolution (including aggregated data from Arbitration Centres, Courts of Peace and Mediation).About these, we collected data on the efficiency rate (which translates into the number of completed cases versus the number of new cases and cases carried forward from the previous year) and the average duration of completed cases, to ascertain efficiency; and the resolution rate (if the indicator indicates a rate higher than 100%, this means that the number of completed cases is greater than the number of new cases), to ascertain effectiveness.The data cover the 2010-2014 period, to enable comparison with CAAD data.
At the second stage of data collection, we used the survey tool of self-administered questionnaires directed at the judges of the arbitration court.According to Bryman (2004), this type of survey is filled in directly by the respondent.These surveys can also be self-administered with or without an interviewer present.In this specific case, it was carried out without an interviewer because IT (the Internet) was the vehicle used, particularly the sending to respondents' e-mail.A pre-test was conducted by means of the sending of the survey through a questionnaire to three judges on 5 June, 2014.The responses were received by 15 June, 2014, and these enabled us to ascertain issues related to the advantages, disadvantages and difficulties of Administrative Arbitration Centres, their results, the participation of users and the success factors of participation, compared to the traditional model.
The survey was of the structured type (Bryman, 2004).We took care to ask questions that covered the area of the respondent's own point of view and that can later help in the perception of the social phenomena under study.Once the pre-test had been carried out, 24 surveys by self-administered questionnaire were sent on 20 June, 2014, to all judges-arbitrators of the arbitration court.The anonymity of respondents was ensured, and 16 duly completed surveys were received by 26 July, 2014 -a 66.7% response rate.The judges were selected among agents of the coproduction process who were not employees or contractually connected to the Arbitration Centre.These are totally independent professionals from various areas of expertise -university professors specialized in taxation, former counsellor judges and associate judges retired from fiscal courts (the Supreme Administrative Court, for example), and taxation law consultants, among others -who are knowledgeable on the reality of arbitration courts in comparison with traditional administrative courts.
It should be remembered that, according to Yin (2009), one of the advantages of using primary sources to collect data collection is the accuracy of the data regarding the names involved.In this particular case the names of the arbitration centre's judges are identified on the centre's website, which facilitates contact with them.Moreover, again according to the author, one of the disadvantages is that it is not always possible to have access to them.In this specific case, the response rate is high, particularly keeping in mind the views of people involved in the process.There are three stakeholders in this process of creation and use of the arbitration centre, namely public administration in the field of tax, taxpayers who are to some degree in dispute with public administration, and also the aforementioned arbitration judges.These judges, as will be shown below, have a duty of independence, impartiality and integrity, being neither representatives nor agents of any party that appoints them.Therefore, they are privileged actors experiencing this social phenomenon because they know the traditional model of public service provision -in this case, the fiscal courts as agents of those services -and because they perceive the advantages and disadvantages of this model regarding the service provided to users.

Traditional administrative courts and alternative dispute resolution
The Laws on Voluntary Arbitration (Law No. 31/86 of 29 August), on the General Contractual Clauses (Decree-Law 446/85 of 25 October), on Organization, Competency and Operation of Courts of Peace (Law 78/2001 of 13 September), and on the Experimental Procedure Scheme (Decree-Law 108/2006 of 8 June), were turning points in the Portuguese legal system in regard to the modes of provision of services aimed at alternative conflict resolution in relation to Traditional Administrative Courts.The general principle of exclusion of arbitrability in administrative litigation was in force for centuries in the national legal system, as a result of the French model of public administration.The latter is seen as a power directed at the pursuit of public services provision.Consequently, the rule of law is observed, where public administration enjoys a higher legal position than citizens (Serra, 2010).Disputes that emerged were usually resolved within public administration itself, without even contemplating the possibility of resorting to other forms of provision (Catarino & Filippo, 2011).
The use of other forms of provision was not allowed on the grounds of the general idea that public administration could not freely have rights and public obligations at its disposal, since it was committed to the primacy of public interest and the principle of legality (Serra, 2010).The State could not conceptually be present in public relations merely as a party, but as primus inter pares (Catarino & Filippo, 2012).
However, the widening of the State's functions generated new social rights, made public management more complex and required the participation of citizens in public administration.It was only in 2004, with the reform of procedural law in administrative courts, which gradually replaced the Portuguese model of objectivist justice to control administrative legality by means of a German-rooted subjectivist model for the protection of citizens' subjective rights.Arbitration, for example, among the possible alternative dispute resolution methods, arises in tax law in a context where the intention is to avoid conflict or as an attempt to resolve such conflicts between states and their citizens (Catarino & Filippo, 2012).
In fact, the State is starting to become more open to ideas of co-production in its various forms (Bovaird, 2007).Such citizens' rights have been complemented, in administrative law, by establishing mechanisms for prior consultation.These rights envisage the participation of citizens in the political and administrative activities of the State.This citizen engagement has generally been shown to be an effective and efficient means in providing conflict resolution services.By analysing Table 1 and considering the period between 2010 and 2014, Alternative Dispute Resolution demonstrated higher efficiency rates than the average of traditional courts: 39.77% in 2011, 41.82% in 2012, 44.67% in 2013 and 39.83% in 2014.The exception is 2010, when the average efficiency rate of traditional courts was 84.65% versus 78.56% for alternative means.It can therefore be said that alternative means are more efficient than traditional ways of providing this service.Due to a lack of available data, the average duration of completed cases could not yet be analyzed.
If we consider the resolution rate, it appears that the figures for traditional courts are as follows: 84.65% in 2010, 79% in 2011, 88.37% in 2012, 112.34% in 2013 and 91.60% in 2014.The figures presented by Alternative Dispute Resolution have higher resolution rates in all the years considered except for 2013 -see Table 1.It can thus be ended, again by comparison, that the effectiveness alternative dispute resolution is greater than that of traditional means.
However, this analysis should include users' assessments of the service provided by alternative means.This was done by the Directorate-General for Justice Policy (2013), which conducted a survey with 505 users and, using a scale of 1 (very dissatisfied) to 10 (very satisfied), ended that dispute settlement speed (average of 7.99), cost of access (average of 7.99) and punctuality in starting sessions when planned (average of 7.94), were the indicators with the lowest scores among those considered in the inquiry.In contrast, the assessment of the work of the judgesarbitrators, justice of the peace or mediator (average of 8.65) and operation of the arbitration centre, court of peace, mediation (average of 8.43 except, and as noted above, punctuality at the start of sessions) were the indicators with the highest scores.In short, alternative dispute resolution shows better rates of efficiency and resolution than traditional courts.There is, however, less user satisfaction when considering associated costs and the speed with which cases are resolved.

The Administrative Arbitration Centre
The adoption of tax arbitration in Portugal as an alternative means to traditional public means (the Courts) for conflict resolution is based on the general idea that it is beneficial for the public sector to use the resources of users and communities to achieve the best results at a lower cost (Bovaird & Löffler, 2013).The resolution of tax disputes through alternative means was demanded in Portugal by the organizational structures of d The average duration of ended cases corresponds to the time between the case's entry date and the date of the final decision (judgement, sentence or order) at its jurisdiction, regardless of res judicata.
economic agents who opposed the monopoly of public justice because it was long-winded and expensive, and did not safeguard their legal rights (Catarino & Filippo, 2012).
Thus, a new non-profit private jurisdiction was created in Portugal in late 2009 -the Administrative Arbitration Centre (CAAD)which theoretically would enable swifter resolution of the issues that previously had to be compulsorily submitted to the Fiscal Courts.This jurisdiction operates under the aegis of the Higher Council of Administrative and Fiscal Courts (CSTAF) and its founding partners are public, private and third-sector organizations.This model enables the connection between the business world and the political, social and tax systems, in particular with the Government and Public Administration, Associative Summits, schools and universities, and the financial and business community.The law allows any public or private entity to conclude association and cooperation protocols with the CAAD, becoming part of its Board of Representatives.
Arbitration is optional, so that taxpayers can choose which of defence channels (judicial, public, or arbitration, private) they wish to pursue.It is a private jurisdiction where neutral and impartial third parties (judge-arbitrators), chosen by the parties or appointed by the CAAD, decide on the issues in dispute, where decisions have the same legal value as court rulings.The engagement of citizens in the process of selecting the arbitrators is an innovation that allows for greater engagement in the conflict resolution process, according to what is proposed by co-production (Pestoff et al., 2006;Eijk & Steen, 2014).Arbitration theoretically has the advantage of swiftly obtaining a decision binding on the tax authorities and on the taxpayer, similar to decisions handed down by a fiscal court of law, as well as ensuring, in some cases, the possibility of appeal.Since the implementation of arbitration as an alternative means for resolving tax disputes, it appears that the number of cases submitted has been increasing, as can be confirmed in Table 2.
Conceptually, the political disposition for the creation of arbitration in Portugal is connected to the attempt to improve the provision of public services both in terms of speed and of quality (Eijk & Steen, 2014;Joshi & Moore, 2004), working from the assumption that service users have access to the necessary resources and the will to achieved this (Bovaird & Löffler, 2013).
On this assumption, and taking into account structural and operational processes, the CAAD is characterized by the following (see Figueiras, 2011): the human resources centre includes a range of specialized technicians who are responsible for all administrative procedures regarding the cases.Therefore, all the steps taken are carried out by administrative human resources who belong to the centre.As stated above, the judges-arbitrators are chosen from a set of personalities that are listed on the centre's website.It should be noted that these judges-arbitrators are not employees of the centre; they only assist in fulfilling the purposes that the centre sets out to achieve.Should the parties who use this service decide to choose a judge-arbitrator, experts, etc., other than those listed by the centre, it is certain that the costs arising from this will be borne by the parties.Despite these increases, "in theoretical terms one actually see a downward trend in arbitration costs in relation to ordinary Courts, at least when the parties do not appoint an arbitrator who is not part of the Centre's list" (Figueiras, 2011, pp. 127-128).
Given these theoretical advantages, the number of cases resolved in that centre has been increasing.This is largely due to the speed at which they are concluded, which is faster compared to other courts that are considered traditional -see Tables 1 and 2. So the time taken to resolve both administrative and tax conflicts is on average four months and twenty days, while in other courts that average is higher -see Table 1.These figures prove that the co-production is more efficient and effective than the traditional provision of public services, as stated in studies on co-production (Eijk & Steen, 2014;Bovaird & Löffler, 2013;Joshi & Moore, 2004).However, one cannot overlook the assessment of CCAD's services, made by its users (see Directorate-General, 2013).Referring now solely to the opinions of the CAAD's users (n = 337), it is clear that they are most satisfied with the performance of the judge-arbitrator responsible for their cases (an average of 8.54).However, they seem less pleased about the swiftness of conflict resolution (an average of 7.77), the cost of access (an average of 7.97), the time between the request and the first session in which they were heard (an average of 7.79), and the punctuality when starting the sessions (an average of 7.87).
According to the judges-arbitrators, the CAAD has advantages over the traditional model of conflict resolution.Of the 16 interviews, all respondents mention the speed of decisionmaking as the main advantage of this model: "We must not forget that 'good justice is always swift justice" (respondent 2); "Under current circumstances, I consider that the main advantage offered by the Centre is the gain in decision time-compared to the saturation and the consequent slowness of administrative and fiscal justice" (respondent 16).
According to respondents, this swiftness is due to the simplification of the process (respondents 7 and 9), which does not require the procedural steps that take place in traditional courts and that generate the slowness shown in the traditional model.On the other hand, beyond the questions of speed and procedural simplicity, the choice of experts to settle disputes between parties is also a great advantage (respondents 3, 5, 7, 9, 11, 12 and 14).This is one aspect where the use of the co-production model in the CAAD produced great benefits.Since the judgesarbitrators are recruited among renowned experts by means of engaging the parties, the average quality of decisions tends to increase because they are legally based (respondent 12).Moreover, in addition to technical quality, integrity and impartiality in decision making are other criteria that are mentioned (respondents 4 and 12).The engagement of the parties in the arbitration process by choosing the judges is an important contribution to the entire process, as is suggested in the literature (Eijk & Steen, 2014;Joshi & Moore, 2004) and this can be seen here, i.e., on average decisions really are faster.Nonetheless, given the costs involved and associated expectations, the CAAD's users do not give such positive scores to the speed of processes.
Even so, these data corroborate one of the main reasons for the creation of the CAAD, which was to swiftly obtain a decision that would be binding on the parties in the case.We would point out that co-production aims to create a possible alternative in order to improve the provision of public services (Joshi & Moore, 2004).Thus, and according to Joshi and Moore (2004), users should participate in the design, implementation and evaluation of the actions and policies adopted.
Although several advantages are enumerated, some disadvantages are also identified.According to respondents, these disadvantages are: the continuing demand there may be for the centre's services and whether this will be matched by increases in logistical and human capacity so as to maintain the current decision-making speed (respondents 1 and 9); and this difficulty necessarily means recruiting judges-arbitrators (respondent 2).
Another of the respondents concerns is the possible lack of impartiality, independence and integrity of judges-arbitrators (respondents 3, 4, 11 and 12).In addition, despite the recognized technical quality of decisions, the judges-arbitrators may also deliver poor-quality technical decisions because it can happen (respondent 4).This disadvantage is associated with another concern -once the CAAD's mediation is accepted by users the decision must be accepted by the parties and this compromises some individual guarantees regarding appeals against the decision, which are provided for in fiscal courts (respondents 7 and 15).But if the parties are involved in the arbitration process by appointing judges, this will foster greater acceptance of the decisions handed down.Finally, costs are higher for users (respondents 5, 12 and 13), which partially contradicts the use of co-production.According to Bovaird and Löffler (2013) coproduction aims to increase the impact of policies on citizens' lives, particularly by lowering costs in the provision of public services.We found in this case that the swiftness is real and that this impacts the pending issues of service users; however, it falls short regarding the reduction of costs, which prevents many users from resorting to this method.This is consistent with what is Figueiras (2001)'s claim that the costs of arbitration should theoretically be lower when compared to the traditional model.This study shows that the main reason for dissatisfaction among the service's users is essentially the costs involved (see Directorate-General, 2013).
Despite the fact that some disadvantages were identified in comparison to the traditional model of conflict resolution, most respondents assert that the CAAD is a model that provides swifter results (except for respondents 7 and 11, for whom, respectively, it is premature to make a real and reliable comparison, and the technical decisions do not seem to be higher in quality).For the remaining respondents this arbitration model provides greater advantages than the traditional model because there is greater agility in hiring judges (respondent 1).This agility results from the engagement of the parties provided for by the co-production model (Hilton & Hughes, 2013).These judges are independent, and according to the respondent 7, demonstrate higher productivity than the career judges of administrative courts.
The last advantage pointed out in this arbitration model is the citizen's ability to be present at all the procedures of the case (respondents 1 and 8).This reason is consistent with Bovaird and Löffler's (2013) assertion: co-production enables users to employ their assets and resources; in this particular case the users can voluntarily add data to the process, thereby involving them in it, which aids decision-making and the swiftness and quality of the decision.
User participation is therefore the biggest asset and innovation in the creation of this arbitration centre.This asset is provided by co-production, as it fosters closeness among all parties regarding the case procedures (respondent 7): "The biggest advantage is the proximity to citizens and the possibility they have of directly following the case through the judges-arbitrators they appoint or the arbitrators appointed by the centre" (respondent 13).
Another innovation is the selection of judges-arbitrators by those engaged in the dispute (respondent 1), and that users commit to accepting and following the decision that is handed down (respondent 11): "User participation is via the judges-arbitrators contained in a list that is pre-approved and made public.These are people of recognized merit in judging on the issues submitted to the arbitration court, so when users resort to arbitration to settle disputes, they already accept the participation of any of these judges-arbitrators in the arbitration proceedings, relying on them to consider the issue and to make judgement on it" (respondent 12).
It should be noted, however, that among the new cases only in fourteen of them did the parties request the appointment of the arbitrator, while in three hundred and five cases there was no request for the arbitrator, who was chosen from among the independent experts.
User participation is not limited to monitoring the process.The creation of this arbitration centre was based on the visibility and reputation of decisions (respondent 9), and these decisions are highly dependent on what is offered by the users themselves.In designing the model, the parties involved can submit proposals for improving services to the centre's Ethics Council (respondent 6).Besides: "(. ..)I think that the participation of users -or, I would say, of the organizations that represent them, rather than of individual users -could be placed precisely at the 'design' and 'implementation', and then at the 'monitoring' stages.The existence of something like a follow-up council consisting of representatives from the organizations of persons and entities that most often request the Centre's intervention, as well as the administration sectors who accede to it, could be a 'monitoring' structure capable of assessing the operation of the Centre and, depending on their experience, suggest and propose the appropriate adjustments" (respondent 16).Such a provision is consistent with what is argued by Eijk and Steen (2014), for whom co-production refers to actions where citizens assist public agencies as volunteers in order to improve the quality of the services provided -which actually happens in this case because users participate in the co-production (not in a narrow sense) of the services that they will use (Pestoff et al., 2006).Some respondents (4, 11 and 12) maintain that, in order for service users to go on benefitting from this service, it is necessary to ensure there is a legal and material framework that is perfectly clear and appropriate.With the existence of this framework, they also argue that it is important to separate and control functions (respondent 3), and that each process can be adapted to the case being tried (respondent 5).This would guarantee the centre's integrity, reputation, credibility, efficiency and effectiveness (respondents 9 and 13).In conclusion, the administrative arbitration centre emerges as an innovative model, not only due to the services it provides but mainly as a result of the procedural and administrative participation of service users.This characteristic sets it apart from arbitration courts, so that: "(. ..) the quality of the services provided -the speed and correctness of the decisions -is an essential factor that the CAAD has practised, in contrast to the very poor service provided by the fiscal courts" (respondent 15).

Conclusions
Given the scarcer resources that States have their disposal to provide services that are fast, efficient, high-quality, and that meet the necessary requirements of its users, co-production has emerged as an alternative means of catering to the collective needs that are individually experienced by populations.The coproduction is understood as comprising all actions in which citizens (individually or collectively considered) assist public agencies on a voluntary basis in order to improve the quality of public services (Eijk & Steen, 2014).
In addition to the limitations of traditional models of service provision, due to citizens' needs and to favourable political and institutional conditions, this way of providing public services inspired, in Portugal, the creation of the Administrative Arbitration Centre in 2009, so as to settle disputes between the administrative and fiscal public authority and its users.The CAAD is private in nature and, by engaging the various parties, has increasingly gained prominence in conflict resolution -the number of cases submitted to it rose by over one thousand percent between 2011 and 2014.This is explained by its swiftness in decision-making, which would not be possible otherwise (Hilton & Hughes, 2013).In addition to the speed of decisions, there are other advantages such as the ongoing engagement of the service user through the appointment of the judges-arbitrators, the procedural monitoring and also the ability to improve the entire service by consulting users'.
However, there are two factors that should be considered.The first refers to the issue of cost, i.e., the cost for users of the administrative centre should be lower.This is reflected in not very positive evaluations by users.As borne out by judgesarbitrators, cost is a barrier for individuals or organizations who would like to resort to this means of resolving existing disputes.This is because the amounts charged to users are higher than those of traditional means of conflict resolution.In spite of the greater efficiency and effectiveness of their service provision, arbitration centres do not receive entirely positive feedback from users, as demonstrated by the study conducted by the Directorate-General (2013).Therefore one of the challenges that should be taken into account is a reduction in the operating costs of this model, to enable citizens to use its services more often.
This case shows that the State has no monopoly in the provision of public services and that it is possible to find more agile ways of providing them by using co-production.Instead, citizens and society in general can themselves provide such services, thus benefiting from goods and services they really need and that help them to resolve everyday issues they may face.It should be noted that this is a preliminary and exploratory study, so ongoing attention should be paid to this particular phenomenon.This attention is due because we cannot yet clearly assert Joshi and Moore (2004)'s argument that co-production 'is the best of all possible alternatives' to improve the provision of public services through users' participation in the design, implementation and evaluation of public policies.Moreover, it is important to proceed with this analysis but this time focussing on the theory of services proposed by Osborne (2013), thus cementing theoretical knowledge in this field.This knowledge should also be expanded to other industries and other social, political and economic realities.

Table 1
Statistics on traditional courts and alternative dispute resolution.

Table 2
Number of cases submitted to Arbitration Courts in Portugal.